Opa1612 in my favourite in this moment. Lm49720 it's too good, but i think opa1612 is better. I taste opa 2604) interested sound with him, but need a time to give results
Last edited:
Good. I would say that if OPA1612 doesn't sound right, it just means there is some more work to do. It means there are still problems with the dac. Of course, there may still be some problems with the dac even if OPA1612 does sound good 
LME49720 is also a very good part, but it can be more sensitive to some RF. Works great for things where there is not too much RF getting into it. DECT wireless phone base stations within 10 feet or so can cause FFT spurs every 100Hz. The base stations transmit up around 2GHz. A shielded case will fix the problem though.
LME49720 is also a very good part, but it can be more sensitive to some RF. Works great for things where there is not too much RF getting into it. DECT wireless phone base stations within 10 feet or so can cause FFT spurs every 100Hz. The base stations transmit up around 2GHz. A shielded case will fix the problem though.
Last edited:
IT'S GARBAGE!!!!!!!!!!!!!!!
IT'S GARBAGE!!!!!!!!!!!!!!!
Dat's cuz you aints 'eard 'em all, Guv. It's a JELLYBEAN part meant for bottom-budget devices.
IT'S GARBAGE!!!!!!!!!!!!!!!
yes i is more difficult to work with, but it may be worth the hassle to try ...at minimum as a simple regulator duty. By experience, it is quite difficult to find a top notch (diy and even commercial i would speculate) 3.3v regulator for audio and this may be a good candidate. Output impedance is really important for digital regulation and it is really for this chip.
as for distortion, it is a concern, but even if THD rises, the opamp is very low distortion to begin with. In any case, i assume opamps are not meant to drive low impedance headphones, there are many top notch diy designs out there for that purpuse, like dcg3, beta22, firstwatt f4, some of member EUVL's designs, etc. But for higher impedance cans, low power shouldn't be a big issue.
like i said why not try?
Last edited:
sorry, i have no experience with opa2604.
how is opa1612 as a simple regulator? how would you compare it to other regs you used /are using?
why was he raving? what did you want to say? I don't understand nonsense. can you have a temperature? Measure it, friend, go to the doctor. Take some pills, maybe it will feel better
Guys,
I'm just trying to be helpful, not tell you what to do. However, I am a little concerned that so much mental focus on opamps is actually misplaced. For example, the only loads for which an opamp regulator is suggested is for AVCC_L and AVCC_R. Did anyone bother to check what the actual load current is going to be? I'll save some time and just tell folks that for ES9038Q2M, each AVCC load pulls about 5mA.
Instead some other questions that can help improve performance should be asked: What bypass caps should be used for the opamps? What layout is best between the dac chip and the opamps? Where will digital and analog ground currents be flowing and what impact might they have on the audio output? What voltage regulators are good to use for the opamps? And so on. Someone who is very thorough might experiment with all of those things to find a good optimum for each.
In other words, please don't over-focus on things that are often given too much focus of attention in this forum. Those aren't the only important things that benefit from careful attention, and from the running of careful experiments. Eventually, many of the things that are not thought about will limit the dac's audible performance.
Regarding OPA1612, it is well vetted. Its not going to be the limiting factor, at least not until a lot of much bigger problems are fixed. If anyone does still feels a need to experiment with AVCC regulator opamps, I would suggest to try AD797. There are tricks that can be done with it that could possibly come in handy at some later point.
Again, no insult to anyone is intended by any of the above. Just trying to help with getting good results.
I'm just trying to be helpful, not tell you what to do. However, I am a little concerned that so much mental focus on opamps is actually misplaced. For example, the only loads for which an opamp regulator is suggested is for AVCC_L and AVCC_R. Did anyone bother to check what the actual load current is going to be? I'll save some time and just tell folks that for ES9038Q2M, each AVCC load pulls about 5mA.
Instead some other questions that can help improve performance should be asked: What bypass caps should be used for the opamps? What layout is best between the dac chip and the opamps? Where will digital and analog ground currents be flowing and what impact might they have on the audio output? What voltage regulators are good to use for the opamps? And so on. Someone who is very thorough might experiment with all of those things to find a good optimum for each.
In other words, please don't over-focus on things that are often given too much focus of attention in this forum. Those aren't the only important things that benefit from careful attention, and from the running of careful experiments. Eventually, many of the things that are not thought about will limit the dac's audible performance.
Regarding OPA1612, it is well vetted. Its not going to be the limiting factor, at least not until a lot of much bigger problems are fixed. If anyone does still feels a need to experiment with AVCC regulator opamps, I would suggest to try AD797. There are tricks that can be done with it that could possibly come in handy at some later point.
Again, no insult to anyone is intended by any of the above. Just trying to help with getting good results.
Last edited:
One opamp could drive Ian's dual ES9038Q2M AVCC, I think. It might be about 20mA total. Getting up there in terms of current but I think its doable with one OPA1612 output. You might find it sounds better to ground the negative rail of the AVCC opamp, since that's the reference voltage for most opamps. A power supply is often not as stable and noise free as good quality ground can be. You could try it either way though, see if you notice any difference.
I've built up one as per the schematics here with LME49720 so dual rail psu for the opamp and configured for the 2 AVCC as per the Chinese board we are talking about on this thread.
So I can just is this as is with +/- 15v and the 3v3 ref input and just use one output?
Sorry I am pretty green with op amps!
So I can just is this as is with +/- 15v and the 3v3 ref input and just use one output?
Sorry I am pretty green with op amps!
That should work, I think. Safe try it and see what you think. Maybe try to keep RF away from the opamps if possible. There is EMI/RFI that radiates from some of the digital circuitry.
If you wanted to tie the two outputs together, then you would have add resistors in series with each opamp output (after the feedback connection). That's because they will probably have slightly different offsets, and we don't want them fighting each other for exact control of the output voltage.
If you wanted to tie the two outputs together, then you would have add resistors in series with each opamp output (after the feedback connection). That's because they will probably have slightly different offsets, and we don't want them fighting each other for exact control of the output voltage.
Last edited:
SUPERB!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
I may have been SUPERBLY mistaken here ... when I Googled the opa in question, the Google SE spat back an $0.89 "jellybean" part. Which I assume was the opa y'all we're 'ferrin' ta.
It turns out that was not the OPA1612 ... a >$6.00 part!!!!!!!!!!! Whoa Nelly!!!!!!!!! Now how does that 'spensive 'amp figure into a "budget" DIY project? What a 'fusin' project y'all gots goin' on!!!!!!!!!
As far as opamps .... AD825.
I may have been SUPERBLY mistaken here ... when I Googled the opa in question, the Google SE spat back an $0.89 "jellybean" part. Which I assume was the opa y'all we're 'ferrin' ta.
It turns out that was not the OPA1612 ... a >$6.00 part!!!!!!!!!!! Whoa Nelly!!!!!!!!! Now how does that 'spensive 'amp figure into a "budget" DIY project? What a 'fusin' project y'all gots goin' on!!!!!!!!!
As far as opamps .... AD825.
do you imply it may be unstable with digital loads?
not rejecting argument, but for info, the flea regulator by «ray« from netherlands (can't remember his surname) is designed for a clock.
http://www.acoustica.org.uk/flea_manual.pdf
i tried ad797 in that flea amp for a clock in previous design and it sounded natural, but not very dynamic. i liked lm317 better... i supsect maybe open loop outpu Z of the opamp to be on the higher side.
Last edited:
When I was modding these dacs I used LT1763, with the optional NPO noise reduction capacitor. Used it for clocking, VCCA, and DVDD. The clock and VCCA may be considered as being analog RF circuitry, whereas DVDD looks like its more digital. Clock stability also depends on choice of power rail decoupling methods (suggest to avoid ferrites, high quality film bypass caps may sound better than X7R despite what the textbooks say -- have to try it and see). The reason for doing it that way was because the regulators are small and can be built dead-bug style on the PCB solder side ground plane. It keeps inductance between the loads and the regulators reasonably low. In addition, I would to suggest to try loading each LT1763 output with a resistor to ground, perhaps in the range of 47R to 68R. Often trying that can result in a step improvement in sound quality. Its an old trick that is often forgotten, but it can still work and may be worth trying.
As far as the opamp circuit goes, one could try it. I doubt it would turn out to work as successfully as the method described above.
Three of the regulators can be seen in the pic below.
As far as the opamp circuit goes, one could try it. I doubt it would turn out to work as successfully as the method described above.
Three of the regulators can be seen in the pic below.
Attachments
Last edited:
OMG, yet another massacred board
.Here we go again, power supply film cap bypass sounds better than X7R cap bypass, I am sure you can hear the difference, isn't it?
And where can I read about this old trick of loading the regulator output and its relationship to "a step improvement in sound quality".
Hearing the difference between X7R and film or C0G bypass caps?
Well I can.
But more importantly, Gilles can, my wife can, my mother can, my son can. All the latter don't even know what we are talking about and couldn't care less. All were used as guinea pigs with 2 identical amps, as the winner was to be duplicated for bi amping purpose. Caps were added step by step, findings were all similar in the comments.
As an ex race engineer I won't get into the technical debate: I have seen so many things that shound't have worked per simulation or classical theory around active chassis and cam profile that I could fill a book.
Do I know what I hear? Am I right with my thinking? I don't know. But I don't care because it sounds better to me, clearly, and I am the one who pays, so...
Will I convince you? NO.
But German spell : Probieren geht über studieren. it cost a few pennies, give it a try and discuss the result with yourself. Take a unit you know by hart since ages. Try any decent OPA ouput stage that is already decoupled by the book and add just a 0.1uF C0G or fim cap. Perhaps between the legs of the already existing X7R 0.1uF cap.
Easy reversal and no real risk
' nough said, I am out, no time for debates and not the topic anyway. Plus after returning the D50s, the M200 (which all sounded so different despite differences being supposedly so inaudible and yet so unsatisfying at the end despite the massive catch up once a proper PS came to play) for my friend's system we decided to stick to SU-9 ... and you just convinced me that their engineers might be just like you so that I should start tweaking the 0805 bypass caps of its 1612 output amps and a few other bits just to see if we enjoy the placebo even more
Thanks for that, order just placed
Well I can.
But more importantly, Gilles can, my wife can, my mother can, my son can. All the latter don't even know what we are talking about and couldn't care less. All were used as guinea pigs with 2 identical amps, as the winner was to be duplicated for bi amping purpose. Caps were added step by step, findings were all similar in the comments.
As an ex race engineer I won't get into the technical debate: I have seen so many things that shound't have worked per simulation or classical theory around active chassis and cam profile that I could fill a book.
Do I know what I hear? Am I right with my thinking? I don't know. But I don't care because it sounds better to me, clearly, and I am the one who pays, so...
Will I convince you? NO.
But German spell : Probieren geht über studieren. it cost a few pennies, give it a try and discuss the result with yourself. Take a unit you know by hart since ages. Try any decent OPA ouput stage that is already decoupled by the book and add just a 0.1uF C0G or fim cap. Perhaps between the legs of the already existing X7R 0.1uF cap
.Easy reversal and no real risk
' nough said, I am out, no time for debates and not the topic anyway. Plus after returning the D50s, the M200 (which all sounded so different despite differences being supposedly so inaudible and yet so unsatisfying at the end despite the massive catch up once a proper PS came to play) for my friend's system we decided to stick to SU-9 ... and you just convinced me that their engineers might be just like you so that I should start tweaking the 0805 bypass caps of its 1612 output amps and a few other bits just to see if we enjoy the placebo even more
Thanks for that, order just placed
Last edited: