PCM56K iv resistor value?

[Zoran]

Gabdx1
If You are planing to play with the tubes at high voltages
please stay out of that idea
joke sorry...
*
try to realise simple circuit without mistakes...
and maybe You are damage the dac?
if You have hard distorsion than it is a case
cheers
*

 

[aparatusonitus]

Was the 100ohm in series that I instaled acting as a voltage divider with the input impedance of the Transimpedance I/V amp?

I don't think you should put any resistor in series with the dac output...the dac will see a 100R+Re impedance, instead (theoretically) zero impedance from active I/V. What I would do is to half the values for R19/20 'cos that is the point where actual current to voltage conversion is taking place. Adjust values of C2/3 appropriately, as well as values of R17/18 (just a bit higher, start with 210R and up) until you get the lowest output DC offset without servo opamp installed.

 

[Gabdx1]

Hi Zoran, I am doing ok with high voltages and tubes, but thanks for warning people.

I think one day ill put up a website to denounce the gimmicks of audio once and for all, when some people advertise products that don't work or solutions for deaf people, it's not good for the society. This is just bad days for me and I am revolted against capitalism driven industry.

For good vintage dacs ill look elsewhere: UltraAnalog D20400A

There is even all the schematics from audio research if someone want to build it.

 

[champ04]

Well this is a very interesting thread. And a bit confusing as well.

I have an X64.4 just like georgehifi has. Lately I've been looking at fiddling with replacing the opamps used for l/V and the output buffer they feed.
I started by thinking about using a Zapfilter, and I still am. But I'm aslo curious about the passive idea after having read this thread. I'd also like to consider the Counterpoint l/V........
(I'm also thinking, in the deep recesses of my mind, of getting really crazy and figuring out how to feed the dacs into the output section of a B.A.T. Vk-D5se. Yikes!)

Anyway......
George,
First, how do you remove the aluminum covers that hide the dacs on the X64.4?
And correct me if I am wrong, but based on what I've read here, it seems like you are only feeding one phase of your dacs into the Zapfilter. Why is this? Having parced out the Sledgehammer I am fairly certain that the dacs have fully balanced outputs.

Also, (to anyone) I'm a little confused regarding the Zapfilter MkII. Their website mentions that it is a single ended design but that it is capable of fully balanced output. I'm assuming its doing some inverting on its own. Can it be safely fed both phases of a balanced dac?

The passive idea is interesting but, like george, I've kind of grown to appreciate the powerful output of the original Sledgehammer. Then again, I like simplicity..........

The most important factors for me are zero feedback and discrete (ie. no opamps)

 

[georgehifi]

I don't think you should put any resistor in series with the dac output...the dac will see a 100R+Re impedance, instead (theoretically) zero impedance from active I/V. What I would do is to half the values for R19/20 'cos that is the point where actual current to voltage conversion is taking place. Adjust values of C2/3 appropriately, as well as values of R17/18 (just a bit higher, start with 210R and up) until you get the lowest output DC offset without servo opamp installed.

Correct me if I'm wrong but the way I see it the dacs themselves are already 1200ohm output impedance, another 100ohms is not going to hurt.

Anyway......
George,
First, how do you remove the aluminum covers that hide the dacs on the X64.4?
And correct me if I am wrong, but based on what I've read here, it seems like you are only feeding one phase of your dacs into the Zapfilter. Why is this? Having parced out the Sledgehammer I am fairly certain that the dacs have fully balanced outputs.

Also, (to anyone) I'm a little confused regarding the Zapfilter MkII. Their website mentions that it is a single ended design but that it is capable of fully balanced output. I'm assuming its doing some inverting on its own. Can it be safely fed both phases of a balanced dac?

The passive idea is interesting but, like george, I've kind of grown to appreciate the powerful output of the original Sledgehammer. Then again, I like simplicity..........

The most important factors for me are zero feedback and discrete (ie. no opamps)

1 The whole board has to come out and the cover screws removed from the other side.

2 The dacs themselves are single ended not balanced output they (Wadia) make a pseudo balanced output after the dac, not real balanced as the dac is not. (see sketch attached this is what's under the covers minus a couple of caps here and there) .

3 The Zap will give balanced or single ended output even when only fed by the PCM56p single ended output. again it's pseudo

Cheers George

 

[georgehifi]

and the dac board

 

[champ04]

Thanks George!

Wow, I think I'm actually a little disappointed now. All this time I though the balancing was done in the digital domain.
I guess I should have known better considering how many dacs it takes to achieve the 64X sampling.

Again, thanks!!!

It looks like the Zapfilter is the way to go for now.

 

[aparatusonitus]

Correct me if I'm wrong but the way I see it the dacs themselves are already 1200ohm output impedance, another 100ohms is not going to hurt.

The dacs output is AC current source/sink and by definition current source/sink is high impedance, but the important thing is what dacs output see in front of it...in your case that is 100R+Re while there should be only Re.

 

[georgehifi]

The dacs output is AC current source/sink and by definition current source/sink is high impedance, but the important thing is what dacs output see in front of it...in your case that is 100R+Re while there should be only Re.

People like Berhard are using 300ohms odd for passive I/V resistor loading of the dacs, if the zap is 25ohm then now with the 100ohm that I put in in series it sees 125ohm, still way under (better) than what the passive I/V resistor brigade are saying, and it sounds magnificent.
Cheers George

 

[champ04]

I took the top off the dacs. Mine is a slightly different configuration than george having two extra trims.
The dacs themselves are Wadia part # 400022.

The resolution on the pic george posted above is not high enough for me to determine what they say.........
Do they also say "Wadia" on them. How were you able to figure out they were the PCM56K?

opamps on mine are OPA606KP.

Thanks.

 

[georgehifi]

I took the top off the dacs. Mine is a slightly different configuration than george having two extra trims.
The dacs themselves are Wadia part # 400022.

The resolution on the pic george posted above is not high enough for me to determine what they say.........
Do they also say "Wadia" on them. How were you able to figure out they were the PCM56K?

opamps on mine are OPA606KP.

Thanks.

Same same, the earier ones had PCM56P K , Wadia started to put their own name on the later ones, and they could even be PCM61P K as they did use them later, 18bit vs 20bit identical in every other way.
As for the opa606 you don't use these if you are wiring in a Zap, cut the track shown (to get rid of the OPA606's) and take the signal from pin 13 shown (which is all 4 dacs) use a good quality thin shielded mic cable take the earth from the point shown and then the signal goes to the + input of the zap via a 100ohm resistor and the shield to ground on the input of the zap, and Bobs your uncle.

 

[georgehifi]

Close up of mine before mods

 

[champ04]

Mine wont be quite as straight forward.
The trace going to the opamps must be on the underneath side of the board now. And the pin 13 trace may be a little more difficult to find. But I dont think these things will be a deal breaker.

Thanks again!

 

[champ04]

Incidentally, in reading the literature of Wadia's newest best DAC (922) it would seem they have done pretty much what george has to his. That being using a fully discrete, zero feedback, class A l/V output stage.

They also seem to have actually implemented what I had thought they were doing all along........
Inverting the signal in the digital domain and achieving true balanced operation by using a total of 8 dacs per channel.
Now THATS a serious DAC!

 

[Bernhard]

Incidentally, in reading the literature of Wadia's newest best DAC (922) it would seem they have done pretty much what george has to his. That being using a fully discrete, zero feedback, class A l/V output stage.

They also seem to have actually implemented what I had thought they were doing all along........
Inverting the signal in the digital domain and achieving true balanced operation by using a total of 8 dacs per channel.
Now THATS a serious DAC!

I have 16 DACs (2x8 true balanced) per channel and
no active I/V or output stage.

 

[champ04]

I have 16 DACs (2x8 true balanced) per channel and
no active I/V or output stage.

Thats very interesting.
Have you posted more information about this somewhere? I am very curious.

Are the 8 dacs per phase summed for the purpose of increasing the current output?

 

[Bernhard]

increasing current = decreasing impedance

 

[champ04]

increasing current = decreasing impedance

Which chip are you using?
Any pictures?

Thanks.

 

[Bernhard]

I knew somebody would ask...
Here is a picture taken with my phone , not finished yet, 3 pcbs have to be replaced and one with yellow & black caps has to be added, plus interconnects & box for the PSU & box for the DAC, next winter...

It is playing though.

I have posted a picture of the TVC which is part of the DAC elsewhwere.

DACs are PCM56.

 

[georgehifi]

Mine wont be quite as straight forward.
The trace going to the opamps must be on the underneath side of the board now. And the pin 13 trace may be a little more difficult to find. But I dont think these things will be a deal breaker.

Thanks again!

Lift or snip the leg on pin 2 of the first of the OPA606 coming from any of the dacs pin 13, that'll get them out of the circuit. I would say it's the one on the right of your pic .
Cheers George