View Full Version : The Birth and Evolution of a Phono Stage.
See pic for where I am currently.
Giz a clue then?... :)
Seriously though. Bloody good for you sir!
I started this thread in the hope that I could 'get the ball rolling' with this DIY lark here on AoS. So, if only one person gives it a go I'll be well chuffed! :youtheman:
Yeah!... Bring it on! :smoking:
I started this thread in the hope that I could 'get the ball rolling' with this DIY lark here on AoS.
And you did a grand job, sir! Look at this section of the forum now :)
Marco.
I've done this as it will stop me from drifting off course and ending up not bothering to all that I initially want. Also having everything in a nice chassis with sections on their own sub chassis makes it a lot easier to build and then modify neatly.
What I want:
Power Switch, Always a good start
High/Low Gain Switch (MM/MC)
5 selectable input impedances
5 selectable input capacitances
2 Stage valve topology with passive RIAA
A switchable pre stage for high gain
Wading Through Treacle:
What Pre gain stage, something very quiet like solid state or step up transformers.
What valves to use, I seen ECC88 circuits but also heard that they're not good for phono stages. Point to point as the enclosure is already set up for this.
Whether to use the spare outlet of my WD Pre PSU and regulate this further, to have a separate outboard PSU or a inboard one.
Hopefully at the end of it I can have something better than my Project Tube Box for a couple of hundred quid.
As you can see at our website P-13 pre-amp (http://www.hawkaudio.nl/amp/p13.htm) we designed a phono pre-amp with tubes and an op amp at the input.
Looking at the schematics there's V1. This is a 6N1P at a collector current of approx. 3 mA. This tube has a low noise level and low capacities so it's far better then an ECC83 for instance having a bandwidth of over 300 kHz.
Thereafter there's the passive RIAA.
At the output we use an SRPP with a 6922. That tube is very reliable even at high currents (and dissipation). In this case the current is approx. 12 mA and the output impedance is around 1,5 kOhm. The latter is brought down by the divider at the output (needed there because the amplification factor is too high!) to around 700 Ohm.
At the input you see an AD797. This is a very nice kind of op amp for this purpose. It's bipolar, so no fet's, and the inner configuration consists on amplifying current mirrors.
Around the op amp you could configure some switches (as in our P-14) to be able to fix the amplification factor, input impedance and input capacity.
John
Primalsea
02-09-2008, 19:08
Oh yer, heres the pic.
It would seem the most crucial element is low noise so I guess smooth PSU and low noise valves as Johnrtd suggested.
Is there a elegant way for a valve circuit to naturally have a HP filter with a roll off of around 10 to 15hz without having signal capacitors on the input.
I could bolt on a filter but would prefer not to
As shown in the P-13 diagram there's a capacitor at the output of the SRPP. It's needed there otherwise you'll have 100VDC at the output.
The value we use there is 2,2 µF MKP giving a roll off of 3 dB at 18 Hz.
There's also a capacitor at the output of V1, C1 = 1 µF MKP. This gives a -3 dB at 2 Hz. If you want to filter more effectively it is thinkable to put an extra RC between the op amp and the V1 input. In that way creating a decay of 12 dB/oct as of 18-20 Hz.
Essential in the design, as in most audio applications, is the power supply. Both circuits, op amp and tubes, are fed by a regulated voltage. The op amp supply has double filtering by applying a series regulator, the common LM317/337, and a shunt regulator. The noise value of that supply is at a healthy >-100 dB.
The noise value of the entire circuit is around -70/-75 dB.
John
Primalsea
03-09-2008, 20:02
OK,
http://i44.photobucket.com/albums/f23/primalsea/DSC00117.jpg
http://i44.photobucket.com/albums/f23/primalsea/DSC00118.jpg
I've got this far.
a Phono stage found on the web by RJM Audio has 6dj8's with plate voltage of 100V and current at 5ma, B+ 250 and plate resistor is 30K. After farting around for 30 mins I decided that its a good idea just to do the same. There's no point in doing differently for difference sake.
I decided upon using a low noise opamp as the first stage. It seems daft just to use the just to boost the output of an MC cartridge. I'm going to use split RIAA, one after the opamp and 1 after the the first valve gain stage. The last stage will give the final gain and buffer the 2nd RIAA stage.
I plan on balanced connection from cartrige to input for decent CM rejection.
Next I need to work out the output impedances, input impedances & stray capacitances of each stage so the RIAA can be calculated accurately.:scratch:
ooh!... This looks serious! :eyebrows:
I'm currently doing some 'chin rubbing' over things like this:
http://www.sacthailand.com/Transformer_LCR.html
Or:
http://home.earthlink.net/~jeremyepstein/6c45riaa.html
:scratch:
Yes you've got a wee pal to play with now, Mike :eyebrows:
How are your 'escapades' progressing?
Marco.
Already thinking about the next one.
Strictly speaking though, I should do a pre-amp first. :confused:
Have you 'perfected' your phono stage yet? Best to complete one project first before moving on to another.
I liked it at mine - it wasn't the finished article but it showed lots of potential.
Marco.
No such thing as 'perfected' in this game. ;)
I'll still be fiddling with this one, hence the 'evolution' edit to the title.
I was that impressed, though, by Andrew's LCR Phono stage that it just has to be done at some point! :)
*Note* Paul is at least doing it properly by designing his own from scratch! Load lines & everything. Rah! :clap:
Primalsea
03-09-2008, 21:43
Its all relative Mike, I might seem like a DIY warrior to some but to others who know more than me it's probably laugh worthy if they saw me having to keep looking in books and trying to work it all out.
OK,
http://i44.photobucket.com/albums/f23/primalsea/DSC00117.jpg
http://i44.photobucket.com/albums/f23/primalsea/DSC00118.jpg
I've got this far.
a Phono stage found on the web by RJM Audio has 6dj8's with plate voltage of 100V and current at 5ma, B+ 250 and plate resistor is 30K. After farting around for 30 mins I decided that its a good idea just to do the same. There's no point in doing differently for difference sake.
I decided upon using a low noise opamp as the first stage. It seems daft just to use the just to boost the output of an MC cartridge. I'm going to use split RIAA, one after the opamp and 1 after the the first valve gain stage. The last stage will give the final gain and buffer the 2nd RIAA stage.
I plan on balanced connection from cartrige to input for decent CM rejection.
Next I need to work out the output impedances, input impedances & stray capacitances of each stage so the RIAA can be calculated accurately.:scratch:
I have found LTSpice is invaluable for fine tuning RIAA designs.
to others who know more than me it's probably laugh worthy if they saw me having to keep looking in books and trying to work it all out.
I know the feeling! ;)
I have found LTSpice is invaluable for fine tuning RIAA designs.
I've found 'Old Spice' invaluable for keeping flies away :eyebrows:
Marco.
I think it's a bad idea to make the input differential. When using op amps it means using three of them and all in the signal path.
You could design a discrete differential circuit but that would take some 12 transistors.
It's far easier to have the preamp at a short distance from the TT. Also the impedances of cartridges are that low that it's almost unthinkable getting and detecting any stray field.
The 6N1P is nice but not reliable at high currents. So at the output I suggest using something else (and preferably a simple SRPP).
John
Primalsea
04-09-2008, 18:35
Noted Johnrtd but for now I've settled on the balanced opamp input stage as I can easily reconfigure it should it be better to goto unbalanced.
The opamp input stage is just a test anyway as technically it should be better than a transformer for MC. However this has to been seen in the real world as overload from clicks and pops my be a issue. I believe that active feedback stages can make them sound horrible while stages with none or passive local feedback can subdue them.
I have a program called Tina which is a Spice simulator and it does make things a lot easier.
So actually the next thing is to design the gain stages then calculate their impedances and strays.
Mikey what's the view on those valves, mate? :)
If you don't want them there's someone else who might.
Marco.
Err... umm... errr...... :scratch:
LOL. You didn't strike me as a dithery woofta! :eyebrows:
;)
Marco.
TBH I've not had much chance to play with them this week, been busy with work and farting about with a DAC thingy.
I guess it depends how much you want for them? I'm a bit poor for the rest of this month. :(
Primalsea
05-09-2008, 19:58
The input stage is taking a bit longer to design. The previous circuit I posted turned out to be a stinker. Also to change the gain and keep good CMRR you need to change 3 resistors per channel. That sounds like chest pain to me.
However Good Old Rod Elliot at ESP Audio has this circuit on his site which only have 1 resistor per channel to change gain. This looks a winner!!
CMRR
???
common mode rejection ratio, maybe?
Primalsea
06-09-2008, 08:46
???
common mode rejection ratio, maybe?
Everyone knows that, don't they :scratch::scratch:
The input stage is almost designed. R6 changes the gain and this will be selected by a DPDT switch. The gain is higher on this stage than some may expect. This is to give the phono amp a higher output overall. As this stage will be the most quiet it seems a good idea to squeeze a bit of extra gain from it. Later on this could allow passive attenuation on the valve stages to reduce their noise floor without ending up with a phono amp with low gain.
R1, R7, C3, C4 are for cartridge loading. There will be double pole 5 way rotary switches on the front panel of the amp to allow changes while listening.
I will probably use PCB mounted screw connection blocks to hold the gain and loading resistors and capacitors so I can adjust on test easily.
Output impedance should be very low and if luck I wont need to consider it.
Glad you guys seem to know what you're doing. I'm still at the stage of assembling ready made boards into boxes with clearly pre-drilled holes (I can see why you possibly need to smoke, with frustration(?) Shian7 :) I don't smoke, but I explode instead - too regularly right now... If the lung cancer doesn't get me, a heart attack will).
Good luck with your D.I.Y. projects all
Primalsea
06-09-2008, 16:41
Had a change of mind and have now come up with this circuit. It's really squeezing the gain out of the opamp stage and I'm not sure how well it will work.
If it works it means I can use single stage passive RIAA correction after the opamp circuit then use a low output impedance SRPP stage to boost gain to compensate for the 20db of loss from the RIAA stage. This means I can reduce the number of gain stages to 2 which will reduce noise as long as the high gain opamp stage is quieter than a lower gain opamp stage plus a valve stage.
Probably a good idea to make and test it now to see if it works ok.
Sounds promising. :eyebrows:
I particularly like the switchable loading, this is something I thought of incorporating in mine too. I was thinking of using those switches that you get for making stepped attenuators, but having that many loading resistors/caps might be over egging the pudding somewhat! :mental:
That SRPP is a good idea. You'll get a low output impedance and won't have to use expensive (low capacitance) interlinks.
That input circuit is quite complex. And why use a TL071??? That's a fet input one and for sure not a low noise type. In applications like this you could better choose a bipolar op amp and then a low noise type. A single 2132 would do the trick or better an AD797.
John
Primalsea
07-09-2008, 12:28
Hi John, Actually I've just ordered some AD797 opamps. The TL071 was just what was mentioned on the schematic that I found.
More design work:
This is the PSU for the opamp stage. Nothing really exiting I just like to pla it out so I know what to order and roughly how big it will be.
http://i44.photobucket.com/albums/f23/primalsea/DSC00121.jpg
I've changed my mind again and opted for a Mu Follower stage which is similar to SRPP. I've choosen to run it at 4ma but I'm nor sure if it should be a little more as its the final output stage of the amp. I like my amps to have a bit of bollocks, and hairy ones at that. No tizzy sounding crap here (if it goes well that is). Does increasing the plate current lower output impedance?? Anyone please tell me.
http://i44.photobucket.com/albums/f23/primalsea/DSC00122.jpg
This is the load line for the Mu Follower if you're really interested.
http://i44.photobucket.com/albums/f23/primalsea/DSC00119.jpg
The new plan with 2 gain stages
http://i44.photobucket.com/albums/f23/primalsea/DSC00120.jpg
I've ordered the stuff for the input stage and should be able to make a start next week when it comes.
That power supply is really funny. Why add noise into your circuit? And why use capacitors in parallel?
That MU-follower will give a lower output impedance. But the disadvantage could be the extra capacitor, which is in the signal path. I suggest using a polystyrene type there. Also the value of that 36,6 k resistor is far to high. IMHO around 3 k is a maximum for this configuration.
The used voltage is very very high, why not stay around 200 Volt.
The 6N1P does a good job at 3 - 4 mA. A larger current will result in a lower output impedance but at the cost of the life time of the tube. If you want a larger current use the 6922 instead. That one (from Sovtek) will give you 5 - 10 years without failure.
John
I'd be tempted to use 4 schottky diodes rather than a bridge. Maybe?
Schottky diodes might give a little improvement. But the main thing is that the power supply should be "seen" as part of the audio channel. So you need to design it carefully and "think" about what happens.
As stated before the rectifier will generate "spikes" which can easily be terminated. The capacitor should have a low inductance over the entire audio band up to 100 kHz (and preferably over that!).
The stabilizers you use have an internal capacitance from input to output. So, apart from their own generated noise, you will feed through any noise coming from the mains.
Best thing to do is measure what's happening with a reasonably wide band oscilloscope. Preferably with a 20 - 50 Mhz bandwidth and NOT a digital one.
Measure the supply with a load. In your case use the preamp as a load.
You could compare the output of the preamp, with shorted input, with the power supply output. And if you don't have an oscilloscope use a sensitive amplifier and listen! The latter is an old fashioned trick but it helps a bit and can give you some "insight" in what's happenning in your circuit.
John
Primalsea
08-09-2008, 12:51
I thought the PSU was fairly standard and although not the best it would be good enough for its purpose. Bare in mind that each opamp will have its own local filter cap probably of around 100 - 220uf which will reduce noise further.
When you say why parallel caps I think you mean why use 2 smaller caps in parallel instead of 1 big one. What I have done is bypass each larger cap with a smaller 100n cap that has much better high frequency performance. I have actually ordered Ultra Fast recovery diodes and will make the bridge myself. Each diode will be snubbed by 100n caps.
As for noise, where is the noise being added???
I looked at reducing HT and the 36.6K resistor to 3K but then it seems that the valve will start to run into grid current leakage. I though that you have to keep the grid well away from -1 volt or higher to avoid this. My design is set for the grid at -2.1 to keep the previous stages signal well away from -1V on negative swings.
As for noise, where is the noise being added???
Those three legged regulators are very noisy. The best solution I have found, is to use the three pin reg to regulate for source variations, then use a HFE multiplier with its base driven from a very well filtered DC source to produce a low noise source.
In the case of the Mu Follower, I don't see how the 36k resistor will change the grid conditions, they should be being set by the two 475R resistors.
You could always consider using a mosfet on the top.
Primalsea
08-09-2008, 18:50
In the case of the Mu Follower, I don't see how the 36k resistor will change the grid conditions, they should be being set by the two 475R resistors.
I was working on the basis that if the resistor was reduced to 3K I would have to reduce the HT as well otherwise the current would be too high. This would have the knock on effect having to operate the valves with a lower plate voltage which would mean a higher grid voltage (was -2.1V, now much closer to -1V).
I read that you get problems with grid current when the grid voltage gets too high as it starts to attract electrons.
This is the first time I have designed a Mu Follower let alone anything more complicated and than a cathode follower (which didn't work very well by the way). Have I missed something???
I'm aware that 3 pin regulators have a bad rap but I've never had their noise be a problem when using them with opamps. The bits have come so time will tell. I will post how well it goes with the PSU. If it goes badly I will have to badger you for the hfe multiplier info.
I was working on the basis that if the resistor was reduced to 3K I would have to reduce the HT as well otherwise the current would be too high. This would have the knock on effect having to operate the valves with a lower plate voltage which would mean a higher grid voltage (was -2.1V, now much closer to -1V).
I read that you get problems with grid current when the grid voltage gets too high as it starts to attract electrons.
Yes, you want to avoid A2 operation unless its intended. I wasn;t suggesting that 3k would be right or wrong, only that the grid voltage would be set by the resistors in the DC cathode feedback circuit, which wasn;t the lower valves load.
Part of the reason you are near the diode line is because your anode voltage is low, so to get the current, you end up with a low grid voltage. If you went for something like 150v @5ma, then you could have the grids at -3v giving pleanty of headroom.
6n1p are cheap and hardy, if you live down the bottom of the curves, you are in the less linear higher anode resistance area. that op point is still just over a third of the max diss.
I thought the PSU was fairly standard and although not the best it would be good enough for its purpose. Bare in mind that each opamp will have its own local filter cap probably of around 100 - 220uf which will reduce noise further.
When you say why parallel caps I think you mean why use 2 smaller caps in parallel instead of 1 big one. What I have done is bypass each larger cap with a smaller 100n cap that has much better high frequency performance. I have actually ordered Ultra Fast recovery diodes and will make the bridge myself. Each diode will be snubbed by 100n caps.
As for noise, where is the noise being added???
I looked at reducing HT and the 36.6K resistor to 3K but then it seems that the valve will start to run into grid current leakage. I though that you have to keep the grid well away from -1 volt or higher to avoid this. My design is set for the grid at -2.1 to keep the previous stages signal well away from -1V on negative swings.
Paralleling caps was a good thing to do when we had old fashioned electrolytics optimized for 50 - 60 Hz. Nowadays, the world is changed and newer technologies are developed, we now have electrolytics with low ESR over a far wider bandwidth. These were primarily developed for use in SMPS circuits, but in audio they also have an advantage. No paralleling is needed any more so you don't have to worry about the phase behaviour (when paralleling different caps).
Although you're using fast diodes in the bridge you didn't look at the generated spikes!
Those µA78.. and 79.. also are very old types (1972?) and for sure a nuisance in audio equipment. A far better way to do it is using a higher voltage from the transformer, for instance 24 Volt. Then stabilize using LM317/337 down to around 20-24 VDC. Then use a shunt regulator with a resistor in series going down to 15-18 Volt. That series resistor isolates the amplifying part from mains nuisances. A further improvement is possible using a CCS instead of that resistor.
If you stay using the proposed circuit mount a small resistor between the supply output and that 100-220 µF caps mounted near the chips. A 10 Ohm will do.
You didn't tell us why you're using a mu-follower instead of a simple SRPP. What's the benefit? It's more complicated and you'll have extra components in the signal path.
In our designs we have around 100 Volt per triode and a 180 Ohm resistor in the cathodes when using a 6N1P. And that works fine (THD around 0,01 %).
John
Primalsea
09-09-2008, 12:13
Hi John,
Good point about the shunt regs. there's loads of headroom with the 15v supplies so I still could fit the shunt mod later on. I was thinking of an RC filter but was looking at about 1 ohm, but current would probably be low enough that 10 ohm is possible.
I looked at SRPP & Mu Follower in detail and the Mu Follower looked more stable with regards to distortion over a larger range of loads. I do concede though that most preamps are a fairly standard with a stable load 47K and between 50p and 300p unless something odd is going on or high capacitance cables are being used.
As always its a case of weighing up the ratio of Time, Cost, Agro, Real Performance.
Well Paul if it's about "real performance" I would still consider using a simple SRPP. The behaviour (distortion) of the circuit can be optimized by loading it with an optimal load value. You don't have to worry about long cables if the control amp is near by.
I still think you didn't optimize the power supply by suppressing (filtering) the generated "spikes"???
John
I've got some time off work today so I'm going to do some playing with my own phono stage. Let's see if we can get this C4S bit working and slotted in. Hopefully I wont pop too many of these wee three legged critters! :)
Primalsea
10-09-2008, 11:47
Well Paul if it's about "real performance" I would still consider using a simple SRPP. The behaviour (distortion) of the circuit can be optimized by loading it with an optimal load value. You don't have to worry about long cables if the control amp is near by.
I still think you didn't optimize the power supply by suppressing (filtering) the generated "spikes"???
John
I have fitted 100n snubber caps across the diodes and I plan to use low pass RC filtering local the to opamp. What else can I do to clean the spikes up other than shunt regulators which I will use if noise is a problem???
You could find the "spike" matter at tips (http://www.hawkaudio.nl/tips.htm) and some power supply ideas at P-14 pre-amp (http://www.hawkaudio.nl/amp/P14.htm)
John
I have fitted 100n snubber caps across the diodes
I think I'll give this a go myself, can't do any harm. :)
I got the C4S section built and fitted the other day, mixed results though. At first I could have sworn it sounded a little smoother but now I'm not so sure, there is also a very slight hum with the volume wound right up and no record playing. I'm sure that wasn't there before, hmmm. :scratch:
Check things again with the scope.
Ah!... funny you should say that, I've just plonked it on the floor next to the rack! :)
This is what the B+ looks like on the scope.
X = 5ms/Div
Y = 5mV/Div
Phono stage and TT are connected and everything has been warmed up for at least half an hour.
http://img398.imageshack.us/img398/8198/p9131971hq6.jpg
By Shian7 (http://profile.imageshack.us/user/Shian7) at 2008-09-13
NB. I have no idea when this scope was last calibrated. Or indeed cleaned!!!
Have you checked the CCS with some 9v batteries and a 1k resistor?
Are all the LED and VR tubes lit?
Have you checked the CCS with some 9v batteries and a 1k resistor?
Yes but with a 100R resistor (I'm sure that's what you said originally), I needed about 12 of them to get both LED's to light up. 17.5mA
Are all the LED and VR tubes lit? Yes. It even plays music. :)
Yes but with a 100R resistor (I'm sure that's what you said originally), I needed about 12 of them to get both LED's to light up. 17.5mA
1k or 100 will do as the load. The reason you need that much to get the leds lit is the dropper between the leds and ground. To run with 27 or so volts (100R would be a better load with 18v) you need to reduce the size of that to give you the 3-5ma or so to get the LED's into their constant voltage region.
But once you get the voltage high enough, you need to ensure the same current is passed as the voltage increases. And using a 10k load you should get the same current on the end of the 300v supply. The only reason you increase the load is to stay within the safe operating area of the larger transistor in the CCS.
Hmmm... Something isn't quite right. :scratch:
Either I've got one of the transistors arse about face (again) or I'm testing it incorrectly.
Should I be testing it like this?
http://img118.imageshack.us/img118/5777/c4svi3.gif
By shian7 (http://profile.imageshack.us/user/shian7) at 2008-09-13
Yes, though the load needs to be such that the expected current will produce a volatge less that the source voltage, but not too much less that the transistor is overloaded. And the 180K resistor needs to be changed to allow the two LEDS to pass a couple of ma from the appled voltage.
So, for 18v, maybe a 500R load, and a 10k in series with the leds/
Ahhh.... Then I think it's working OK. :)
No time to elaborate ATM as we're expecting company at any moment.
More tomorrow....
Primalsea
13-09-2008, 19:40
SOOOOOOoooooo,
http://i44.photobucket.com/albums/f23/primalsea/DSC00140.jpg
http://i44.photobucket.com/albums/f23/primalsea/DSC00139.jpg
I have tarted up the enclosure with a wood facia and made use of the 2 chrome knobs from my WD88. Its looking like I may not have enough room inside so may have to have an outboard psu.
The low voltage psu is done for now and just needs a test. The input stage is also almost done as well. Just a few more things to do and then I can start testing it.
This is what the B+ looks like on the scope.
X = 5ms/Div
Y = 5mV/Div
Phono stage and TT are connected and everything has been warmed up for at least half an hour.
So it's around 1 mVpp and 100 Hz. If you're using LM317/337 try using bigger C's on the control pin, 50 - 100 µF might help.
Led's can generate a field of noise. So be careful when using those and keep them far from the input.
John
What I see with that scope trace is the regulation dropping out on the lower part of the sine wave, its not just the input ripple. If you have any upstream resistor, I would try removing them for a test, to see if more voltage before the CCS makes a difference. I would also compare the trace before and after the CCS to see if the lower part of the ripple is dropping to low.
What voltage before the resisor did you have before replacing with the CCS?
Primalsea
16-09-2008, 20:25
I was short shipped the transformer for the input section so cant go much further until it comes.
Meantime I've just written a handy Excel spreedsheet that takes the arseache out of calculating the values for a Mu Follower. I wanted to post it for anyone to use but it seems you can't upload Excel sheets. Is there some other way of posting it, a downloads section maybe?????
Primalsea
19-09-2008, 11:15
OMG I've been truly overwhelmed by the response for my spreadsheet!!
Anyway, I've just spotted something that may have caused some confusion. In the initial circuit I put 6N1P as the valves to use. I had read somewhere (probably from a careless Ebay seller) that 6N1P was a direct interchange for a 6DJ8. Now that I have looked at the curves for both types I see that they are not interchangable without modification.
I wonder how many people have stuck the wrong valve in and watch it go BANG???
Primalsea
22-09-2008, 18:30
I've finally made the input circuit and its PSU and begun testing it.....
Could I get it too work, could I F**K. Scrathed my head for 5 minutes as to why and then realised I hadn't turned the thing on!!
:doh::doh::doh::doh::doh::doh::doh::doh::doh:
Begun again and just got all sorts of weird shapes on the scope. Scrathed my head for 5 minutes as to why and then realised I had connect the scope ground to the wrong place!!
:doh::doh::doh::doh::doh::doh::doh::doh::doh:
Begun again and still got all sorts of weird shapes on the scope. Scrathed my head for 5 minutes as to why and then realised I had'nt connected all of the ground wires up.
:doh::doh::doh::doh::doh::doh::doh::doh::doh:
At that point I realised it was a good idea to leave it for another day as I obviously wasn't in th eright headspace for it.
:(:(:(:(:(:(:(
Primalsea
05-10-2008, 21:27
Update...beep beep ber beep beep beep beep!!
The balanced input circut seemed to be unreliable at high gain so I had to change in to single ended input. I seems to work OK but I have to admit that background noise may be an issue. However the amount of noise changes drastically depending on where the scope is connected to ground.
Looks like one of the mose important things is the ground plain when looking into low noise.
Primalsea
26-10-2008, 12:25
I managed to get 2 small switchmode PSU's from Ebay for £9 total to use as the heater supplies. I cracked them out of the cases and glued them to some protoboard. They have a few settings one of which is 6.5VDC and can supply upto 1200ma, they only have to supply 360ma each.
They do have some noise which can be seen on the scope pic below. With luck this will not be coupled to the output but if it is a problem I can increase the voltage to 7.5V and put in some extra filtering.
http://i44.photobucket.com/albums/f23/primalsea/DSC00176.jpg
http://i44.photobucket.com/albums/f23/primalsea/DSC00174.jpg
This is getting close to a working prototype now. All I need to buy is the transformers and components for the HT PSU. I will probably use 2 9V transformers back to back as sourcing a single HT transformer has become a problem. These only have to supply 8ma at idle so wont have to be very large.
I will probably use 2 9V transformers back to back as sourcing a single HT transformer has become a problem. These only have to supply 8ma at idle so wont have to be very large.
Hmmm.... If you're going to do that, would it not be better to 'tap off' some of the 9V and use it for the heater supplies? That way you could add whatever filtering/smoothing you want instead of mucking about with those noisy SMPS bits?
Just a thought... ;)
Primalsea
26-10-2008, 19:27
That was my original idea but as I have a mu follower one heater has to be at 0v and the other flaoting at 220v (something that I didnt spot until yesterday). So I would need an extra transformer and by the time all the other bits are included space would be at a high premium. So I thought I would give the SMPS a go and see whats happens. I can always remove them if it goes tits up.
We have very bad experiences with SMPS supplies. In various (High End?) audio chains such a supply induced some noise via the mains connections. The result was that whatever the source, be it CD or record, the stereo image was somewhat blurred, less "open". It took some time to find out it was that supply (in a DVD-player) causing the trouble. Mind you that player was not connected in audio way with the audio set.
We have been using E88CC, 6N1P and 6922 tubes in various follower configurations for over 15 years. The supply voltage always was around 200 Volt so the follower output was at 100 Volt. The heater is grounded (when rectified always the minus to ground!) and we never had a problem with the cathode-heater voltage difference. The tubes, at around 12-13 mA, will last for at least 5 years and mostly over 10 years without degradation of the sound quality.
There could be a problem though using a µ-follower or Lofton-White. The distortion increases if the output of the amplified audio is going over 5 Volt eff.
John
New bits!
http://img508.imageshack.us/img508/5937/pc092308mt0.jpg
By shian7 (http://profile.imageshack.us/user/shian7), shot with u30D,S410D,u410D (http://profile.imageshack.us/camerabuy.php?model=u30D%2CS410D%2Cu410D&make=OLYMPUS+CORPORATION) at 2008-12-09
Can you tell what it is yet?
<Clue> not all the bits are for the phono stage. :)
Ali Tait
09-12-2008, 19:28
Stepped attenuator! and looks like some teflon tubing.Whadda I win? :)
'fraid there's no prize for 2 out of 6! :(
i_should_coco
09-12-2008, 19:39
Hmmm... passive pre-amp?
Nah!..... the little rotary switch/knob will have those matched pairs of Takman resistors stuck on to provide switchable cartridge loading for the phono.
The switched attenuator (dead cheap) is going into the WD88, which is getting rewired with that solid silver wire which will be stuffed down the Teflon tubing.
:)
i_should_coco
09-12-2008, 19:57
Nah!..... the little rotary switch/knob will have those matched pairs of Takman resistors stuck on to provide switchable cartridge loading for the phono.
The switched attenuator (dead cheap) is going into the WD88, which is getting rewired with that solid silver wire which will be stuffed down the Teflon tubing.
:)
Obvious now! :doh:
Ali Tait
09-12-2008, 20:32
:) Well I kinda thought the other stuff was self evident-roll of solder etc.
Arrrrrrrh, (in a pirate stylee) t'is no solder... t'is silver wire! ;)
Pete,
I hear your little project trounced a few well know items at the weekend mate nice work :-)
You be ready for production run soon
Ali Tait
09-12-2008, 22:07
That's a lot of wire! How much is there and where did you get it from? How much if you don't mind me asking?
Hi Ali,
It's 10M of 0.6mm (dia) 4N solid silver from wires.co.uk, it cost £16.28 + VAT.
http://wires.co.uk/acatalog/si_bare.html
The Teflon tubing is from RS.
Cheers...
Powered by vBulletin® Version 4.2.3 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.