Hi everyone,
I've built quite a few DACs with the usual opamp methods.
My best DAC so far is based on PCM1792 and LME49710, it does have a respectable sound 🙂
Just recently i found out about the Zen V/I method and it looks quite interesting.
Is there an implementation compatible with PCM1792?
I'm sure that the answer to this is covered somewhere of these 78 pages or the other 100 pages related so i apologise is advance but don't have the time to go through them.
My concern with the PCM1792 is that it only sinks current and is idling on -6.2mA, also how about using the diff output to cancel distortion?
I'm very interested to listen to this method... that's why i'm looking for an implementation that has already been tested...
Also i noticed a lot of discussion on batteries... Considering that my job for 3 years used to be designing battery management systems i might be able to help you with something...
Thanks.
I've built quite a few DACs with the usual opamp methods.
My best DAC so far is based on PCM1792 and LME49710, it does have a respectable sound 🙂
Just recently i found out about the Zen V/I method and it looks quite interesting.
Is there an implementation compatible with PCM1792?
I'm sure that the answer to this is covered somewhere of these 78 pages or the other 100 pages related so i apologise is advance but don't have the time to go through them.
My concern with the PCM1792 is that it only sinks current and is idling on -6.2mA, also how about using the diff output to cancel distortion?
I'm very interested to listen to this method... that's why i'm looking for an implementation that has already been tested...
Also i noticed a lot of discussion on batteries... Considering that my job for 3 years used to be designing battery management systems i might be able to help you with something...
Thanks.
Attachments
I've been using PCM1794 which is similar I think.
SEN is very simple and there are simple explanations about how to deal with the bias current without using servos in the thread above.
You should probably try those first for the best sound.
My room temperature varies by about 25C throughout the year and my jfets are not matched perfectly so I have been trying to come up with a circuit with servos to deal with the bias current and unmatched jfets.
I posted a circuit in #674 which works for the bias current but the servo in that circuit introduces some high order distortion and the jfets need to be well matched.
I posted a later circuit in #749 which DOES NOT WORK (the filter is noisy and the two servos are unstable) but I have a 2 servo circuit similar to #749 working at home which is stable and quiet with measured 2nd and 3rd harmonic about -95dB for an unbalanced half.
This latest circuit is currently on a breadboard and is picking up some RFI. I want to do a soldered balanced version this weekend.
If it sounds good, I'll post the schematic soon.
Here is an example of CEN used with a current sinking DAC.
http://www.diyaudio.com/forums/digi...minimalistic-iv-converter-12.html#post2935106
You just need to add a current source upfront of the standard CEN.
But as Calvin said before, I would start with a single JFET CCS first.
As to distortion cancellation, simply use balanced output.
If you have to add a balanced to SE converter (opamp) or otherwise, you add both complexity and distortion.
I would then just use the +ve phase and leave pin 3 of the XLR open.
Patrick
http://www.diyaudio.com/forums/digi...minimalistic-iv-converter-12.html#post2935106
You just need to add a current source upfront of the standard CEN.
But as Calvin said before, I would start with a single JFET CCS first.
As to distortion cancellation, simply use balanced output.
If you have to add a balanced to SE converter (opamp) or otherwise, you add both complexity and distortion.
I would then just use the +ve phase and leave pin 3 of the XLR open.
Patrick
Thanks guys i'll look into these and try them out,
Just as an indication, how much difference (compared to Op-amps) do you hear with these circuits?
Cheers,
Panagiotis
Just as an indication, how much difference (compared to Op-amps) do you hear with these circuits?
Cheers,
Panagiotis
Couple of questions,
Why so many parallel JFETs?
What caps are you guys using for the output? polyprop @ 200uF it's not easy to find 🙂
I've used the AMPOHM caps for ac coupling in other amps before but 30uF is about the max value you get. There are some Vishay MKPs at 200u but are 50quid a pop...
Why so many parallel JFETs?
What caps are you guys using for the output? polyprop @ 200uF it's not easy to find 🙂
I've used the AMPOHM caps for ac coupling in other amps before but 30uF is about the max value you get. There are some Vishay MKPs at 200u but are 50quid a pop...
Go to Linear Audio Magazine website and download the article there for free.
It will give you a lot of information that allows you to build a very good sounding system.
Of course I shall not stop you from experimenting with exotic parts.
But that would be the last 2% at most.
And before that you should figure out how the other 98% appeal to you.
Patrick
It will give you a lot of information that allows you to build a very good sounding system.
Of course I shall not stop you from experimenting with exotic parts.
But that would be the last 2% at most.
And before that you should figure out how the other 98% appeal to you.
Patrick
The PCM1794 likes to drive a virtual ground. If you parallel them up you get a lower input impedence and so closer to a virtual ground.
I'm using Rubycon electrolytics. They are very cheap.
Patrick thanks very much for the tip.
I read your article and i understand much more of this now.
It made me even more curious to try it out.
Also now that i got a change to study this a bit better i see your point about the current source to bias the output of the DAC.
I'll try a couple of things on simulation first 😉
You mention quite a few times about the spice sims you did, what jfet sim model did you use? do you have a model for the 2sk170/2sj74?
Cheers,
Panagiotis
I read your article and i understand much more of this now.
It made me even more curious to try it out.
Also now that i got a change to study this a bit better i see your point about the current source to bias the output of the DAC.
I'll try a couple of things on simulation first 😉
You mention quite a few times about the spice sims you did, what jfet sim model did you use? do you have a model for the 2sk170/2sj74?
Cheers,
Panagiotis
Search the forum and you will find all the Spice models you need for the JFETs.
Ask those who build one by PM how they like the sound, and how it compares to the commercial IV kits / modules.
In the time you spend to sim this, you would have finished building the real thing.
😉
Patrick
Ask those who build one by PM how they like the sound, and how it compares to the commercial IV kits / modules.
In the time you spend to sim this, you would have finished building the real thing.
😉
Patrick
Last edited:
Patrick,
This may seem to be a bit of a "wackie" idea but if I could manage to keep the input impedance of a Lightspeed (Passive Vol Control) to 2.7kR (Uriah's 'LighterNote' version can do this okay), would this be a reasonable proposition for the Riv as the current would still generate the o/p volts but via an L-pad instead of a singe resistor?
It seems too simple to be viable.
This may seem to be a bit of a "wackie" idea but if I could manage to keep the input impedance of a Lightspeed (Passive Vol Control) to 2.7kR (Uriah's 'LighterNote' version can do this okay), would this be a reasonable proposition for the Riv as the current would still generate the o/p volts but via an L-pad instead of a singe resistor?
It seems too simple to be viable.
If your DAC has I_out = 1mA, there is no reason you cannot try.
The problem is that LDRs do not like current. They heat and then distortion goes up.
And what ever the impedance, they still see the 1mA max.
I guess you know how to calculate dissipation then.
Patrick
The problem is that LDRs do not like current. They heat and then distortion goes up.
And what ever the impedance, they still see the 1mA max.
I guess you know how to calculate dissipation then.
Patrick
Thanks Patrick,
Ah, that might be a problem as the 1541A dac outputs max of 4mA, and even tho each of the Siliconex devices that make up a Lightspeed are 'specced' to be able to absorb 50mW, it probably becomes quite 'non-ideal' at anything like that.
Possibly with the vol control's input impedance down to less than 1kR, it would reduce the 'guestimated power' consumption to 16mW max, and if one of your neat little heatsink fitted, it could be viable.
As I'll be using an Aikido gain stage to follow that has a minimum gain of about 18dB with the 6SN7s, too much voltage gain anyway to drive the amps.
I will give this a try with some unmatched devices (with Uriah's control boards) and see if it functions at all well - very interesting, hmm.
Ah, that might be a problem as the 1541A dac outputs max of 4mA, and even tho each of the Siliconex devices that make up a Lightspeed are 'specced' to be able to absorb 50mW, it probably becomes quite 'non-ideal' at anything like that.
Possibly with the vol control's input impedance down to less than 1kR, it would reduce the 'guestimated power' consumption to 16mW max, and if one of your neat little heatsink fitted, it could be viable.
As I'll be using an Aikido gain stage to follow that has a minimum gain of about 18dB with the 6SN7s, too much voltage gain anyway to drive the amps.
I will give this a try with some unmatched devices (with Uriah's control boards) and see if it functions at all well - very interesting, hmm.
You can set the LDR attenuator total resistance to say >> 10k, and then parallel this with a 680R R_iv for the 1541.
Should work fine. But now you need a good 680R of course which defines the sound.
Patrick
Should work fine. But now you need a good 680R of course which defines the sound.
Patrick
Yeah, pretty happy with the sound of the NeoOhms 8E16D's (Rhopoint) so no problems there.
I meant to ask awhile ago, which Conductive Polymer caps have you had success with and where do you get them from?
Thanks.
I meant to ask awhile ago, which Conductive Polymer caps have you had success with and where do you get them from?
Thanks.