PDA

View Full Version : CS4397 dac



Pages : [1] 2

leo
16-03-2008, 16:35
I got one of these cheap dac pcbs to play about with, its based on the CS4397.

I've changed the PPL filter for the CS8416 as I noticed it was the wrong values, it now locks onto the spdif much better and sounds better too.
I have also changed the op-amp to a OPA2111 and bypassed the coupling caps on the dacs differential output
Voltage is being supplied from ALW super regs

Sound so far isn't bad, its quite a way behind both my diy TDA1541A double crown based dac and diy PCM1794 based one but we'll see how it goes with some more work, never been a huge fan of voltage output dacs but its all good fun:)

I'm going to build a discrete differential to line out output stage so to ditch the op-amp, also have a play about with the decoupling
Replace the phono in socket for the spdif to a 75R BNC and probably add a pulse traffo on the input

http://i273.photobucket.com/albums/jj220/kingbusoms/f14c4aa7.jpg

Edit added 01/06/09

I recently had a PM from a forum member under the name of Raptor , he's spent some time going through the mods and thought it would be helpful for others to compile list and info so here it is:)

- R2 becomes (BC Components minimelf) 3k, C22 becomes 22nF, and C23 becomes (COG/NPO) 1 nF ((Read datasheet for reccommendations))
- Remove and link the pads of the 4 DAC output capacitors, (2.2u 50v C27,C28,C30,C31)
- Remove and link the pads of the output capacitors near RGL socket (10u 16v C6,C54)
((Use your own MKP))
- Link the two exposed pairs of pads underneath on the top and top left of the small IC ((This bypasses it completely.))
- C9 (C7/C13 for TOSLINK) should be REMOVED and replaced by a Vishay/ERO MKP1837 0.1uF or equivalent. (Possibly also a small 0.1uF Ceramic dipped in hot glue to remove some microphonics. SHORT LEADS)
- Lift or unbridge pin 2 of the CS4397 and connect to Pin 16 of the CS8416.
- R8,R9,R14,R17 to 30k WELL MATCHED
- C32,C33,C34,C35 to 150pf SMD or Polystyrene
- R28 and R29 should be missing.
- C17,C20,C2 100uf-220uF Oscon
- C16 100uF Oscon
- C3, C18, C11 can be left alone (Just pulling pins high)
- C25 5.6!!-10uf Tantalum or 5mm MKP.
- C12, C8, C13, C10, C14(C41) bypass with MKP 0.1uF
- C19 should be 1 decimal step smaller than the last filter cap on your +5 supply. (1000ish Not OScon.)
- C21 should be 1 decimal step smaller than C19. (100-220ish Not OScon.)

- Tube out - http://theartofsound.net/forum/showpost.php?p=36311&postcount=325 or page 33.
- Digital out - Pages 22-27ish have a board layout referenced from earlier page's schematic. 30-35 have modifications to the components around the opamp socket for output. I suggest just pulling off the DAC's 2.5+v outs with a coupling cap or trafo.

"BTW don't always trust what you read regarding low ESR caps, their not always better in every application" -Leo

jonners
17-04-2008, 14:40
I've just ordered one of these after reading this: http://www.lampizator.eu/LAMPIZATOR/LAMPUCERA/CD%20DAC%20Lampucera%20lampizator.html
I will report back when I get it going. Meanwhile, I would be interested to know, Leo, what were the specific changes that you made to the PPL filter?

leo
21-04-2008, 17:11
C22,R2 and C23 is the PPL filter for the CS8416, I changed them to whats used in the updated datasheet 3k/1N and 22N , I used a minimelf resistor for the 3k and PPS for the caps

I don't see much point in adding a valve output stage to this type of chip, nothing wrong in it though I guess, beauty with diy:eyebrows:

Marco
21-04-2008, 17:55
Sound so far isn't bad, its quite a way behind both my diy TDA1541A double crown based dac and diy PCM1794 based one but we'll see how it goes with some more work, never been a huge fan of voltage output dacs but its all good fun:)


Interesting comments, Leo, and congrats on your latest DAC project.

What's your opinion on the TDA1541 compared to Burr Browns and Wolfsons? Ashley is of the view that Burr Browns are better because they measure better, but my ears, and those of others I respect, tell me nothing beats the TDA1541, whether single, double, or triple-crown. They seem to give music a 'weight' and 'presence', and an 'analogue' type of presentation (in a good sense) that the top modern DAC chips lack, which to my ears are often thin and 'bleached' sounding in comparison.

What are your views?

Marco.

leo
21-04-2008, 18:30
TDA1541A double crown is still my favourite chip to date, I've tried and heard so many different commercially made units using BB, Wolfson, AD, Cirrus etc, I've also built quite a few diy dacs using these various chips and I always come back to this classic, implemented correctly no other device has that realism in the midrange, decay of piano that this one has IMHO.

Of course theres a lot of units using this classic chip that sounds god awful, as everything else its the implementation which is why I only stick to diy with this chip.
Most components have moved on since these chips was first made and used in commercial equipment, problem regarding dac chips is the latest and so called greatest ones have the important stuff like oversampling, current to voltage etc locked inside the actual chip so your limited to what you can do, where as the older ones have it all external giving you far more to tweak about with

Just my opinion of course;)

leo
21-04-2008, 18:38
The only other dac I have kept hold of is a diy based around the BB PCM1794, its very good but I still prefer the old classic

The CS4397 I'm playing about with now is basically for a bit of fun on the cheap

Marco
21-04-2008, 18:41
TDA1541A double crown is still my favourite chip to date, I've tried and heard so many different commercially made units using BB, Wolfson, AD, Cirrus etc, I've also built quite a few diy dacs using these various chips and I always come back to this classic, implemented correctly no other device has that realism in the midrange, decay of piano that this one has IMHO.


I totally agree! I think I'll flag this one up as a sticky for Ashley :eyebrows:


Of course theres a lot of units using this classic chip that sounds god awful, as everything else its the implementation which is why I only stick to diy with this chip.


Yes indeed - implementation is all, but correctly implemented, the TDA1541 is the best DAC chip ever made bar none. It's quite simply the definitive choice for discerning audio enthusiasts :smoking:

Marco.

leo
21-04-2008, 18:56
I don't care too much what others say or think about this chip, until somebody comes along with a unit that I think sounds better than what I use I'll stick with this old classic;)
I give everything a chance even the hyped up stuff, still waiting to have my socks blown off though:lol:

Marco
21-04-2008, 19:01
Same as me, Leo. You are indeed a man of class :smoking:

Marco.

jonners
21-04-2008, 21:25
C22,R2 and C23 is the PPL filter for the CS8416, I changed them to whats used in the updated datasheet 3k/1N and 22N , I used a minimelf resistor for the 3k and PPS for the caps

I don't see much point in adding a valve output stage to this type of chip, nothing wrong in it though I guess, beauty with diy:eyebrows:

Thanks for your help.:)
John

leo
27-04-2008, 12:24
onboard output stage Implementation is now based on this http://www.diyaudio.com/forums/attachment.php?s=&postid=1444038&stamp=1204478039


I use minimelf resistors and smd PPS caps, AD826 seems to work well in this new stage where as it wasn't as good with the older implementation

http://i273.photobucket.com/albums/jj220/kingbusoms/DSCF0535.jpg

Yomanze
29-04-2008, 13:34
Haha Marco. Sharing a love of the TDA1541A, SL1210 and DL103 - spooky!

Marco
29-04-2008, 14:26
Two words sum it up, mate:

"Discerning ears" ;)

Marco.

nickthevoice
07-05-2008, 14:46
Me too I buy this after view web site CD DAC LAMPUCERA.

this man don't use and modify capacitor around opamp buffer output,

that's the best capacitor modification in ANALOG section and around OPAMP ?

First modification I change this OPAMP for LM4562 but AFTER ???

if you have complete list of your change in this DAC, because on LAMPUCERA web site it's complicated to know the brand and the type ??? WIMA MKP2 mixed with TANTALUM with OSCON oufffff ... ???

Is this a good idea to change all surface mount capacitor ! ?

thanks in advance ;)

nicK

Marco
07-05-2008, 15:17
Hi Nick,

Welcome to the forum :)

I presume you aren't English?

Marco.

nickthevoice
08-05-2008, 11:22
HELLO :) ???

Nobody want help me with my complete DAC KIT upgrade ? :(

:doh:

nicK

Filterlab
08-05-2008, 11:25
Nobody want help me with my complete DAC KIT upgrade?

I'm sure Leo will help you as he has a vast amount of experience in this field, however he's not on every day so you may have to wait a short while.

iihay
14-05-2008, 21:36
Just picked up one of these myself and stumbled in here looking for info. I'm a clueless noob so pointers on this dac would definitely welcomed by me as well. Looking forward to being here.

Cheers

Iain

alb
16-05-2008, 12:21
Any questions Iain, i may be able to help as well.

I also use one of these and am just about to start doing a few mods to it.
Very good value for money, i reckon.
Where are you up to with it?

leo
16-05-2008, 22:59
Sorry I'm late chaps, been busy and not had a lot of time

Currently still using the modified op-amp based output stage but got another one of these cheapo dacs so will try the Lampizator output stage on this one to compare

Theres only so much you can do with these dacs but it makes sense to get as much as you can out of them.
I'm still playing about with the decoupling, again the dac is limited because of the layout, I'll post more info as soon as I can

alb
19-05-2008, 07:24
This is how my version of this DAC looks when hooked up to a Lampizator style tube output stage.

http://img359.imageshack.us/img359/7970/lampy005fm0.jpg

Another pic showing updates to the DAC mainboard. Getting rid of a few surface mount parts and replacing with quality capacitors. Seems to have paid off. Still work in progress though.

http://imgcash3.imageshack.us/Himg359/scaled.php?server=359&filename=dacupdate003nr0.jpg&xsize=640&ysize=480

Filterlab
19-05-2008, 07:46
Very nice indeed.

iihay
19-05-2008, 22:40
Sorry I'm late chaps, been busy and not had a lot of time

Currently still using the modified op-amp based output stage but got another one of these cheapo dacs so will try the Lampizator output stage on this one to compare

Theres only so much you can do with these dacs but it makes sense to get as much as you can out of them.
I'm still playing about with the decoupling, again the dac is limited because of the layout, I'll post more info as soon as I can

Cool, I'm not ready to tinker yet anyway got a humm on my amp to resolve first, causing me some significant grief:scratch:

Interested to hear more.

Cheers

Iain

leo
20-05-2008, 00:01
A few very simple things to try first if you intend on using the standard op-amp based output stage, don't expect giant killing performance but its not too shabby IMHO for a dirt cheap thing:)

Recommending a specific capacitor type is hard for the decoupling, it mainly depends on what regulated PSU your going to be using, just be careful if using lots of low ESR caps and check for instability

For any op-amp you want to try carefully check the datasheet first to make sure its suitable

http://i273.photobucket.com/albums/jj220/kingbusoms/modded2-1.jpg
http://i273.photobucket.com/albums/jj220/kingbusoms/modded1.jpg

jonners
20-05-2008, 09:50
This is how my version of this DAC looks when hooked up to a Lampizator style tube output stage.

http://img359.imageshack.us/img359/7970/lampy005fm0.jpg

Another pic showing updates to the DAC mainboard. Getting rid of a few surface mount parts and replacing with quality capacitors. Seems to have paid off. Still work in progress though.

http://imgcash3.imageshack.us/Himg359/scaled.php?server=359&filename=dacupdate003nr0.jpg&xsize=640&ysize=480

alb - you seem to have soldered the replacement through-hole components very close to the board, and I'm wondering how you managed to do that so well?
The boards supporting the valves look very cool - can you tell me where I can get some like those please?
Thanks, John

lurcher
20-05-2008, 11:31
The boards supporting the valves look very cool - can you tell me where I can get some like those please?

http://www.hificollective.co.uk/components/audionote_connectors.html

And scroll down

alb
20-05-2008, 13:12
Hi John.


you seem to have soldered the replacement through-hole components very close to the board, and I'm wondering how you managed to do that so well?

The photo angle makes them look closer than they actually are. However, i did make an effort to keep the leads short as possible. My solder station has an iron with a long and quite thin pointy tip, that will get underneath em.

The Glasshouse tube mounts are useful, as they bring all connections to the top. Makes access much easier when swapping bits around, which one inevitably does.:)

jonners
20-05-2008, 13:43
Thanks, alb and lurcher. Now I just need my CS4397 board - coming on a very slow boat from China, I think. :confused:

nickthevoice
21-05-2008, 20:26
Do you had same value capacitor for all ?

nichicon Muse FG 1000uF is a great value for C19 or more bigger ?

you bypass SMD by MKP capacitor with same value ?
because BYPASS create 2 time the original value (2uf // 2uf = 4uf)


Thanks in advance.

nicK

alb
21-05-2008, 21:21
Nick.

I just used caps and values that i already had in my parts box.
C19 is Pana FC 1000uf, C21 is Pana FC 150uf.
The gold Samwha caps are 220uf and the Oscons 150uf.
The cream grey box caps are 0.22 polyester and are probably not ideal.

I have already changed C9 (input cap) from .22 polyester to a .47uf Obbligatto which is much much better, although way too big to fit on the board.

Please don't think that this random assortment is my best combination. To find that, could take some considerable experimentation.
However, what i have ended up with is still a significant improvement on the unmodified board.
Lots more work to be done yet, and always open to suggestions.

jonners
29-05-2008, 11:08
My DAC has finally arrived and I have it up and running (optical input). However, it operates on 44kHz and 48kHz sample rates but not on 96kHz. I'm a bit out of my depth here, but from my reading of the CS4397 data sheet it looks as though for 96kHz operation I need to connect M4 (pin 2) to +3.3V. (It's currently connected to Ground). Can anyone - Leo?? - confirm if I've got that right?
If I am right, changing it looks unfortunately like it would be rather a delicate operation. The board was advertised as a 24bit/192kHz DAC so it seems a bit amiss for it to be currently incapable of selecting the higher sampling rates.

Filterlab
29-05-2008, 11:13
Optical (TOSlink) will only carry a maximum of 48Khz. For 96Khz and above you'll need to use an electrical connection. However the upsampling will still go to 192Khz as long as the input is 24bit (48Khz, 96Khz, 192Khz or multiples/divisions thereof).

jonners
29-05-2008, 13:21
Optical (TOSlink) will only carry a maximum of 48Khz.

Thanks - that bit of information has somehow passed me by. :scratch: Will organise elec. input asap.

jonners
29-05-2008, 16:50
I now have electrical input, but still nothing doing at 96kHz - just lots of hiss. :confused:

Filterlab
29-05-2008, 17:37
Is your input definitely 24bit?

jonners
29-05-2008, 17:56
Yes. I can select 16, 20 or 24bit output on my Behringer SRC2496 upsampler. 44 and 48kHz work OK at all bit rates, 96kHZ at none.

Filterlab
29-05-2008, 18:38
How odd. Without being there I'm at a bit of a loss to be honest. Maybe Leo can help more.

jonners
31-05-2008, 16:43
I wonder if the problem could be as a result of the PPL filter having the wrong values - as noted by Leo when he started this thread? I haven't changed anything on my board yet.

jonners
01-06-2008, 18:32
New PLL filter values are in (not SMD though) - they improve the sound but there's still nothing doing at 96kHz.

iihay
04-06-2008, 20:53
Just got a load of caps in the post today Oscons and wima as per the lampucera site, will be having a crack at my board this weekend. What's the best approach with smd caps, am I ok just sticking the new cap across it or is it better to get the little blighters off first?

Cheers

Iain

alb
04-06-2008, 22:24
Iain.

Depends how confident you are with a soldering iron. The board has survived a couple of attacks from me, but there's always a danger of lifting PCB tracks.

The ceramic surface mounts are reckoned to be good with digital so i'd just bypass em for the time being.

The input cap C9(another ceramic which doesn't sound bad) can be improved on. I've been doing an experiment with different bypass caps for C9. Haven't really finished yet,(been at it for a week or so) but have discovered that some caps which can sound wonderful in analogue applications, don't seem to cope with high frequency digital information.
Exceptions are Obbligatto caps and Vishay/ERO 1837 which live up to expectation. Both of these improve on the original. Even a cheap tantalum sounded quite nicely balanced.
I tried a red Wima here but it sounded muffled especially at the bottom end.
Anything polyester sounded awful.
The clear winner amongst those tested so far, would have been a 0.33uf Ansar polyprop which has stunning clarity at the top end, but in this application, very weak bass.
Of course you may find that different values yield better results, but i have a feeling that very low inductance is important here and that's why the big Obbligatto cap works better than other big caps.

I even tried bypassing the bypass cap in an effort to get a good compromise, but then things start to deteriate. To be expected.

Next step is to remove the original C9 and see which one works best in its place.

Good luck.:)

jonners
05-06-2008, 10:00
What's the best approach with smd caps, am I ok just sticking the new cap across it or is it better to get the little blighters off first?

Cheers

Iain

I found it quite easy to remove the smd caps by heating them with the iron and then just sliding them off. You can then get a good, reliable connection between the new component and the pad.

iihay
05-06-2008, 10:18
Thanks chaps, I'm not too bad with an iron after my recent amp build so I should be ok. Keep us posted on the C9 experiments, I don't have too many caps lying around to do this kind of trial and error so I am very interested to hear how it goes.

Cheers

Iain

alb
05-06-2008, 11:45
A good reason to leave C9 in place for the time being, is that you then have something to use as a reference. As i found out after trying several different caps, you need to be able to compare with the surface mount.

It's quite difficult to remember the sound of a cap for any length of time.
Having narrowed it down to four, things should be easier.

If a particular capacitor sounds good as a bypass, then theres a reasonable chance it will be good in place of C9.

jonners
05-06-2008, 20:15
It seems to me it would make sense to look at C24 in conjunction with C9, since both are part of the SPDIF input circuitry.

alb
05-06-2008, 21:40
Good point, C24 is on my list of things to do. It hasn't been touched yet because i didn't want to fill the board up with caps and find out i couldn't get access to anything.

C9 is a real pain to get at, when trying to fit a chunky cap as close as possible to the board. I might remove the onboard input socket since i'm not using it. That should help quite a bit.

Having slagged off polyesters in a previous post, i now find that the little "orange drop" types work rather well. Don't usually like these, but they're not bad here.

jonners
07-06-2008, 09:47
Has no-one else tried to use this at 96kHz? Mine still won't play ball and I'd like to know why. By the end of next week I expect to see your clear answers, in good English and with no spelling mistakes. (Stern headmaster smilie would be here if we had one!).

alb
08-06-2008, 08:18
I'm sure Lee will come in when he can, Seems to be having a spot of bother with his PC at the mo.
I can't answer your questions.

leo
08-06-2008, 15:59
Don't talk to me about PC's :doh:

TBH I've not tried 96khz, reason being I don't have any 96khz material , I mainly use a modified SB3 to feed the dacs and its limited what it can output.
I'll have to look at the dacs datasheet, apart from this cheapo pcb I've not used the CS439* family before, could be to do with how they set the mode control on this board, the PLL was wrong so who knows what other mistakes there is;)


Have you tried another dac to be sure the digital signal is actually 96khz?

leo
12-06-2008, 16:53
Just knocked up this output stage to try
http://i273.photobucket.com/albums/jj220/kingbusoms/cs4397cheapo.jpg

iihay
12-06-2008, 18:45
Cool, be interested to know how it sounds. I've cleared my board of all smd caps and I'm in the process of sticking on the new caps. Half of the OScons I've been waiting for have got lost in the post though so not close to listening yet. Those little caps came off far easier than I expected.

I'll ultimately be listening to this using a CD transport from a JVC box (shigaclone from diyaudio) against my modded NOS arcam alpha 5. Really interested to see how such a cheap project fares against it.

Iain

jonners
12-06-2008, 21:09
Have you tried another dac to be sure the digital signal is actually 96khz?

Yes, I've had two other dacs working OK at 96kHz off that digital signal.

leo
12-06-2008, 23:35
The Arcam A5 is a unit I know well, these are nice and ripe for modding and sound pretty damn good too once properly fettled!

I'm giving the discrete stage a good run in, initial impressions are promising ;)

leo
12-06-2008, 23:38
Yes, I've had two other dacs working OK at 96kHz off that digital signal.

You could be right regarding pin2 of the CS4397, I've no idea why they grounded it:confused:
3,4 and 5 go straight to 3.3v

jonners
13-06-2008, 18:14
The mode table on the Crystal CDB4396/7 evaluation board sheet gives 'Automatic Mode Operation Settings' : M0=1 M1=0 M2=1 M3=1 M4=0 - which I think are the settings we have on our board. However, the Cirrus Logic CS4397 data sheet doesn't seem to mention 'automatic mode' at all. According to the tables there, these same settings would give I2S operation at up to 24-bit, 'single speed' (16 to 50kHz). I'm confused. :confused:

Puffin
13-06-2008, 18:38
iihay. I am using mine with a JVC Shigaclone. Sounds pretty good. What do you think ?

alb
13-06-2008, 19:08
Fitting a pulse transformer on the DAC input cleans up some noise that we don't always realise is there, and makes a significant difference.

Also i found that fitting a BNC socket to my player instead of using the PCB mounted RCA socket made quite an improvement. Same applies to the DAC board.

jonners
13-06-2008, 19:28
Fitting a pulse transformer on the DAC input cleans up some noise that we don't always realise is there, and makes a significant difference.


What pulse transformer are you using, alb?

iihay
13-06-2008, 20:14
iihay. I am using mine with a JVC Shigaclone. Sounds pretty good. What do you think ?

Glad to hear it, you're the same puffin from the diya thread then. I've not listened yet still waiting for a some missing caps to arrive, getting very impatient!


Leo - agreed lovely CD player, I didn't mod it. Chap from called Kevin green audiocellar in oxford did this one and it sounds absolutely wonderful. It will be a good reference and very hard to beat.

Iain

alb
13-06-2008, 20:15
What pulse transformer are you using, alb?

The one i have is a Newava S22083, available from Digi-key. I was lucky to have this sent to me by another member here.
Farnell do some Murata versions with similar specs, but i dont know how they compare.

Puffin
14-06-2008, 06:25
The guy who calls his version the Lampucera, says that the chip supports balanced operation. It seems that you can get a balanced out by connecting 4 wires plus grd to the vacant pads left by removing c27, c28, c30 and c31 ?

http://www.lampizator.eu/LAMPIZATOR/LAMPUCERA/CD%20DAC%20Lampucera%20lampizator.html

lurcher
14-06-2008, 08:55
Glad to hear it, you're the same puffin from the diya thread then. I've not listened yet still waiting for a some missing caps to arrive, getting very impatient!


Leo - agreed lovely CD player, I didn't mod it. Chap from called Kevin green audiocellar in oxford did this one and it sounds absolutely wonderful. It will be a good reference and very hard to beat.

Iain

Ah, another who has tried the JVC, I built one up a month or so ago, and it was remarkably good, it was more of a shock when it went to its owner and I had to go back to my old tosh DVD as a transport. Mo on here was at mine at the time and he heard what a difference it made.

I have another still in its box to try when I find time, but I might see if I can connect it directly to my DAC I am working on, as the chip on the transport does have a digital (not the spdif one) output, its just not i2s, but I should be able to manipulate that into i2s in the cpld I am using.

jonners
14-06-2008, 09:36
It seems that you can get a balanced out by connecting 4 wires plus grd to the vacant pads left by removing c27, c28, c30 and c31 ?


That's right. If you remove the caps you will see the pads are marked R+ R- L+ and L-.

Puffin
14-06-2008, 10:29
Goody Goody, more time spent bodging !. I bought a kit from White Noise years ago to accept a balanced in and balanced out. I didn't then and haven't since had a pre with balanced inputs, but I found a significant improvement by using the balanced outs on a Dacmagic and into the module and out as single ended. Far superior to the single ended outs. So, will have to give it a go.

iihay
14-06-2008, 14:15
I have another still in its box to try when I find time,...

I'm planning on buying a couple more for tinkering. They are hard to find but ASDA near me has loads left :)

Iain

leo
14-06-2008, 16:01
The guy who calls his version the Lampucera, says that the chip supports balanced operation. It seems that you can get a balanced out by connecting 4 wires plus grd to the vacant pads left by removing c27, c28, c30 and c31 ?

http://www.lampizator.eu/LAMPIZATOR/LAMPUCERA/CD%20DAC%20Lampucera%20lampizator.html

Yes, the dac chip already has differential outputs so running balanced is easy enough

jonners
17-06-2008, 14:18
I have finally got my board working at 96kHz, thanks to some help from error401 over at diyAudio. In case anyone is interested, the mod. involves lifting pin 2 of the CS4397 (to remove its ground connection) and wiring it direct to pin 16 of the CS8416. The board will then process all sampling rates up to and including 96kHz (and maybe 192kHz, though I'm not able to test that).
Doing this involved micro-surgery more delicate than I generally care to undertake :eek: so don't do this at home unless you have to, or you are comfortable with such things. It does sound really good at 96kHz, though!

alb
17-06-2008, 16:14
That's a useful discovery, nice one.

I've come to the end of my search for the ideal input capacitor. Vishay/ERO MKP1837 0.1uf does most things very well indeed, and as a bonus it actually fits on the board. Quite happy with it all now.

Puffin
19-06-2008, 16:28
I've now got mine working in balanced mode. IMHO much better than single ended.

jonners. Is that all there is to it (96khz) direct connection from pin to pin ?and you get 44.1 and 96khz ?

jonners
20-06-2008, 11:35
I've now got mine working in balanced mode. IMHO much better than single ended.

jonners. Is that all there is to it (96khz) direct connection from pin to pin ?and you get 44.1 and 96khz ?

Yes it will then receive and convert a 96kHz signal but it won't upsample of course. You have to lift pin 2 of the CS4397 away from its ground connection on the board. Its no use trying to cut the trace leading from it - possibly the pad connects to the ground plane in some inaccesible place under the board. You then solder a thin wire from that pin to pin16 of CS8416 - no need to lift pin 16.

I'm using balanced out too - direct into a TVC. I found that with no filtering after the dac outputs the hiss level rises somewhat, so I'm experimenting with a bit of passive filtering. At present I've got the + and - outputs going through series resistors of 240 ohms followed by 2.2nF bridging between + and -. That cuts the noise out, but I might experiment with slightly smaller caps.

Puffin
20-06-2008, 12:59
jonners. Thanks for your reply. Some years ago Elektor had an article for an Up-Sampler. I got the board and sourced all the parts, but as I was fairly new to bodging, I messed up a couple of SMD's and I think I lifted the pads. I wll have to dig it out and see if I can resurrect it.

I am using 22ohm resistors and 15uf caps. Maybe the caps are a little too high in value. I tried some 2.2uf, but the 15's seem to suit it better.

Good fun this lark in it.

Puffin
22-06-2008, 07:56
In 2001 I built the up-sampler mentioned above. The main chip and supporting caps/resistors are SMD's. Knowing very little then I built the kit and hoped it worked. (standard of soldering is incredible looking at it now as I didn't really know what I was doing) It didn't work or at least I didn't think it did until last night!. I had kept the Elektor mag and article and decided to re-read it thouroughly. With the greater knowledge I have now, things started to make more sense. There are 3 jumpers one for Coax/Opt another for Master/Slave and a third. They only sent me two jumpers with the kit and so I ignored the third. Reading the article with my better understanding told me that the third set of pins could be used to add an additinal crytal oscillator and if you were not going to do this you should............fit a jumper on it!

So, that's what I did and after 7 years, it works !

Next step to lift pin 2 and and enjoy 96khz CD's !

Puffin
22-06-2008, 19:42
Well, looks like it's my go again. I like this game I get to win all the time !

Seriously now. I have done the surgery to get 96khz. Up-sampler is working and will give it a few days to gel and will report back.

If anyone wants to see the Elektor article, let me know.

jonners
22-06-2008, 20:51
Well, looks like it's my go again. I like this game I get to win all the time !


That's right, and talking to yourself ensures that you get a sympathetic listener! :lol:

Will be interested to hear what you think of it at 96kHz. Slighly smoother and more spacious in my system, but not a night and day difference.

Mr. C
23-06-2008, 07:16
John if your up-sampler can output 88.2khz give that a try, you may find it gives you back some of the music. Nice works by the way guys :-)

STIFA
23-06-2008, 08:07
@alb @all
About C9 and best cap...I can't seem to be able to find VIshay anywhere near me and as for ERO i found following:
0.22uF/100v MKC
0.33uf/250v MKC
0.47uF/100v MKC
1.0uf/63v MKC
1.0uF/100v MKT
0.30uF/250v MKP
0.33uF/400v MKP
0.33uF/250v MKP 1840 5%

so could I use some of the above? Also what is the voltage for C9 and other cap bypasses and their uF values?

Also what is the voltage of OS-CONs , I plan to replace all stock with 150uF Sanyo os-con, but is there a minefield in choosing voltage? Options are 6.3v 10v and 16v.

Thanks

jonners
23-06-2008, 08:28
John if your up-sampler can output 88.2khz give that a try, you may find it gives you back some of the music. Nice works by the way guys :-)
Yes I've been trying that - don't know which I prefer just yet though.

Puffin
23-06-2008, 19:13
Used it for a couple of hours tonight and I like the combination a lot. Can't quite put my finger on exactly what the up-sampling does, I suppose it's quite subtle, but at the same time it seems to add depth and more realism.

STIFA
27-06-2008, 18:31
on Lampizator.eu I fount this text :
DO NOT TOUCH THE circuit consisting of C22, C23 and R2 !!! This is a resonator.

Problem is that the leads on capacitors act like antenna I think.

Should they be replaced with high quality SMD parts then instead?

What are your impressions?

jonners
27-06-2008, 19:08
There is some info. about this here: http://rockgrotto.proboards39.com/index.cgi?board=m&action=display&thread=1672&page=5

leo
27-06-2008, 22:00
on Lampizator.eu I fount this text :
DO NOT TOUCH THE circuit consisting of C22, C23 and R2 !!! This is a resonator.

Problem is that the leads on capacitors act like antenna I think.

Should they be replaced with high quality SMD parts then instead?

What are your impressions?

Be careful what type of caps are used here, I've found some in the PLL can make it sound surprisingly nasty
I'd certainly not use those ceramic disc plates which are microphonic
Its hard finding nice parts in SMD, COG/NPO should be ok for 1N , the 22N maybe harder

leo
27-06-2008, 22:02
The resistors and filter caps in the standard differential to unbalanced don't make any sense to me in this circuit:confused: the + is supposed to mirror the -
They don't so it gives large DC offset:doh:

More weirdness with this little dac board, no wonder it was cheap:lolsign:

iihay
01-07-2008, 18:42
Finally getting around to working on this again as some more caps arrived. I am a little confused about the 4 smd caps L+- R+- next to the opamp. Should these be replaced or linked out>

Cheers

Iain

jonners
02-07-2008, 22:08
Does post no. 24 from Leo help with this?

STIFA
03-07-2008, 07:53
Be careful what type of caps are used here, I've found some in the PLL can make it sound surprisingly nasty
I'd certainly not use those ceramic disc plates which are microphonic
Its hard finding nice parts in SMD, COG/NPO should be ok for 1N , the 22N maybe harder

What did you use?
What about that 3K resistor? What 4 that ?

Has changing original SMD parts on the PLL filter with new ones given improvement and how big?

Btw , maker of AYA dac will be soldering these parts as a favor, I just need to know what to put where. See you made one, gerat choice.
Thanks

STIFA
03-07-2008, 09:47
LEO
I really cant find SMD for resonator PLL filter, can I use normal ones and which?

Fikus stated that they act like an antenna? True?

Thanks

leo
03-07-2008, 19:54
What did you use?
What about that 3K resistor? What 4 that ?

Has changing original SMD parts on the PLL filter with new ones given improvement and how big?

Btw , maker of AYA dac will be soldering these parts as a favor, I just need to know what to put where. See you made one, gerat choice.
Thanks

For the resistor I use minimelf made by BC components
These are a tight fit because they are equivalent to 1206 size but they will fit if you have a steady hand http://uk.farnell.com/jsp/level5/module.jsp?moduleId=en/207337.xml

you can use 5mm sized through hole caps if you're struggling to find any decent SMD's, the key is to trim the legs as far as you can to the body.
Avoid types that are inductive, a thin film would be better than a stacked film

The sound improvement is subtle but worth it because its an easy mod to do.

The best advice I can give for this dac if you use co-ax in for the digital, DITCH the supplied phono socket and replace with a 75R BNC, the standard phono is real poor and causes problems

Maker of AYA dac, I presume you mean Pedja? if so please tell him Leo sends his regards;)
The AYA is a lovely dac!

iihay
03-07-2008, 20:02
Does post no. 24 from Leo help with this?

Doh :doh: Knew the answer was in here somewhere, don't know how I missed that page when going back through!

Ta

Iain

leo
03-07-2008, 20:18
If your using the standard output stage and intend on bypassing the four caps after the dac you MUST use some caps on the output (left and right out on the pcb socket) because of the DC offset, I'd bypass the pair of crappy standard tiny things and use better quality fitted to the output cable feeding the phono sockets

if you fancy a bit of smd resistor soldering there is some alterations you can do to eliminate most of the DC offset

iihay
03-07-2008, 23:32
If your using the standard output stage and intend on bypassing the four caps after the dac you MUST use some caps on the output (left and right out on the pcb socket) because of the DC offset, I'd bypass the pair of crappy standard tiny things and use better quality fitted to the output cable feeding the phono sockets

if you fancy a bit of smd resistor soldering there is some alterations you can do to eliminate most of the DC offset

Ok, any thoughts on values for this. I have various ones spare but very few decent "audio" caps apart from some 100uF Oscons and some 10uf Wima's. Alternatively I could be up for having a crack at the resistors. Taking the smd caps off was easy enough. Should I measure the DC on it?

Apreciate the help, learning as I go :)

Iain

leo
04-07-2008, 01:10
10uf Wima should be fine, I personally wouldn't use an oscon for signal but feel free to try it:)

The standard output stage +/- should be mirrored, it isn't for some reason and this gives the DC offset.
If you look at resistors R8 right channel and R9 left channel for the inverted side they are 36k, the mirrored side to these for the non inverted R14 right channel and R17 left channel are 22k1.
Basically R8 and R9 should be 22k1 to match 22k1 for R14 and R17.
If you change those resistors to match, and then measure the output pins of the op-amp (pins1 and 7 to ground) the DC offset should be a couple of mV's instead of the hundreds of mV's in the standard circuit.
You may lose a bit of gain but it shouldn't be a problem.
BTW a thanks to Ray for pointing this out

theres also a cap across each of those resistors, they don't match either:confused: only thing here is that on the non inverted side they are 1N and the inverted side they are 56p, I'm not sure if having 1NF across the NFB resistor would sound worse than 56p, its something I'd need to play about with, these caps shouldn't effect the DC offset though, its mainly the resistors mentioned above.

This dac is weird:lol:

I'd still use a input coupling cap in the amp or pre just as a precaution

iihay
04-07-2008, 12:37
Thanks Leo for the excellent explanation. I'll check this out at the weekend and replace the resistors. 10uF WIMAs are no problem I have loads.

I'll probably stick an input coupling cap in the amp and see how it goes. Although I guess this depends on if the DC is variable at all as DC in the amp I'm using (SKA GB150's ) can be nulled so a few mvs should be ok. Got my newly built speakers at the end of this chain though :eyebrows:

leo
04-07-2008, 17:09
No probs:)
The only thing is if those ceramic caps are left with the odd value it may sound a bit odd, if measured probably makes the signal look all over the place:lolsign:

Once the dac is sorted and your confident the DC offset is low enough you should be able to remove the coupling cap, not sure I'd risk it with this thing though tbh:smoking:

iihay
04-07-2008, 19:45
No probs:)
The only thing is if those ceramic caps are left with the odd value it may sound a bit odd, if measured probably makes the signal look all over the place:lolsign:

Once the dac is sorted and your confident the DC offset is low enough you should be able to remove the coupling cap, not sure I'd risk it with this thing though tbh:smoking:

lol, this dac is just for and cheap and fun way to get a dac for my transport. my main system will still be the modded arcam so I think I can risk it a bit Ultimately (i.e when not skint) I fancy something far better, twisted pear buffalo maybe.

Iain

leo
04-07-2008, 23:34
I have my eye on that ESS sabre dac too;)

leo
20-07-2008, 14:54
Had a couple of PM's from folks wanting to try the simple mods
I've added the below, hope its easier to understand:)

For the coupling cap mod simply bypass
C27,C28,C30,C31,C6,C54

Doing the simple coupling cap mod gives a 800mV DC offset on the output so you need to add a decent quality coupling cap on the outputs feeding the phono sockets to protect the vol pot in pre-amps etc, even if your amp uses input coupling caps before the circuitry this offset from the dac can still damage the plastic track vol pots and certainly won't do speakers much good if theres no input coupling caps at all

C1 and C2 can be around 1uf-10uf, its best to try different types.
R1 and R2 can be 22R-100R, this is just for a bit of loading for the op-amp, some op-amps like AD826 can drive heavy loads so lower values are fine, about 68R-75R should be fine in all cases

http://i273.photobucket.com/albums/jj220/kingbusoms/simplemods1-1.jpg

The pic below should be easy to understand, it simply bypasses the chip on the bottom of the board thats on the analogue output, this does similar job to muting transistors, the sound is better without it IMHO, just add blobs of solder onto the little pads next to the chip

http://i273.photobucket.com/albums/jj220/kingbusoms/DSCF0840.jpg

If you want to improve the measurement of the output stage and also reduce the DC offset to very low mV's change the following parts

R8,R9,R14,R17 to 30k
C32,C33,C34,C35 to 150pf

Measure the output L to ground and R to ground at terminal marked J5 on the board, it should now read roughly 1mV
You should be able to get rid of external C1,C2 (added above) without any damage from DC offset, I'd personally leave in C1,2 as a precaution though;)

leo
20-07-2008, 15:21
User gabdx PM'd me, I got the following after trying to reply

gabdx has chosen not to receive private messages or may not be allowed to receive private messages. Therefore you may not send your message to him/her.

iihay
20-07-2008, 15:24
Thanks leo, nice explanation, reducing down the DC is top on my list. I have a small problem with mine in that I have managed to lift the pads off on the chip underneath on the left channel. Is there any other way to do do this mod or should I remove the blob from the right channel to switch back in the chip on both sides?

Also I got carried away with cap replacement and pulled out the 2 25V ones on the rhs of the socket. On checking I don't have any decent 25V caps and I binned the ones I took off :doh: The best I can manage at present is 16V, I assume these will not do?

Cheers

Iain

leo
20-07-2008, 15:36
Hi Iain,

Hmm, let me think, its possible to still do the bypass, how hard would it be for you to desolder that SMD chip and remove?

16v caps is fine, highest DC for this dac is 12v for the op-amp decoupling:)

Cheers
Leo

iihay
20-07-2008, 15:40
Hi Iain,

Hmm, let me think, its possible to still do the bypass, how hard would it be for you to desolder that SMD chip and remove?

16v caps is fine, highest DC for this dac is 12v for the op-amp decoupling:)

Cheers
Leo

Should be ok removing that if I pick up some more braid tomorrow, depends if there is much danger of removing more pads which are then needed! Will I need to link some when it's off?

Actually would cutting the legs be an option here or does the thing need to go back on after?

Cheers

Iain

leo
20-07-2008, 15:44
Hi Iain,

Hold on a mo, it is the pads nearest where its mark U1 on the PCB that has lifted? for the left channel?

If so you can run small wire from the SMD pins on that chip which will do the same job, give me ten minutes and I'll edit the pic to show you which pins to link on that chip for the left channel.

What you then need to do is set your multi meter to continuity and measure from the output socket Left to the linked C6 to be sure it measures continuity.
Does that make sense?

leo
20-07-2008, 15:59
Should be ok removing that if I pick up some more braid tomorrow, depends if there is much danger of removing more pads which are then needed! Will I need to link some when it's off?

Actually would cutting the legs be an option here or does the thing need to go back on after?

Cheers

Iain

Got a pic coming, hold on;) should make it easier for you.
Just as long as its the left channel

leo
20-07-2008, 16:01
This is incase the pads for the left channel are lifted
http://i273.photobucket.com/albums/jj220/kingbusoms/untitled-2.jpg

iihay
20-07-2008, 16:10
Splendid that's easy, I'll give that a go. On inspection it looks like my blob on the R channel has linked to a pin as well which I'll have to sort out, bit of a large blob!

Appreciate the help.

Iain

leo
20-07-2008, 16:13
No probs, of course you can easily just cut this chip out and link the appropriate pads, its not needed anyway once the linkings been done

leo
20-07-2008, 16:15
Also if your blob of solder for the right channel touches pin3 or 4 then don't worry about it, the pads go to those pins anyway

iihay
20-07-2008, 16:35
Also if your blob of solder for the right channel touches pin3 or 4 then don't worry about it, the pads go to those pins anyway

Left it as it is then :) Not having a good day soldering wise, I just lifted the a pad from C6 as well. Fortunately it was still connected so I have linked it out and hope all is ok. The soldering iron is turned off now as I obviously have lost the knack today!

Ordered up some others bits from RS so hopefully will have the DC mod completely by tomorrow.

Thanks for the help

Iain

iihay
20-07-2008, 16:42
BTW, are those cap values for the DC mod crtitical, the ones around the chip, or can they be a little higher? I know I have nothing quite that low but I may have a few spare caps in that sort of range.

Iain

leo
20-07-2008, 17:05
150pf is the values that gives best measurements for that particular filter with the 30k's etc, I guess you can try 100pf, I'm not sure how it would sound, may not make a huge difference.
main thing is getting those 4 resistors the same where as in the standard circuit they are way out and don't make sense, I've no idea why they even fitted such values:lol:

Oh 180pf should be ok too

gabdx
21-07-2008, 02:57
Hi leo , thanks for all the explanations :) sorry for not being able to receive messages, i hope i will soon.

gabdx
21-07-2008, 12:21
Hi, it looks like i can't post or even pm you now, no idea why i could not receive pm from you leo... the receive message from other box is checked in my option..

Marco
21-07-2008, 12:42
Hi gabdx,

All has been explained in an email :)

Please, as per our policy and request to you, pop into the welcome room and briefly intoriduce yourself and your system to our members.

What is your first name?

Thanks in advance.

Marco.

Marco
21-07-2008, 12:55
Hi Gabriel,

Thank you for complying with our request. Welcome, and please continue to enjoy the forum :)

Marco.

leo
21-07-2008, 17:11
Must admit Marco I do much prefer to call people by their proper name, it seems a bit odd saying hi to some of the strange usernames;)

Marco
21-07-2008, 17:17
Yep, that's the whole point of why we ask people to sign posts with their real names. Therefore your username can be 'toxiccockhead405' as long as you sign it as 'Frank', or whatever. You can even make up a Christian name if you like as long as it sounds 'real'! :eyebrows:

Marco.

gabdx
21-07-2008, 17:21
Right now my dac doesnt work at all :( voltages seems extremely low, i am going to recheck the power supply and evrything ,

Leo ! now i can post :) thanks Marco , now for the serious part , i have the bottlehead
preamp with a 33 K resistor in the input ,

I also have another integrated amp with a 100 K volume pot...

Both are unbalanced in, so can i plug (with capacitors of course) the unbalanced cirrus chip into one of these ???

EDIT : the cirrus output used to be 2.5 v DC , now its 0.7 DC and some 0.4 AC... weird , i checked the power supply and its all good 4.7 V and 13 V close enough. The RCA voltages are just around milivolts of DC
On ac the RCA are varying with music around 0.2 Volts AC if i only touch the + or - with one leg of the voltmeter /
If getting a ground point and touching the + or - on the RCA i get around 0.5 Volts

Bottom line i hear no sound !

leo
21-07-2008, 17:32
Yep, that's the whole point of why we ask people to sign posts with their real names. Therefore your username can be 'toxiccockhead405' as long as you sign it as 'Frank', or whatever. You can even make up a Christian name if you like as long as it sounds 'real'! :eyebrows:

Marco.

Yes, I sometimes use my middle name on forums if a member has the same name as my first, as you say though its got to be more convincing than saying Hi toxiccochead:lol:

iihay
21-07-2008, 17:36
Yep, that's the whole point of why we ask people to sign posts with their real names. Therefore your username can be 'toxiccockhead405' as long as you sign it as 'Frank', or whatever. You can even make up a Christian name if you like as long as it sounds 'real'! :eyebrows:

Marco.

Toxiccockhead405 is actually my real first name, I just use Iain for forums :lolsign:

leo
21-07-2008, 17:45
Right now my dac doesnt work at all :( voltages seems extremely low, i am going to recheck the power supply and evrything ,

Leo ! now i can post :) thanks Marco , now for the serious part , i have the bottlehead
preamp with a 33 K resistor in the input ,

I also have another integrated amp with a 100 K volume pot...

Both are unbalanced in, so can i plug (with capacitors of course) the unbalanced cirrus chip into one of these ???

EDIT : the cirrus output used to be 2.5 v DC , now its 0.7 DC and some 0.4 AC... weird , i checked the power supply and its all good 4.7 V and 13 V close enough.

Hi Gabriel

Bloomin heck, how did you manage that l:doh:

If your 5v supplies are a lot lower than normal you probably have something dragging it down, could be anything from a short to bad instability somewhere .
You could try posting a close up pic of your dac and we can look for possible faults.

I presume you mean can you plug the output of the dac chip directly into the pre-amp through coupling caps?
If so, yes you can but I wouldn't recommend doing this with a solid state pre amp if it only has single line in rather than balanced.
You won't get the benefit of the noise canceling without a differential to line out using this type of dac, you'll also have no output filtering, it may sound pretty nasty, you can of course try it, using something like the lampizator valve stage makes things a little more forgiving when using the dacs direct negative left and right out.
You can also try it with your bottlehead pre, I've no idea what it will sound like but just be careful if you give it a go;)

leo
21-07-2008, 17:50
You edited before I posted:ner:

I presume this 0.7v DC is what you are measuring at the dacs differential outputs to ground and not on the line out after the op-amp?

gabdx
21-07-2008, 17:50
Yes i can post a picture very soon,
regarding the voltages on the supply , i think i have partially fried the little regulators feeding it with incorrect wiring on my china kit transfo, these guy in china didnt gave me any instruction how to connect it ! Since i have replaced it with a wall pack 9 V that really is giving me over 11 volts. Ill try it with direct coupling on the preamp we will see.

leo
21-07-2008, 17:51
Toxiccockhead405 is actually my real first name, I just use Iain for forums :lolsign:

Can I call you cockhead for short then ? :eyebrows:

gabdx
21-07-2008, 17:52
You edited before I posted:ner:

I presume this 0.7v DC is what you are measuring at the dacs differential outputs to ground and not on the line out after the op-amp?

Exactly , and it used to be 2.5 V


http://gabdx.webs.com/DACG/Picture%20009.jpg
http://gabdx.webs.com/DACG/Picture%20011.jpg
http://gabdx.webs.com/DACG/Picture%20013.jpg
http://gabdx.webs.com/DACG/Picture%20014.jpg

leo
21-07-2008, 17:54
Yes i can post a picture very soon,
regarding the voltages on the supply , i think i have partially fried the little regulators feeding it with incorrect wiring on my china kit transfo, these guy in china didnt gave me any instruction how to connect it ! Since i have replaced it with a wall pack 9 V that really is giving me over 11 volts. Ill try it with direct coupling on the preamp we will see.

Yes, problem with a lot of this stuff is that its usually cheap for a reason, I won't comment on here my experience with certain products made in China:steam:

leo
21-07-2008, 18:05
Exactly , and it used to be 2.5 V

did you try removing all the links across C27,28,30 and 31 and measure again? this just eliminates any problems on the boards analogue section

Also check the voltage on the dacs onboard U4 SMD regulator, it should be 3.3v

Check for any shorts or track breaks around capacitor C25, this is for the dacs internal bias

Also check the CS4397 analogue supply pin (pin22) it should have 5v going to this pin

gabdx
21-07-2008, 18:52
i did mesure again without the links and it give low reading, mostly around 0.4 volts ac or dc.

the 22 pin indicate - 1 V and 21 +1 V, the U4 has 3 pins, 1 to 3 = -2 V
2- 3 = around 1 V.

Imo it looks bad

leo
21-07-2008, 19:01
i did mesure again without the links and it give low reading, mostly around 0.4 volts ac or dc.

the 22 pin indicate - 1 V and 21 +1 V, the U4 has 3 pins, 1 to 3 = -2 V
2- 3 = around 1 V.

Imo it looks bad


Hmm, thats not good mate:scratch:
The onboard 3.3v seems a little low but its that 1v for the dacs analogue which is the main problem.

Set your multimeter to continuity/bleep, turn the dacs power off and measure pin22 to ground, does it measure short circuit or low ohms?
If so you either have a short somewhere on the supply lines or if your very unlucky the dac chip may have gone

iihay
21-07-2008, 19:13
Can I call you cockhead for short then ? :eyebrows:

I'll wash and then you can drop the toxic ;)

Iain

--
Another totally off-topic post from the keyboard of Toxiccockhead405

leo
21-07-2008, 19:13
Another thing to check, the small connectors inside the big white plastic plug, those things are really shite in all honesty, I've known the plastic sleeving for the cables can sometimes not be stripped back enough so the metal crimp does not make a good contact with the cable core.
Because my dac uses a different regulated supply to the standard one I removed the plastic socket and fitted something else

leo
21-07-2008, 19:19
I'll wash and then you can drop the toxic ;)

Iain

--
Another totally off-topic post from the keyboard of Toxiccockhead405

Thats a little bit too much info there mate:lol:

Marco
21-07-2008, 19:25
Hahaha... :lolsign:

Slightly off-topic but it reminds me of a wee joke...

A chav Scottish couple called Bobby and Betty are trying to mix it with the posh set are at a dinner party. The champagne's flowing and everything's going well. The hostess spotting that they're without nibbles passes them a tray of canapés and asks Betty in a snooty voice: "Cheesy dip?" to which Betty replies: "Naw, hen, he washed his cock this morning..."

:eyebrows:

Right, back to DACs...

Marco.

leo
21-07-2008, 20:36
Thats put me off Dairylea dunkers forever!

Nowt wrong with these odd off topic posts, it adds a bit of flavour to the discussion even if it does involve a bit of cheese :lol:

gabdx
21-07-2008, 22:01
Ground is ok, no shorts , only short to the 5 V , 20 M ohm to the Ground = evrything ok
I RESOLDERED every power connections :)
Good news the cirrus dac show a healthy** weird**1.7 V AC+DC , with some weird 0.4 V DCat output.... i think i am going to try it directly on my buffer ( bottle head foreplay ) then to the integrated tube amp...

...hope nothing blow up



It mesured around 0.1 V DC without any music playing at the output of the buffer same with music playing... ok...i hope... should be ok ... if you have a bottle head pre, never turn it down first it gives 20 V DC... you hear my speaker cry when i did that sometimes... they need mostly only 5 V to play at peak level!


result : just noise, ground problem, sound like a rca socket with defective ground, leo i anticipate your next post, change the kit sockets, lol ill try to tight them very hard and it should work :mental: , but i think even then it's not going to work . Anyone has a good dac to recommend me ?

iihay
21-07-2008, 22:14
....a bit of flavour ...

Not a great thought! Back on DACS I am interested on your thoughts as regard the PSU. How good/bad is the supply supply versus a decent diy job?

Iain

gabdx
26-07-2008, 16:07
My dac looks worst and worst , before i rip evrything apart ill try to revert all the modifications...

leo
26-07-2008, 23:01
Not a great thought! Back on DACS I am interested on your thoughts as regard the PSU. How good/bad is the supply supply versus a decent diy job?

Iain

I never had the standard psu so can't comment on its quality, it certainly helps to use the best you can though, something with low impedance low noise is what I use for both analogue and digital regulation

Theres plenty of cheap diy designs about, its easy to go mad but obviously seems to defeat the object of this cheap diy project.
Could use something like a cap multiplier supplying a shunt set for 5v for the dac.
Adding a current regulator diode from the op-amps output to the supply pin with worst PSRR can add a nice improvement on the cheap, if you can't find a suitable CRD its easy to make one using a Jfet with a resistor through it to set the current, also add a resistor before the CRD to help with the capacitance

leo
26-07-2008, 23:04
My dac looks worst and worst , before i rip evrything apart ill try to revert all the modifications...

I think its just been a bit of bad luck you've had, it makes sense removing the mods to see if the fault can be isolated

gabdx
27-07-2008, 01:47
I think it's completely out of usage , the cs itself seems dead, maybe i did a short some capacitor and it burned something inside... I tried evrything i could and nothing works.

leo
28-07-2008, 15:07
If the CS8416 receiver or the CS4397 dac chip has gone theres the option of replacing the chip if your confident
http://uk.farnell.com/1023396/semiconductors-integrated-circuits/product.us0?sku=cirrus-logic-cs4397-ksz&_requestid=484498

http://uk.farnell.com/1023451/semiconductors-integrated-circuits/product.us0?sku=cirrus-logic-cs8416-csz

Puffin
28-07-2008, 15:13
Got to be cheaper in real terms to buy another board.

gabdx
30-07-2008, 14:52
Thanks guys, especially leo :) and hi iihay i got me too a ska under construction :)))

Well i might buy a Rakk, or some modified lite 60 , i still don't know and my confidence with china products is going down and down ... though i think the dac is my fault it's a warning about it for moders !, had no problem modifying my dared amp tough ! no chips :)

I had 2 other bad experiences with chinese products : buying on ebay this guy a opamp like 85 $ and its a non working one, lost my money,

Now i bought some preamp transformer and the instructions are far from clear.. chinese is so tempting because it looks half price of usa gear but there is always a catch somewhere in the design or on reliability.

iihay
30-07-2008, 18:57
Thanks guys, especially leo :) and hi iihay i got me too a ska under construction :)))

You'll love the SKA, quality bit of kit :)

Puffin
01-08-2008, 09:35
Now i bought some preamp transformer and the instructions are far from clear.. chinese is so tempting because it looks half price of usa gear but there is always a catch somewhere in the design or on reliability.


I bought a Lite remote controlled Pre-Amp from DiyClub.biz. Cost aout £20 with shipping. 4 unbalanced and one balanced input, relay switching for inputs, Alps motorised pot.

Kit came with no instructions at all as to which pins to attach which sets of wires, and more inportantly how to wire up the senser and adjuster. I emailed and got some info back. Pidgin English! Good start.

Turns out the wire patch cables are insufficient. Some pin sets are 6 pins and the cables have only five wires and 5-hole connectors. If I didn't have copius amounts of "rubbish" that I have taken off over the years. I would not have been able to get it working.

The thing is actually very good for the money and the volume control works quickly. However if you were a newbee you would not have got it working.

leo
24-08-2008, 15:40
Some more fiddling, poor things had more alterations :lol:

http://pic20.picturetrail.com/VOL1293/4634025/12749957/331627487.jpg

chakija
25-08-2008, 09:35
Hi guys !
I also have this DAC and i have done those simple bypassing mods and i putted 2 oscons in C20 and C17.However ,i have not tested if it's working.
Offsets with original opa are 75mV for L and 180mV for R .
Are those values ok?
Will it be ok if i solder 75R coaxial cable direct on board instead of RCA socket?
I 2 have one of those JVC transports so i would solder coax on it 2...

leo
26-08-2008, 11:26
Hi guys !
I also have this DAC and i have done those simple bypassing mods and i putted 2 oscons in C20 and C17.However ,i have not tested if it's working.
Offsets with original opa are 75mV for L and 180mV for R .
Are those values ok?
Will it be ok if i solder 75R coaxial cable direct on board instead of RCA socket?
I 2 have one of those JVC transports so i would solder coax on it 2...

Theres a flaw in the original output stage, no matter what chip you use there will always be a DC offset, if you want to lower this you have to properly mirror the differential output, by changing a couple of caps and resistors it lowers the DC offset to +/- 1mV , it also improves the frequency response , the original circuit doesn't make a lot of sense TBH

for now the DC offset is too high so use a decent quality pair of coupling caps on the output to be safe, even though your amp may have an input cap its not recommended to have DC across the pot tracks for the pre, in a lot of cases you get the coupling cap after a pot so any DC tends to give a scratchy sound when adjusting volume, it can damage the tracks

chakija
26-08-2008, 13:31
Well ,i think i menage to kill my dac 2 .
With stock opa bypassed ,i connected to 2.2uF and 47R on 100K pot ,and then to my aleph J iteration amp (babbelfish J).I got some loads on speakers even with pot turned down.I didn't have courage to even touch the pot !
I instantly turned it off after 1 minute.
When i turned it on again and measure offsets ,i had some huge values like 1700mV ,and after few secs i didn't have any value at all.Finally ,i cut bypasses and measure exit's from dac and it seems dead.:wah:

Source is from my jvc transport connected directly on board with 75R coaxial.

chakija
28-08-2008, 20:11
I disconnected spidf and i tried to measure again.
I have now 0.25 V on R+ R- and L+ L- from dac.Does this means it is maybe still alive ??

paradox
30-08-2008, 14:19
I've been following this thread for a while and have finally decided to contribute something, albeit a request...!

Leo, do you have a link to a schematic of this amp? I'm just wondering how you know so much about how to effectively modify it, or have you just worked it all out? I'm more interested in the output stage, as I'm considering using it to feed my main amp.

thanks

leo
30-08-2008, 17:21
Lots of experience in fiddling with things:lol: I find it hard to resist not to meddle with any piece of kit I buy.
Everything is made to a given price, the tweeker does not have these limitations so if theres anything we don't like in a particular circuit we simply faff about until its improved.
Theres also a few design errors on this dac which is probably why they are being sold for cheap

There is a schematic for the dac somewhere (I'll see if I can find it) although a lot of the part numbers on the drawing is different to whats on the pcb so it just requires working out

paradox
31-08-2008, 14:44
I'm also looking for somewhere to draw power for some LEDs to indicate which input is being used. I'm using a rotary switch to select between coax and optical, and could do with an external LED indicator.

jonners
01-09-2008, 18:14
Leo, do you have a link to a schematic of this amp?

A schematic can be found here (scroll down to near bottom of page) : http://lampizator.eu/LAMPIZATOR/LAMPUCERA/CD%20DAC%20Lampucera%20lampizator.html

As Leo said, the part numbers are different from those on the board.

Tripmaster
09-09-2008, 07:17
I tested the dac before fitting parts and it worked fine.

The story so far...

Stage 1

Changed LM4562 OPAMP Sound: ok

---------

Stage 2

Removed and bypassed PS caps (C30, C31, C28, C27 & C54, C6) with thin wire. connected coupling caps and R68 resistors on RCA output, soldered two blobs to bypass chip on the base of board.

Sound: Treble is OK but bass really muffled and distorted - Sounds really bad.

----------

Stage 3

Change all other caps. Sound: Still sounds the same, but it still works!

I have not removed any of the smd caps, just bypassed with vishay

I have not changed any resistors.


Do you have any clues?

alb
09-09-2008, 15:13
Aaaaaarrrrgh! CAPACITOR OVERKILL!!!!!

I think you may have made things difficult for yourself, with the choice of caps. Not easy to see whats what.
Probably best to retrace your steps and find out where it went pear-shaped.
Even put it back to std board and use it for a few days, till its all nicely run in. Then you'll know better what to expect.

I don't use opamps in mine so difficult to comment.
It should when upgraded(and before), sound very clear and dynamic.
Good luck.

leo
09-09-2008, 18:47
A few problems I can see

WAY too many Oscons there I'm afraid, the ESR will be too low, you need to check for possible instability on a scope, Oscons also have a tendency to kill bass if you use too many of these which will give very soft signature
Those film caps are too big physically, their inductance will be too high

A mistake I've seen a lot is people tend to think low ESR capacitors are the best for all apps, infact it can cause a lot of problems with regulation regarding stability

If you have to use this many Oscons and after say a few days it still sounds bad try adding a low value resistor in series with those caps to add some impedance and help cure any possible ringing

leo
09-09-2008, 19:00
Also I noticed you used Sonicaps, these are reported to need some burning in , also the LM4562 tends to be poor at driving capacitive loads, I tried this chip but removed it, it needs care before it sounds half decent IMO

Tripmaster
09-09-2008, 20:38
Hi

I think I may have got little carried away! ;)


The bass is in fact very heavy. It sounds like strumming a loose bass guitar string

I started by just changing the opamp and it didn't sound any worse than the stock card.

I then proceeded to remove only C30, C31, C28, C27, C54, C6 and added two SCR coupling caps and 68R resistors. It was at this point the sound took a turn for the worse. I changed to the Sonicaps because I thought the SCRs may have dried up as they are quite old (18years)

In my parts bin I have the following items

10x 47uf 16v Tant beads
5x WIMA 150pf
10x Panasonic FC 470uf

Would I be better off replacing some of the components with these items, And if so at which points?

Thanks for your help

Richard

Tripmaster
09-09-2008, 20:53
The oscons are 150uf SA

Grey Vishay MKP 0.22uf

Blue Vishay MKP 0.1uf

The Vishays are bypassing the SMDs

alb
09-09-2008, 23:08
Richard.

Sorry for the poor photo but this will show you how mine looks with the upgraded caps.

http://i27.photobucket.com/albums/c164/alnewall/Baffles-1.jpg

C19 is 1000uf Pana FC.
C21 is 150uf Pana FC.
You can use your 470uf at C19. C21 should be smaller.

C9 is Vishay mkp1837 0.1uf. This was the best i found after trying twelve different caps here. Big ones don't work properly and generally mess up the sound, sometimes in a big way.
C9 is in my opinion one of the most important caps on the board. This also the only surface mount that i removed.

C16,17,20 and 25 are Oscon 150uf 16v. Probably enough Oscons.
C3,C18 are 220uf Samwha caps(gold), but any decent quality cap suitable for purpose would perhaps do.

C8,11,12,13 are 3.3uf Tants.

All lytics are bypassed by 0.1uf film caps.
I think blue Vishays are preferable to Wima, anywhere on this pcb.

This DAC works very well. Hope this is helpful.

Tripmaster
10-09-2008, 05:56
Whilst listening to the DAC last night one of the solder connections attaching the right channel lead to the coupling cap broke away.

I re-soldered the lead and at the same time decided to remove the two blobs of solder from the chip on the base of the board (These blobs bypass the chip). I fired it up it up again and only the left channel works. All of the connections seem fine and both channels operate via the DACMAGIC.

I re-soldered the two blobs to the chip and now the output is very quite and distorted.

This is not a good way to finish the evening!

HELP!!!!

Tripmaster
10-09-2008, 05:57
Richard.

Sorry for the poor photo but this will show you how mine looks with the upgraded caps.

http://i27.photobucket.com/albums/c164/alnewall/Baffles-1.jpg

C19 is 1000uf Pana FC.
C21 is 150uf Pana FC.
You can use your 470uf at C19. C21 should be smaller.

C9 is Vishay mkp1837 0.1uf. This was the best i found after trying twelve different caps here. Big ones don't work properly and generally mess up the sound, sometimes in a big way.
C9 is in my opinion one of the most important caps on the board. This also the only surface mount that i removed.

C16,17,20 and 25 are Oscon 150uf 16v. Probably enough Oscons.
C3,C18 are 220uf Samwha caps(gold), but any decent quality cap suitable for purpose would perhaps do.

C8,11,12,13 are 3.3uf Tants.

All lytics are bypassed by 0.1uf film caps.
I think blue Vishays are preferable to Wima, anywhere on this pcb.

This DAC works very well. Hope this is helpful.



Thanks, I will give that a go if the board returns to a stable state ;)

Tripmaster
10-09-2008, 21:50
Thanks, I will give that a go if the board returns to a stable state ;)
Hi

Late on Tuesday night...

Whilst listening to the DAC last night one of the temporary solder connections attaching the right channel lead to the coupling cap broke away.
I re-soldered the lead and at the same time decided to remove the two blobs of solder from the chip on the base of the board. I fired it up again and only the left channel works. All of the connections seem fine and both channels from my amp operate ok via a DAC MAGIC2 box.
I re-soldered the two blobs next to the chip and now the output is very quiet and extremely distorted.

This is not a good way to finish the evening!


Wednesday

Well...I have removed quite a number of the components previously added but am no further forward with regards to sound, still very distorted.

When I desoldered the small blobs from the base of the board next to the chip I think I may have damaged the solder pads (Solder is not sticking any more and they are barely visible).

Do you think I may have damaged this chip due to excessive heat? Or do you think the board is a right off?

Thanks in advance

Richard

Puffin
11-09-2008, 07:08
Just send it to me ;)

gurusan
11-09-2008, 12:13
hey everyone I am trying to bypass the onboard output stage to go directly to my headphone amplifier.

I removed the 4 capacitors in the middle of the board and connected the RCA outputs to where -R and -L caps were.

Now I am reading 2.5V on both outputs :/

What have I done wrong?

http://img386.imageshack.us/img386/9568/img3601di7.jpg

Makgajwer
11-09-2008, 12:48
hey everyone I am trying to bypass the onboard output stage to go directly to my headphone amplifier.

I removed the 4 capacitors in the middle of the board and connected the RCA outputs to where -R and -L caps were.

Now I am reading 2.5V on both outputs :/

What have I done wrong?

http://img386.imageshack.us/img386/9568/img3601di7.jpg

This is normal. You should add a 2,2uF capacitors (polypropylene or so) in series to outputs.

gurusan
11-09-2008, 12:57
ok thanks. I have some 2.2uF MMK EVOX caps and some 3.3uF Philips MKC caps.

Which would be better? sorry I'm new at this....

Makgajwer
11-09-2008, 13:07
ok thanks. I have some 2.2uF MMK EVOX caps and some 3.3uF Philips MKC caps.

Which would be better? sorry I'm new at this....

You should use this one which sounds better.

gurusan
11-09-2008, 13:14
gotcha. thanks :)

Tripmaster
11-09-2008, 20:46
I have been playing around with some of the recommend cap values (I know the caps are a bit large)

I have tested the output from the psu and I am getting 12v & 5v

How do I check when connected to the board?

Tripmaster
11-09-2008, 22:12
Hi

I have checked all of he solder joints and I cant see bridging. I also checked the Digital voltage from the CS8416 and I am getting a reading of 3.33v on pin 23 and 21. I also have a reading of 13.24v from pins 4 and 8 on the opamp.

I have not yet checked all of the tracks for continuity.


After some considering my options I decided to remove the chip from the base of the board.

Can anyone advise which pads I need to short for bypass?

Tripmaster
12-09-2008, 08:57
I have check there is a connection along each track between all of the changed or removed capacitors.

Does anyone know the value of C7?

gurusan
12-09-2008, 09:01
as far as I understand it you already soldered the 2 little pads on the bottom of the board next to the chip which effectively bypasses it. Physically removing the chip doesn't do anything extra.

Tripmaster
12-09-2008, 12:18
as far as I understand it you already soldered the 2 little pads on the bottom of the board next to the chip which effectively bypasses it. Physically removing the chip doesn't do anything extra.

I had problems soldering these pads at first because there may have been some lacquer/dirt over the contact point. I think the excessive heat trying to get a good contact may have damaged the chip. There was very little left of the solder pads prior to removing the chip.

alb
12-09-2008, 15:14
A useful link to identify surface mount caps.

http://library.solarbotics.net/pieces/parts_elect_pass_cap_code.html#sm

C7 is labelled 104 according to the schematic.

The first two digits are the base numbers ie 10.
The third digit is the multiplier, in this case 4.

So C7 is 100000pf = 100nf = 0.1uf.

Tripmaster
12-09-2008, 17:03
A useful link to identify surface mount caps.

http://library.solarbotics.net/pieces/parts_elect_pass_cap_code.html#sm

C7 is labelled 104 according to the schematic.

The first two digits are the base numbers ie 10.
The third digit is the multiplier, in this case 4.

So C7 is 100000pf = 120nf = 0.1uf.

Thanks for the info, very helpful!

Do you happen to know which chip solder pads need to be connected to enable a bypass?

iihay
12-09-2008, 22:45
Thanks for the info, very helpful!

Do you happen to know which chip solder pads need to be connected to enable a bypass?

Answer is lurking here:

http://theartofsound.net/forum/showthread.php?t=315&page=11

Iain

Tripmaster
13-09-2008, 19:25
YES!

I've just soldered the the link in the past five minutes. The sound is back albeit only the right channel, and its not distorted. It is however sadly lacking in the bass department. Progress!

I have checked the connection between pin one of the opamp to the left output pin and its fine. I attached the meter to the right output pin but could not make a connection with any of the other opamp pins. I do have a connection between the blob on the base of the board and the right output.

I have tested the leads to the rca sockets and they appear fine.

Any suggestions?

Richard

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/SL272591.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/SL272593.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/SL272595.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/SL272596.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/SL272597.jpg

leo
14-09-2008, 02:37
well done, at least you have some progress!

Don't suppose you have a scope? it would make finding whats causing the missing channel much easier

Regarding the bass, see the bottom pic with the three oscons? try changing the middle oscon to say a standard 10uf electrolytic, let it run for an hour and see if it makes any difference, that cap is for the dacs internal analogue bias, low ESR is good here but I personally didn't like a oscon in that position

also the large pair of caps in the middle of the board seem to be non polar, these are not really suited for supply decoupling, try a couple of polarised caps there instead

I know its hard but also try to keep the legs of the caps as short as you can, your adding inductance with the long legs, the caps should have as short lead length as possible to the chips, it makes them more effective

leo
14-09-2008, 02:50
I'll be starting another dac thread soon:eyebrows:

Tripmaster
14-09-2008, 08:25
well done, at least you have some progress!

Don't suppose you have a scope? it would make finding whats causing the missing channel much easier

Regarding the bass, see the bottom pic with the three oscons? try changing the middle oscon to say a standard 10uf electrolytic, let it run for an hour and see if it makes any difference, that cap is for the dacs internal analogue bias, low ESR is good here but I personally didn't like a oscon in that position

also the large pair of caps in the middle of the board seem to be non polar, these are not really suited for supply decoupling, try a couple of polarised caps there instead

I know its hard but also try to keep the legs of the caps as short as you can, your adding inductance with the long legs, the caps should have as short lead length as possible to the chips, it makes them more effective


Hi Leo

Thanks for your reply

I meant the left channel is working, not right. I intend to cut the legs much shorter when this bloody thing works! ;)

I have put a meter between pin 1 of opamp and L-OUT and its fine. But I don't get a reading between pin 7 on the opamp and R-OUT. If I turn up the volume on the amp and turn the balance control right around I can hear I faint sound (Probably cross-talk)

Completely stumped

Tripmaster
14-09-2008, 08:28
Is it definitely pads 3+4 to short for the chip on the base of the board?

fordgtlover
14-09-2008, 12:36
Hi to all

This is my first post on this forum.

A brief intro first. My name is Rob, and while I don't really know a great deal about electronics, I do enjoy building DIY audio gear.

I'm posting here as it seems to be the most active site for mods to this neat little DAC.

I've had mine for a few weeks now, and have done a bit of tinkering. Now that I have a configuration that I think sounds pretty good, I thought I'd post it.

Here is my current configuration:
R2 - 3K (SMD)
C3 - 330uF 6.3V os-con
C8 - .22uf MKT EPCOS
C9 - .1uF MKP416 BC Components
C10 - .22uf MKT EPCOS
C11 - .22uf MKT EPCOS
C12 - .22uf MKT EPCOS
C13 - .22uf MKT EPCOS
C16 - 10uF 16V Tantalum
C17 - 330uF 6.3V os-con
C18 - 10uF 16V Tantalum
C19 - Rubycon MBZ 1500uF
C20 - 330uF 6.3V os-con
C21 - 470uF Panasonic FM
C22 - 22nF (SMD)
C23 - 1nF (SMD)
C25 - 10uF 16V Tantalum
Opamp - opa2107

Leo's bypass mod
C27,C28,C30,C31,C6,C54 - Bypassed
L & R out C - 2uF Obbligato PIO
L & R out R - 51 Ohm Kiwame

I am running this from my PC via a HAGUSB USB->S/PDIF convertor, which sound stacks better than the S/PDIF straight from my motherboard, into my DIY AMB labs M³ amp and finally out to my AKG K701 headphones.

This little DAC sounds better inplace of either of my CD players, a NAD 5320 and a Rotel RCD-855.

Thanks to all here (and other sites) for the ongoing tinkering and information - particularly Leo.

Cheers

Rob

fordgtlover
14-09-2008, 13:15
Sorry to be post-whore on my first day.

I just wanted to relay one of my experiences with this DAC.

One of the first mods I made was to replace C19 & C21 with a Rubycon MBZ 1500uF and an os-con 330uF respectively. I found that the bass rolled off dramatically.

Noting that Leo mentions in one of his posts that the os-cons can kill the bass, and that he used 10uF tantalums in these positions, I swapped the caps for the tants but found that the DAC sounded rather dull and un-dynamic (sorry Leo). I then swapped the tants for the Rubycon MBZ 1500uF at C19 & this time a Panasonc 470uF FM, rather than the os-con, at C21. The dynamics returned, as did the bass.:)

Cheers

Rob

leo
14-09-2008, 14:28
Reason I use a tant for C19/21 is because I use a far superior external psu , these two caps are for the input and output of the 3.3v regulator coming straight from the main 5v, if the main 5v isn't up to scratch you need to add some good low esr there, with my supply large values destroy this fairly neutral sound I currently have

C25 is the only place I'd recommend a 10uf tant as that cap is quite critical, it needs to be low leakage, of course some electrolytics may be better but tend to be more leaky , its something you just have to experiment with

I tend not to post fit this type of cap here and this type there mainly because my external regulated psu is nothing like the supplied unit so results would vary, what sounds good with mine may be crap with the standard one

leo
14-09-2008, 14:39
Is it definitely pads 3+4 to short for the chip on the base of the board?


Don't get too excited but I may have spotted the problem (unless its the pic you posted playing tricks on my eyes)

Look at the two pads next to that chip which we short out for the right channel, look closely, is the metal pad missing on both?
The third chip pin on the left must have continuity to the top pad to be shorted out and go to the tiny VIA on the left, if you can post a larger pic I can edit I'll try to explain what I mean and what needs measuring

leo
14-09-2008, 15:56
http://i273.photobucket.com/albums/jj220/kingbusoms/SL272591.jpg

alb
14-09-2008, 15:58
Yes, i think it's very much a case of suck it and see, with this dac.

I posted details of mine because it works nicely and doesn't have any obvious weakness. A good starting point.
There are of course an unlimited number of ways that this can be achieved.

Another thing that i think is worth trying.
Remove the crappy pcb mounted phono socket and fit a BNC socket to the dac case and source. Connect with some proper 75 ohm cable and you should notice an improvement.
One advantage of doing this early on, is that it creates more space on the pcb, and makes life easier, especially when experimenting with input caps.

leo
14-09-2008, 16:06
Good point Al, those supplied sockets are really quite poor anyway, I had some problems with mine , intermittent SPDIF lock caused clicking/ticking noises

Tripmaster
14-09-2008, 19:50
http://i273.photobucket.com/albums/jj220/kingbusoms/SL272591.jpg

Thanks Leo

I have revealed the copper track and if I put a meter between here and pin 7 on the opamp I get a connection. I didnt get a connection between the blob and the copper. I desoldered the blob and tried to solder a link between the copper and the blob but I ended up removing the solder pad ARGHHHH!!!

I also pulled away the pad from the C19 negitive side, double ARGHHHH!!!

This is one project thats definatly not gone to plan :(

Can you recommend any other hidden tracks?

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/lifted.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/padremoved.jpg

Tripmaster
14-09-2008, 20:07
I have sorted C19. I scraped away some of the screen I now have a good connection to the cap

leo
14-09-2008, 20:18
I think (going from memory, will have to check to be sure) you can wire from that solder blob (next to the R) to the first VIA (hole) between the chip pad on the left in that pic, it will need some scraping to expose the copper

Stick your meter on that first hole (inside that white rectangle between pin3 and 4) to see if it measures continuity to the Rout socket pin

leo
14-09-2008, 20:24
Hold on, let me edit the pic

leo
14-09-2008, 20:33
Try that
http://i273.photobucket.com/albums/jj220/kingbusoms/padremoved.jpg

Tripmaster
14-09-2008, 20:50
Yippee! STEREO

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/RemoveScreen.jpg

I decided to scrape off the screen around the pins and it became obvious where everything should go.

Leo you are a life saver! :)

leo
14-09-2008, 21:18
Well done!

You just have to tune it up to your taste now;)

alb
14-09-2008, 22:09
I love a happy ending:).

fordgtlover
15-09-2008, 08:03
Reason I use a tant for C19/21 is because I use a far superior external psu , these two caps are for the input and output of the 3.3v regulator coming straight from the main 5v, if the main 5v isn't up to scratch you need to add some good low esr there, with my supply large values destroy this fairly neutral sound I currently have

C25 is the only place I'd recommend a 10uf tant as that cap is quite critical, it needs to be low leakage, of course some electrolytics may be better but tend to be more leaky , its something you just have to experiment with

I tend not to post fit this type of cap here and this type there mainly because my external regulated psu is nothing like the supplied unit so results would vary, what sounds good with mine may be crap with the standard one

Which PSU are you using, if you don't mind me asking?

Tripmaster
15-09-2008, 17:28
I love a happy ending:).


Me too!

I haven't finished soldering yet ;)

alb
15-09-2008, 19:47
I too had problems with lifting tracks. Ended up with a piece of wire from input cap to chip pin.
It's the first time i've had to work with surface mounts, so i'm not exactly an expert.
All good fun though.

leo
15-09-2008, 20:00
Which PSU are you using, if you don't mind me asking?

Schotty bridge, C-R-C + Cap multiplier feeding a shunt to lower the impedance, its knocked up on vero

alb
15-09-2008, 20:18
Any chance of a diagram Leo.
just in case i get bored one evening.

Tripmaster
15-09-2008, 22:04
http://i374.photobucket.com/albums/oo189/Tripmaster_photo/WithTants.jpg

Hi,

Legs a little shorter.

Considering the components have had very little burn-in time its sounding rather good. Its a little bit bright at the moment but I guess this will settle down with time. Bass is still a little shy.

Richard

alb
15-09-2008, 22:41
Richard.

The surface mounts are the bypass caps, in most cases anyway.


Most of the electrolytics are already bypassed with an adjacent surface mount cap. Its these we are trying to improve, on as much as anything.

http://www.lampizator.eu/LAMPIZATOR/LAMPUCERA/schemat.jpg

I 've bypassed the surface mounts with a variety of 0.1uf poly caps which are hopefully better quality. This now means the lytics have bypass caps of 0.22. I could have added some 0.01uf as well, but space is tight.

I can't remember much about the burn in except that it didn't seem to change much after a few hours constant use. It's quite some time since i first fired it up.

Tripmaster
16-09-2008, 10:09
I have a few questions…

I intend to house the DAC pcb and transformer/power supply in separate aluminium enclosures. I have some 4 pole Speakon connectors and five core mains cable for the umbilical. Can I combine the analogue and digital grounds together?

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/speakonconnector.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/speakonconnector2.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/powersupply.jpg

I also have some light weight coax for connecting a coax socket to the pcb. Will this be ok?

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/coax.jpg

Leo said the LM4562 needs a lot of tweaking for it to sound half decent. Will it sound better than the stock opamp without tweaks?

Thanks


Richard

p.s I am available for hand modelling if anyone is interested ;)

leo
16-09-2008, 22:25
Any chance of a diagram Leo.
just in case i get bored one evening.

It was cobbled together from various circuits so I'll have to try and draw it out

Basically a centre tapped transformer, centre tap as ground, both the secondaries go to a MBR20200 double schottky, 2200uf- 1R resistor - 2200uf- cap multiplier-shunt
Cap multiplier is based on this http://www.pinkfishmedia.net/forum/showthread.php?t=39990
Shunt uses a transistor as a current source

I might seperate the 3.3v supply on the dac board and feed it its own shunt, I don't like the layout of this dac, also the fact that the main 5v is too far away from the circuit

leo
16-09-2008, 22:42
Leo said the LM4562 needs a lot of tweaking for it to sound half decent. Will it sound better than the stock opamp without tweaks?

Thanks


Richard

p.s I am available for hand modelling if anyone is interested ;)

The LM4562 is a double edged sword, it offers good performance but only when used with care, its also crap at driving capactive loads, it needs a really good clean low impedance supply to work its best , ideally regulation needs to be close not run from lengths of cable which will likely result in a thin sterile sound.
Basically a slower lower spec op-amp may be better and easier to use

The dacs analogue and digital grounds are all on the same ground plane on the dacs pcb, normally best to keep the ground impedance as low as possisble , I'd personally run them all separate from the regulated PSU to the dac board, you can try running all together but it may sound worse
I don't have one of those supplies, do the ground wires from that supply have separate grounds or are they just fed from a ground plane same way as the dac board? if the regulated psu has a single ground plane for both supplies multiple/separate ground wires to the dac probably won't make any difference, if that makes sense:)

Nice pics BTW, pin sharp on my monitor!

Tripmaster
17-09-2008, 12:55
The LM4562 is a double edged sword, it offers good performance but only when used with care, its also crap at driving capactive loads, it needs a really good clean low impedance supply to work its best , ideally regulation needs to be close not run from lengths of cable which will likely result in a thin sterile sound.
Basically a slower lower spec op-amp may be better and easier to use

The dacs analogue and digital grounds are all on the same ground plane on the dacs pcb, normally best to keep the ground impedance as low as possisble , I'd personally run them all separate from the regulated PSU to the dac board, you can try running all together but it may sound worse
I don't have one of those supplies, do the ground wires from that supply have separate grounds or are they just fed from a ground plane same way as the dac board? if the regulated psu has a single ground plane for both supplies multiple/separate ground wires to the dac probably won't make any difference, if that makes sense:)

Nice pics BTW, pin sharp on my monitor!

Hi Leo

Thanks for your reply.

Here is a picture of the rough layout I am proposing. Please ignore the cables in the picture.

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/DACLayout.jpg

I plan to run all signal cables along the top and power along the bottom of the picture. In your previous reply you suggested keeping the power supply close to the DAC pcb. Am I going to suffer any RFI with this configuration? If this is OK I intend to mount the transformer in a separate enclosure with mains umbilical between the two.

Which opamp do you think works best for this project?

Picture not so good this time ;(

Thanks

Richard

leo
17-09-2008, 14:47
It should be ok, try not to have the lead from the transformer too long either if possible, its normally a transformer when run too close to signal that causes noise pick up, toroidals radiate more crap

What I'd do if possible is before drilling holes etc in the case try to run it as you intend to lay it out first, obviously you have to be careful not to short anything etc, if it sounds ok then you can then start bolting things together.
Using shielded cables for the signals should also help prevent unwanted noise pick up, having unshielded signal cable running close to supply cable as you know can give hum

Regarding the best op-amp, I've honestly no idea, they tend to be a personal taste thing, I mainly use AD826 because of its good output drive, its also much harder to implement than the LM4562 though
OPA2111 sounds ok and usually behaves, some people like OPA2107, opa2132 AD8066 and AD8620, the last two are smd so need to be fitted on brown adaptors, they also have a max voltage of +/-12v
The fet input based ones sound smoother if thats what you want

I have now added leads to allow using optional valve output stage, something I'm still messing about with:)

STIFA
20-09-2008, 11:18
Hi Leo

Thanks for your reply.

Here is a picture of the rough layout I am proposing. Please ignore the cables in the picture.

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/DACLayout.jpg

I plan to run all signal cables along the top and power along the bottom of the picture. In your previous reply you suggested keeping the power supply close to the DAC pcb. Am I going to suffer any RFI with this configuration? If this is OK I intend to mount the transformer in a separate enclosure with mains umbilical between the two.

Which opamp do you think works best for this project?

Picture not so good this time ;(

Thanks

Richard

Hi Tripmaster, just wondering what are those 2 black caps for? Could you explain your setup basically? I have done similar job, but still waiting to get Vishay for c9 . My board's analog part is intact, all I will be doing is remove those e-lytes after digital and steal signal for my lampizator tube stage.
Just interested in your approach...and those 2 black caps :)

jackenhack
24-09-2008, 22:55
Hi!

I just want to thank for all the great help I found here.

Just for fun I did a measurement of my Lawrence DAC before I started modding.

This is the frequency measurement before doing anything.

http://img.skitch.com/20080924-pp4n29dp3u5r6ekfnsg4npqr64.jpg

As you can see, the -3db point is at 13.000hz!

After changing out the resistors to the same value and then changing C32, C33, C34, C35 to 150pF it now looked like this!

http://img.skitch.com/20080924-kjb489jci1hjk3c7ftc4srp1et.jpg

And it sound great! My wife said that it sounded better than my extensivly modified Zero DAC. And I have to agree. Next step, better opamp and modify the power supply.

Thanks guys for the help!

alb
25-09-2008, 06:55
A good result then.
You are fortunate to have a wife who has sufficient interest, to be able to hear differences.
My wife seldom comments on anything, unless i emerge from the garage with some ridiculously large home made speakers.:)

Tripmaster
27-09-2008, 10:59
Im back!

Here are a few more photos of my project.

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/rear1.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/rear2.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/rear3.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/inside1.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/inside2.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/inside3.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/psu.jpg


Here is a list of the components changed so far...

C3 Jamicon 220uf
C18 Jamicon 220uf
C11 3.3uf Tant
C10 Vishay 0.1uf ERO 1837
C7 Vishay 0.1uf ERO 1837
C9 Vishay 0.1uf ERO 1837
C16 Oscon 150uf
C8 3.3uf Tant
C12 3.3uf Tant
C25 Oscon 150uf
C26 Oscon 150uf
C13 3.3uf Tant
C41 Wima 0.1uf
C19 Panasonic FC 2200uf
C21 Panasonic FC 470uf
C17 Oscon 150uf

Bypass caps are SCR 8uf with R68 resistors
Using standard JRC opamp and have also tried LM4562

The output sounds a little bit veiled and lacking sparkle. The internal coax is a 50ohm cable from Maplin. I was after 75ohm to match the 75ohm socket but the chap at Maplins said it would not make any difference because the cable length was so short.

Leo recommended keeping the transformer close to the DAC. Do you think I have mine too far?

Can you recommend changing any of the parts to help improve the sound?

Thanks

Richard

Tripmaster
27-09-2008, 11:04
Hi Tripmaster, just wondering what are those 2 black caps for? Could you explain your setup basically? I have done similar job, but still waiting to get Vishay for c9 . My board's analog part is intact, all I will be doing is remove those e-lytes after digital and steal signal for my lampizator tube stage.
Just interested in your approach...and those 2 black caps :)

Hi Stifa

Sorry for taking so long to respond I have been up to my neck in solder!

The two black caps are SCR 8uf coupling caps for the output. These were pinched from a pair of speaker cross-overs.

How are you getting on with yours?

Richard

Makgajwer
27-09-2008, 16:08
There is my project.
DAC is built in to the Cambridge Audio D500SE. Original output stage is unusable for me, so I decided to build tube stage. It's SRPP using E88CC. Sounds very good, far better than original opamp based circuit.
Photos and comments:
-purple frame original trafo moved to make room for additional trafo (for tubes)
-green frame additional trafo for tubes
-yellow frame filament supply
-red frame anode supply
-turquoise-coloured frame DAC board
-pink frame tube output stage
http://www.audiostereo.pl/zalaczniki/1167490_1.jpg
http://www.audiostereo.pl/zalaczniki/1167490_2.jpg
http://www.audiostereo.pl/zalaczniki/1167490_3.jpg

fordgtlover
28-09-2008, 02:42
There is my project.
DAC is built in to the Cambridge Audio D500SE. Original output stage is unusable for me, so I decided to build tube stage. It's SRPP using E88CC. Sounds very good, far better than original opamp based circuit.
Photos and comments:
-purple frame original trafo moved to make room for additional trafo (for tubes)
-green frame additional trafo for tubes
-yellow frame filament supply
-red frame anode supply
-turquoise-coloured frame DAC board
-pink frame tube output stage
http://www.audiostereo.pl/zalaczniki/1167490_1.jpg
http://www.audiostereo.pl/zalaczniki/1167490_2.jpg
http://www.audiostereo.pl/zalaczniki/1167490_3.jpg

What are those great little tube PCBs and where can I get some. Nice job BTW.

Cheers

Rob

fordgtlover
28-09-2008, 02:44
Hi!

I just want to thank for all the great help I found here.

Just for fun I did a measurement of my Lawrence DAC before I started modding.

This is the frequency measurement before doing anything.

http://img.skitch.com/20080924-pp4n29dp3u5r6ekfnsg4npqr64.jpg

As you can see, the -3db point is at 13.000hz!

After changing out the resistors to the same value and then changing C32, C33, C34, C35 to 150pF it now looked like this!

http://img.skitch.com/20080924-kjb489jci1hjk3c7ftc4srp1et.jpg

And it sound great! My wife said that it sounded better than my extensivly modified Zero DAC. And I have to agree. Next step, better opamp and modify the power supply.

Thanks guys for the help!

Thanks for the heads up on the FR. Did you change the resisters as discussed by Leo in this (http://theartofsound.net/forum/showpost.php?p=14693&postcount=96) post?

Where he suggests R8,R9,R14,R17 to 30k and C32,C33,C34,C35 to 150pf.

Puffin
28-09-2008, 06:47
Has anyone tried to connect a clock to this board?

Makgajwer
28-09-2008, 09:25
What are those great little tube PCBs and where can I get some. Nice job BTW.

Cheers

Rob



Thanks.
"Those great little tube PCBs" are custom made in Poland.

fordgtlover
28-09-2008, 13:05
What are those great little tube PCBs and where can I get some. Nice job BTW.

Cheers

Rob



Thanks.
"Those great little tube PCBs" are custom made in Poland.

Do you sell them?

Makgajwer
28-09-2008, 14:35
I don't sell them. I buy them ;-)

jackenhack
29-09-2008, 12:12
Thanks for the heads up on the FR. Did you change the resisters as discussed by Leo in this (http://theartofsound.net/forum/showpost.php?p=14693&postcount=96) post?

Where he suggests R8,R9,R14,R17 to 30k and C32,C33,C34,C35 to 150pf.

Yes, changed to 30k.

Jacken

madigwann
05-10-2008, 19:10
Yes, changed to 30k.

Jacken

Hi everyone ! Thank you for all the good infos provided here ! I've been reading about your mods with interest ! I'm not expert when it comes to electronics...I'm trying to get good results from this pretty dac too. Do you need to change the caps WITH the resistors, or can just the resistors be changed to 30k to get that amazing response curve jackenhack showed ? Thanks...

fordgtlover
06-10-2008, 11:15
I thought I might try the output stage that Leo previously posted here (http://theartofsound.net/forum/showpost.php?p=11637&postcount=50). So, I thought I would put my newly developed Eagle skills to work, and try a PCB layout.


I have recreated the schematic here:
http://img291.imageshack.us/img291/5071/os13schub6.jpg

And, here is my PCB layout:
http://img508.imageshack.us/img508/182/os13un3.jpg

The board is 40mm x 60mm and will require two for the complete output stage.

I have assumed BC550C for Q1, Q3, Q4 & Q6 & a BC517 for Q2.

I have used the Murata 1800R inductors, but I might swap them for the small Murata 2200R from Farnell.

You might note that for the output cap I have provided the option to use either a large(ish) film cap (c2) or an electrolytic bypassed by a small film cap (C9 & C10). Is this a sensible option and implementation?

Does C2 need to be 10uF? Can it be a smaller value?

I'd be interested in any feedback about the layout or the parts selections.

This is my first board layout, so all comments welcome.

Cheers

Rob

jackenhack
06-10-2008, 14:07
Hi everyone ! Thank you for all the good infos provided here ! I've been reading about your mods with interest ! I'm not expert when it comes to electronics...I'm trying to get good results from this pretty dac too. Do you need to change the caps WITH the resistors, or can just the resistors be changed to 30k to get that amazing response curve jackenhack showed ? Thanks...

I would go for both changes, but to alter the filter, change the caps. Here's a comparison with both res and cap change (MOD5) and just the resistors.

http://img.skitch.com/20081006-1y41xj4q4kwdmhmqi9futu7af.preview.jpg (http://skitch.com/jacken/a5h5/fr)
Click for full size (http://skitch.com/jacken/a5h5/fr) - Uploaded with plasq (http://plasq.com)'s Skitch (http://skitch.com)

StanleyB
06-10-2008, 14:10
Very nice longtailed pair cct. No feedback in the O/P stage:scratch:?

leo
06-10-2008, 17:08
This circuits proven to work well with the CS4397 Stan which is already high enough output, its also dead simple and cheap.
Its just a differential input to line out using a simple emitter follower.
Normally the emitter follower is a Darlington

fordgtlover
07-10-2008, 12:26
I wasn't very happy with the middle part of my board layout so I've redone it, I've also changed the indiuctor to the smaller version.

Any comments before I go ahead and get a couple of test boards made?
http://img80.imageshack.us/img80/81/os151xh5.jpg

madigwann
10-10-2008, 01:25
I would go for both changes, but to alter the filter, change the caps. Here's a comparison with both res and cap change (MOD5) and just the resistors.

http://img.skitch.com/20081006-1y41xj4q4kwdmhmqi9futu7af.preview.jpg (http://skitch.com/jacken/a5h5/fr)
Click for full size (http://skitch.com/jacken/a5h5/fr) - Uploaded with plasq (http://plasq.com)'s Skitch (http://skitch.com)

Hello ! Thank you very much ! I'll make both changes as I like to see that white curve so much ! How good was the sound after you did that change (MOD5) ?

leo
15-10-2008, 14:44
Just noticed , Q1 emitter should go to R3
You can alter the gain by adjusting the emitter resistors on the differential pair

This circuit is based on this http://home.quicknet.nl/qn/prive/ra.vdsteen/pdfs/SACDenhancer_evo3.pdf

leo
15-10-2008, 14:46
Another version here including pdf with layout http://home.quicknet.nl/qn/prive/ra.vdsteen/index_en.html

Ciu
20-10-2008, 12:49
Hello !

That's what I 've done :

http://img206.imageshack.us/img206/3926/jrbcfecc86rt8.png

I redrawn an original schematic published by JR Broskie in Tubecad.com
This is a buffer with no gain, and filtered input, with low voltage tube ECC86/6GM8
The B+ is 24v, simple LM317, with a slow turn-on , about 30s
The heater,6.3v, 600mA, is also a simple lm317, with a slow turn-on , about 10s
Works perfectly for a symetrical voltage DAC...
R.C.

Makgajwer
20-10-2008, 13:29
Your link not works

Ciu
21-10-2008, 10:48
Hello !
Works better now!
Sorry

Tripmaster
25-10-2008, 09:50
Im back!

Here are a few more photos of my project.

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/rear1.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/rear2.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/rear3.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/inside1.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/inside2.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/inside3.jpg

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/psu.jpg


Here is a list of the components changed so far...

C3 Jamicon 220uf
C18 Jamicon 220uf
C11 3.3uf Tant
C10 Vishay 0.1uf ERO 1837
C7 Vishay 0.1uf ERO 1837
C9 Vishay 0.1uf ERO 1837
C16 Oscon 150uf
C8 3.3uf Tant
C12 3.3uf Tant
C25 Oscon 150uf
C26 Oscon 150uf
C13 3.3uf Tant
C41 Wima 0.1uf
C19 Panasonic FC 2200uf
C21 Panasonic FC 470uf
C17 Oscon 150uf

Bypass caps are SCR 8uf with R68 resistors
Using standard JRC opamp and have also tried LM4562

The output sounds a little bit veiled and lacking sparkle. The internal coax is a 50ohm cable from Maplin. I was after 75ohm to match the 75ohm socket but the chap at Maplins said it would not make any difference because the cable length was so short.

Leo recommended keeping the transformer close to the DAC. Do you think I have mine too far?

Can you recommend changing any of the parts to help improve the sound?

Thanks

Richard

Hello again

I have been running this for a while and the bass is still poor. I don't like a really pronounced bass but after listening to a variety of music it is obviously lacking.

Can anyone provide a suggestion?

Thanks

Richard

leo
25-10-2008, 16:14
Hi Richard,

I still think the bass issue maybe the choice of caps fitted, ideally you need to look at the datasheet for the dac etc to see what each of the pins are for.
The cap on the dacs CMOUT has an influence, a tant will be different to say an Oscon.
I currently use a 5mm spaced 10uf MKT cap there, I wouldn't fit much higher than 10-22uf there

Seriously though I'd remove most of those tants,Oscons and FC's and fit some cheap normal ESR types to start with, see how it sounds and then do changes to lower ESR parts in stages

Also it seems you still use the original values for the differential output stage, changing the values to whats posted earlier does improve the sound , it also measures much flatter on the output

THE CS4397 bass is reported to be light , mine sounds pretty deep and tight

gurusan
26-10-2008, 12:25
So....without any warning my DAC started buzzing a whole ton through the outputs...the little crappy transformer that came with the DAC got VERY hot and started to melt my case.

Any idea what could have caused this? it was working fine for over a month. Just a cheapo transformer? What kind would you recommend for a cheapish but good replacement?

Also this morning I desoldered some bits from the PSU board to check out the components and see if I wanted to replace anything. The little green caps look very cheap and I don't know what sort of brand they are. According to the writing on the front I assume they are rated at 100V and .0035uF?

What caps would be good replacements for these? I was looking at Wima MKP10 .022 or .033uF? Anything cheaper?

And the big electrolytic caps are Rubycon YK for the 4700uF, and Elna RJB for the 1000uF.

Would I benefit in changing these to something else? What would you recommend?

http://img371.imageshack.us/img371/7447/img3663ys2.jpg
http://img49.imageshack.us/img49/9602/img3664mu3.jpg

Tripmaster
26-10-2008, 12:32
Hi Richard,

I still think the bass issue maybe the choice of caps fitted, ideally you need to look at the datasheet for the dac etc to see what each of the pins are for.
The cap on the dacs CMOUT has an influence, a tant will be different to say an Oscon.
I currently use a 5mm spaced 10uf MKT cap there, I wouldn't fit much higher than 10-22uf there

Seriously though I'd remove most of those tants,Oscons and FC's and fit some cheap normal ESR types to start with, see how it sounds and then do changes to lower ESR parts in stages

Also it seems you still use the original values for the differential output stage, changing the values to whats posted earlier does improve the sound , it also measures much flatter on the output

THE CS4397 bass is reported to be light , mine sounds pretty deep and tight

Hi Leo

What would you recommend I remove/replace to start with?

Do you have a list of the components/brands you have used on your board and their location?

I really appreciate your time

Richard

leo
26-10-2008, 17:14
I'm not using that standard regulated psu (I just brought the dac module only) so the caps I use may not work so well with that supply you use, problem is your using a lot of low ESR caps like Oscon and Panny FC/FM which are good caps but can cause stability problems if a standard reg like LM317/337 is used, the output pin on those regs can ring with too low ESR.
Apart from measuring worse ringing can make the sound bad too

I mainly started the thread as it was a fun and a cheap project, the dac as standard has obvious flaws like having all its 5v and 3.3v supplies shared from a single source, the PLL filter values are wrong and so is the Differential output filter, most things can be improved though and most is covered in the thread.

TBH I'd have changed each cap in stages and evaluated the differences, its easy to go changing a lot of parts at once but this can make things hard trying to isolate which parts are ruinning the sound


Only thing I can recommend is taking out the Oscons,tants and replace with a decent quality general purpose type, give it a few days to run in and see how it sounds. I'd then try changing the caps in stages, this also ensures you don't go too far in the wrong direction.
Modding this stuff to your own personal preference can be tedious but with a bit of patience usually pays off

BTW don't always trust what you read regarding low ESR caps, their not always better in every application;)

JDMz
26-10-2008, 18:21
Hi, I am planning to get this DAC and mod it according to the lampucera website AND the mods Leo showed. Will this have any conflicts?

Thank you

Tripmaster
26-10-2008, 20:24
I'm not using that standard regulated psu (I just brought the dac module only) so the caps I use may not work so well with that supply you use, problem is your using a lot of low ESR caps like Oscon and Panny FC/FM which are good caps but can cause stability problems if a standard reg like LM317/337 is used, the output pin on those regs can ring with too low ESR.
Apart from measuring worse ringing can make the sound bad too

I mainly started the thread as it was a fun and a cheap project, the dac as standard has obvious flaws like having all its 5v and 3.3v supplies shared from a single source, the PLL filter values are wrong and so is the Differential output filter, most things can be improved though and most is covered in the thread.

TBH I'd have changed each cap in stages and evaluated the differences, its easy to go changing a lot of parts at once but this can make things hard trying to isolate which parts are ruinning the sound


Only thing I can recommend is taking out the Oscons,tants and replace with a decent quality general purpose type, give it a few days to run in and see how it sounds. I'd then try changing the caps in stages, this also ensures you don't go too far in the wrong direction.
Modding this stuff to your own personal preference can be tedious but with a bit of patience usually pays off

BTW don't always trust what you read regarding low ESR caps, their not always better in every application;)


I was foolish to change so many components at once. Having read the http://www.lampizator.eu/ website and and also heard a friends modded DAC I decided to go for it!

I normally like to take things slowly and evaluate each change. I have recently built Peter Daniels Shigaclone transport with great results and wish I had taken the same approach with this project.

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/Use1edit-1.jpg

Should I stick with 150uf for the Oscon replacements and 220uf for the C3 and C18? And what about the tants, should I choose a larger value than 3.3uf?

Thanks Leo

Richard

JDMz
26-10-2008, 21:12
Do you guys think i should ignore the Lampucera site mods and do these ones instead? How about bypassing the opamp? Is that better or should i keep the opamp? I plan to replace the opamp to the HDAM sun ones from here http://www.audio-gd.com/enweb/pro/diy/OPA.htm

leo
27-10-2008, 00:43
I was foolish to change so many components at once. Having read the http://www.lampizator.eu/ website and and also heard a friends modded DAC I decided to go for it!

I normally like to take things slowly and evaluate each change. I have recently built Peter Daniels Shigaclone transport with great results and wish I had taken the same approach with this project.

http://i374.photobucket.com/albums/oo189/Tripmaster_photo/Use1edit-1.jpg

Should I stick with 150uf for the Oscon replacements and 220uf for the C3 and C18? And what about the tants, should I choose a larger value than 3.3uf?

Thanks Leo

Richard

Not foolish, just a little too eager;)

The original values for most of the decoupling caps was 100uf, the value isn't really critical so anything around 100-220uf should be ok

BTW, first easy thing to try is swapping C25 to a 10uf or 22uf cap, it looks like you have a big Oscon there, 150uf is too big for the CMOUT pin in my opinion, thats for the dacs internal bias
If that don't help much try changing just the other oscons and tants to general purpose 100-220uf, its worth leaving those two panasonic FM/FC caps until last and see what its like without the oscons and tants, we need to find a point where you start to hear improvements and then work from there

Very nice job on the Shigaclone :smoking:

leo
27-10-2008, 00:50
Hi, I am planning to get this DAC and mod it according to the lampucera website AND the mods Leo showed. Will this have any conflicts?

Thank you

No, the caps don't seem as critical if using a valve output stage, one lot of caps may sound awful when the op-amp is used but tends to be far less awful when using the valves

Also with the Lampizator circuit your only using half of the differential outputs of the dac so don't take full advantage of the noise cancelling that this type of chip has

leo
27-10-2008, 00:58
Do you guys think i should ignore the Lampucera site mods and do these ones instead? How about bypassing the opamp? Is that better or should i keep the opamp? I plan to replace the opamp to the HDAM sun ones from here http://www.audio-gd.com/enweb/pro/diy/OPA.htm

Only mod I'd consider on the Lampizator site is the valve output stage, the use of lots of oscons as shown on there probably won't mate too well if using the op-amp, the lead length of those caps is way too long IMHO anyway

The valve stage sounds quite nice in a coloured way, some people will love it, others won't
I find it pleasant with this dac, it also covers a lot of the flaws, the op-amp stage makes things much harder to get to sound right

My advice is try both, see which you prefer:)

I've never tried that discrete HDAM Sun so no idea how it performs

JDMz
27-10-2008, 01:04
I plan to do your mods, and then get a HDAM opamp which was on the link of my last post. I am kind of new to electronics and need some elaboration on your instructions. What type of resistors are those 30k ones? (like wattage and tolerance) And the 68-75R.

Finally for the 150pf capacitors, would these work? http://cgi.ebay.com/Silver-Mica-Amp-Guitar-Capacitor-150-PF-500-V-3-for-2_W0QQitemZ180297776689QQcmdZViewItem?hash=item180 297776689&_trksid=p3286.c0.m14&_trkparms=72%3A570|66%3A2|65%3A12|39%3A1|240%3A131 8

Thank you

leo
27-10-2008, 01:12
The resistors are just phoenix MRS25 metal film 1% which I matched, nothing special, just what I had spare in my parts box, the 150pf caps are just cog ceramic, again its what I had so nothing special.
Mica is usually technically better but also physically larger, theres not a lot of room on this dac pcb, if you think the mica's will fit then they should be ok

leo
27-10-2008, 01:21
The resistors on the output are not critical, they are just to help with op-amps that don't drive capacitive loads very well, 22R to 100R should be fine

JDMz
27-10-2008, 01:39
Should the input capacitor be a good brand? like Mundorf or something?

What tolerance and wattage were your 30k resistors? Sorry for so many questions.. I am ordering now that's why

leo
27-10-2008, 09:16
It was 1:21am here so I'd gone to bed just after my last post:)

Any resistors I used was MRS25 0.6w 1% tol metal films, use the smallest you can find if it makes fitting easier, wattage is not really an issue here, obviously you don't need anything large over 1w, if the ones you buy come in packs I'd try matching them in pairs with your meter

SPDIF input cap is .1uf (100nf) Vishay MKP1837
My dac does not use output coupling caps, since doing the output stage mod in the differential (30k +150pf) output DC offset is near 0v
I'd only advise using no caps on the output if your pre-amp or amp has input caps just as a precaution

If your going to use output caps use any decent quality film type, I'm not sure what particular brand to recommend, everybody seems to have their own preferences for those

gurusan
27-10-2008, 17:55
How do you go about sizing output caps?