PDA

View Full Version : which caps for 7510 I/V stage



robyonekenoby
02-04-2009, 19:35
I'm trying some different caps in the signal coupling between dac chip and I/V opamp


The stock caps were 22uF standard electrolytics

then tryed rubycon ZA 22uF 35V: better bass, tighter and stronger, but too bright highs, sibilance in S and C vocals

then mundorf mcap (polypropylene metallized film) 10uF 400V: superb details, very clear and open, relaxed, natural and neutral ..... too neutral
... and too relaxed :scratch:

from a sound definition point they are the best caps tried, but lack in involving
I can't explain, lytics were not so clean but warmer, more involving

it's a nonsense, because they introduce distortion, far more distortion than film caps, and add harmonics

maybe I'm missing those (artificial) harmonics?

It's the famous tube effect?

Or what else? Where I'm wrong?

Somebody knows the cut frequency with 22uF? I don't have schematics, so I don't know about I/V stage input impedance, and wondering if using 10uF could lead to audible bass roll off

I'm not sure I want to buy expensive 22uF film caps and discover nothing change

thanks guys

NRG
02-04-2009, 20:23
Well....I use Nichicon Fine Gold in that postion, however, Blackgates would be good either the standard or FK types...

Theory dosen't always follow in the listening experience and its the pleasure in the listening that we are looking for.

robyonekenoby
02-04-2009, 20:42
Yeah, you are right, but unfortunately I'm affected by a perfectionism syndrome, want the sound as close ad possible to the reality

maybe I didn't give them enough break in, just 24 hours

Better wait other 100 hours and make another trial

But are we sure the break in affect components and not the brain ? :lolsign:

I was just carefully listening one of my preferite blues concert at loud volume, and I have to admit I've never heard such a detailed and not fatiguing sound, maybe what I miss is just a bit more bass kicking

hard to find blackgate in europe anymore

StanleyB
02-04-2009, 21:57
If you carry out MOD21, then you need 47uF:). Try http://cgi.ebay.co.uk/2-x-47uF-25v-Elna-Silmic-Audio-Electrolytic-Capacitor_W0QQitemZ220380673761QQcmdZViewItemQQptZ LH_DefaultDomain_3?hash=item220380673761&_trksid=p3286.c0.m14&_trkparms=72%3A1688|66%3A2|65%3A12|39%3A1|240%3A13 18 or http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&ssPageName=STRK:MEWAX:IT&item=380106087936
Both go very well with the LM4562MA upgrade to the TC-7510.

Stan

robyonekenoby
03-04-2009, 10:03
Wow, I got a reply from the 7510 engineer

Thank you Stan, I've already changed the I/V opamp to AD8620, faster slew rate than 4562

So I have to increase value for input coupling caps; was convinced that changing from lytics to film is possible to reduce capacitance

Ok I'll order two mundorf 22uF :doh:

anyway, could you give me an idea of the input impedance of the I/V stage, just to calculate the cut frequency for a given cap

Thanks

NRG
03-04-2009, 13:14
Faster slew rate dosn't mean its better ;) Also, Stan will correct me if I'm wrong, the opamp and cct you are tweaking is not an I/V stage, the PCM1716 outputs a voltage already.

robyonekenoby
03-04-2009, 13:54
... so we are talking about a buffer or what else ?

I can tell you I've already tried to connect 1716 exit directly to the RCA connectors, but the sound was extremely thin and bass lacking

I'd like to skip the entire buffer (?) section, in a minimalist approach, but in this case I don't really know if I need a passive filter or what else

StanleyB
03-04-2009, 15:16
I think that some of you guys are mixing up the design of the PCM1716 with that of say the PCM179X range. I recommend that you download the PCM1716 datasheet from TI and have a read up of the recommended design parameters for the PCM1716. There is no I/V stage output on the PCM1716. It's a direct V output. TI recommends using 10uF electrolytic with the + side facing the PCM1716 analogue output pin. HOWEVER, if you want to listen to the DAC that I designed, then you have to use a 47uF electrolytic with the - side facing the PCM1716 O/P pins. You can try to analyse it and ask questions as to why, but you won't get an answer from me :eyebrows:.

robyonekenoby
03-04-2009, 20:22
Thanks for the reply Stan

but ....

the negative pole facing the dac output?

I understood well?

I measured few volts dc bias, how can it work without burn the lytics?

StanleyB
04-04-2009, 06:33
I measured few volts dc bias, how can it work without burn the lytics?
Its like asking how come a bumble bee flies;).

robyonekenoby
04-04-2009, 09:51
Ok Stan, I'll order those two silmic and give'm a try

Today I repeated the experiment of linking 1716 Vout directly to RCA connector via coupling polyprop film cap, skipping completely the entire analogic section

Well, the result is quite interesting, the most clean, smooth and detailed sound soo far against all the other combinations; curious, I got better bass using 3.3 uF than 10uF, probably because the first is film and foil while the second is metallized film

but

the music is cold, without life, uninvolving, I can't explain why
maybe I should use bigger decoupling cap on VccL and VccR ?

and it's curious also that restoring analogue section and comparing lytics vs film caps for the same value in input, I get a warmer sound from lytics, while they are a little harsh in the highs

I'm starting to believe that we audiophile people love harmonics, don't care if they are natural or not

StanleyB
04-04-2009, 10:05
The TC7510 opamp location has very little PCB space to work with and carry out much in terms of more advanced mods. I would love to design a capacitor free output between the DAC chip and the fixed LINE output for the TC7510 that could fit in the limited space. If anyone has any suggestions or circuit ideas, please feel free to share it.

STan

leo
04-04-2009, 11:22
I'm sure those with a TC-7510 can run the output of a the PCM1716 to a piece of vero, your then free to try a number of circuits, its pretty easy knocking things up on vero especially a simple output stage
Once a rough circuit using through hole parts is proven to work it could be converted to SMD

Come on guys, get your fingers out:lolsign:

robyonekenoby
04-04-2009, 14:08
Sorry but my limited knowledge doesn't permit to understand some issues

Example: does the Pcm1716 really need a buffer, or it's just to ensure compatibility with any kind of amplifier?

As I posted before, seems to work without, but the sound is just too thin for my taste

I downloaded the 1716 datasheet, the application diagram shows 10uF to decoupling VccL and VccR (are the yellow tantalum(?) on the PCB?), maybe making them bigger could help?

I'm wondering if we find the way to bypass opamp and related circuitry, we could improve sound quality a lot

Anyway my next test is to try two panasonic non polar electrolytics connected back to back in series, as suggested by Cyril Bateman, in opamp input coupling position

robyonekenoby
04-04-2009, 14:17
Leo, I have one piece of vero :)

What could I try first? a low pass filter? but the 1716 already has one built in

:confused:

StanleyB
04-04-2009, 15:33
Sorry but my limited knowledge doesn't permit to understand some issues

Example: does the Pcm1716 really need a buffer, or it's just to ensure compatibility with any kind of amplifier?

As I posted before, seems to work without, but the sound is just too thin for my taste

The PCM1716 output level is a miserly 600mV, which is far too low for CD quality playback. So it needs an output amplification stage to take it to something more respectable.

robyonekenoby
04-04-2009, 17:44
Yes I know 600 mV is very low, but in theory you could get same volume just turning more your pot, and effectively it works, but still you feel a lack of energy in the sound

probably it's an ampere issue also, the dac out hasn't enough power to manage bass transient without Volt drop

ok, so we can't stay from an opamp

I continue my search for the best input coupling cap ;)

for the output I already have a winner, 3.3 uF audyn mkp Plus

leo
04-04-2009, 19:09
Leo, I have one piece of vero :)

What could I try first? a low pass filter? but the 1716 already has one built in

:confused:



I was thinking more of an active stage, fudging about with vero is easier and will save the track on your pcb

leo
04-04-2009, 19:14
I was thinking more of an active stage, fudging about with vero is easier and will save the track on your pcb

Something like a rail splitter and DC servo, lots of things to try

NRG
04-04-2009, 22:34
Its like asking how come a bumble bee flies;).

Well not quite, the laws for bumblebee flight have been known for a number of years... http://dragonfly.tam.cornell.edu/publications/2000_PRL_Wang_Hv.pdf but unfortunately the laws for magically rotating capacitors have not.

StanleyB
05-04-2009, 01:24
but unfortunately the laws for magically rotating capacitors have not.
Good:). Since I intend to use the technique for many other purposes in future projects, I shall remain silent on the how and why;).

Stan

robyonekenoby
08-04-2009, 10:29
I've tried a bunch of capacitors in opamp (line out) input coupling position, and I found all lytics give a warm sound, with deep but rounded bass, low ESR give brighter sound while higher ESR (standard cheap caps) are more veiled and laid back, but all of them lack in transparency and detail, and some are harsh and grainy, especially audible in the female vocals

Metallized film caps are way better in definition and smoothness , but I feel a lack in bass weight and low midrange energy, they don't sound really thin but the upper midrange is too forward, too emphasis in the highs

Now I wanna try film & foil caps

I don't have 7510 schematics, and wonder if somebody could provide me, or at least know the input impedance of the analog output stage seen from the dac, just to calculate the cut frequency for a given capacitance

Film & foil are quite expensive, I'd like to buy the correct value at first shot

thanks

jkeny
21-04-2009, 02:02
Why don't you try bypassing your favourite lytics with a your favourite film caps?

NRG
21-04-2009, 06:35
Just use a good 'lytic in the first place

jkeny
21-04-2009, 11:41
yea, you're probably right, bypassing is not such a hot idea - better to find a single cap that gives you the sound you like.