Re: What numbers????
The question was:
TI/BB advices to use the DAC with a given opamp to be able to obtain the performance numbers they give in the datasheet. Do you mean that this can also be done with a discrete IV?
I think that they precised to use this particular opamp, because they got rid of people complaining because they didn't succeed to obtain the given numbers, with a different circuit
Jocko Homo said:
The question was:
TI/BB advices to use the DAC with a given opamp to be able to obtain the performance numbers they give in the datasheet. Do you mean that this can also be done with a discrete IV?
I think that they precised to use this particular opamp, because they got rid of people complaining because they didn't succeed to obtain the given numbers, with a different circuit
and i think they specified that opamp because its something like standart. you see something like that in most dac datasheets. What leads you to the suspicion IC must be better than discrete?
Also i think those specs and details are not really of interest. If we look application notes for precision DACs with less than 24 Bits, lets say 16 or 18 - and not into audio - we see much more afford in the link between digital analog circuit, like isolation circuits. This makes me suspect its not of interest if we choose a 16 or 24 bit resolution DAC, and nit if we do discrete or IC i/v conversion, but how much we care for isolation of digital/analog, and power supply quality. And is suspect to measure noise of a power supply is something most of the people posting wisdom on this board can not really do as their equipment will not show noise if frequency is only high enough.
Also i think those specs and details are not really of interest. If we look application notes for precision DACs with less than 24 Bits, lets say 16 or 18 - and not into audio - we see much more afford in the link between digital analog circuit, like isolation circuits. This makes me suspect its not of interest if we choose a 16 or 24 bit resolution DAC, and nit if we do discrete or IC i/v conversion, but how much we care for isolation of digital/analog, and power supply quality. And is suspect to measure noise of a power supply is something most of the people posting wisdom on this board can not really do as their equipment will not show noise if frequency is only high enough.
till said:
Perhaps someone might point out where I'm wrong here, because this variation seems rather obvious to me.
The current out from a DAC is sent to the opamp negative input, with a resistor from the output closing the loop. That means the opamp output must settle quickly as it's a series of steps which must then be anti-alias filtered.
Has anyone thought of adding a parallel capacitor to the feedback resistor of the IV converter to provide the first stage of filtering? It seems to me this would relax the settling requirements of the IV stage.
Am I missing something here?
Francois.
Yes, post #20.
As well as the fact the datasheets often, at least as often as they do not, show such capacitor in the typical application circuits. For example those of TDA1541A and PCM1792 do that.
For certain solution of the problems with settling requirements you need to low pass the signal before it enters the opamp.
As well as the fact the datasheets often, at least as often as they do not, show such capacitor in the typical application circuits. For example those of TDA1541A and PCM1792 do that.
For certain solution of the problems with settling requirements you need to low pass the signal before it enters the opamp.
Bernhard said:
If it measures better it'll sound better, isn't it ?
Almost any modern 1 bit DAC chip outperforms TDA15xx chips by a large margin. The point is that they often sound absolutely worse despite their technical superiority. They also are more sensitive to jitter because of the used technique. Which creates a dillema by itself.
Old, obsolete, more distortion ( at -60 dB ) & good sounding versus modern, very low distortion & bad sounding.
How does the Technics compare to the Philips ( measured with the best instruments ) ? Had several PCM56/58 machines but even modified they could not keep my attention. PCM63 is another story, measures well and sounds OK. I made most DAC's with PCM63...
It's just a counter remark for Bernhard's when-it-measures-better-it-is-better filosofy which has some more angles to view on. Things are not so simple as they seem.
It should be both IMO; measuring well and sounding well. Never heard a PCM56/58 machine that could compete with a TDA1541A machine. I reread some reviews of cdplayers of that time and there were more people thinking the same about this. When PCM63 was introduced things changed obviously. Most modern 1 bit chips can compete technically with TDA15xx easily but most don't sound that well. Howcome ?
I've modified quite some digital audio gear and ( amonst other things ) one thing became very apparent. The influence of a low jitter clock is absolutely highest on 1 bit systems which gives me the impression that jitter is very dominating in the overall experience of listening cdplayers/digital audio. As seen in the measurements of Bernhards implementation of a better clock the better clock did not make a significant difference in harmonic distortion measurements. Please check post #41 in this thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=33430&pagenumber=3
Don't come up with the crap that we like it when gear distorts, we don't.
It should be both IMO; measuring well and sounding well. Never heard a PCM56/58 machine that could compete with a TDA1541A machine. I reread some reviews of cdplayers of that time and there were more people thinking the same about this. When PCM63 was introduced things changed obviously. Most modern 1 bit chips can compete technically with TDA15xx easily but most don't sound that well. Howcome ?
I've modified quite some digital audio gear and ( amonst other things ) one thing became very apparent. The influence of a low jitter clock is absolutely highest on 1 bit systems which gives me the impression that jitter is very dominating in the overall experience of listening cdplayers/digital audio. As seen in the measurements of Bernhards implementation of a better clock the better clock did not make a significant difference in harmonic distortion measurements. Please check post #41 in this thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=33430&pagenumber=3
Don't come up with the crap that we like it when gear distorts, we don't.
i never heard from bernhard what measures good must sound good.
but every day someone posts some wisdom like:" whatever you do, a nos 1543 will sound better" no matter what the question was.
Those NOS 20 year old DAC prophets seem to be unable to accept someone asks question and wants answers on topic, or simply to measure and test different circuits for resarch or interest.
Following that simple minded philosophie wie would still fight pest and cholera, use in best case horses for transport, and would kill everyone who only think about using wheels as a nonbeliver.
No matter what question or topic is brought up in the digital section of diyaudio, someone will say - all crap, use a TDA154x ... This can´t be the final answer for someone who wants to do experiments and not belive in some obscure kind of resistor sound guru religion.
I ´m really interested if that CDP uses the DACs really in a each one halfwave configuartion, or if its parallel or balanced configuration.
but every day someone posts some wisdom like:" whatever you do, a nos 1543 will sound better" no matter what the question was.
Those NOS 20 year old DAC prophets seem to be unable to accept someone asks question and wants answers on topic, or simply to measure and test different circuits for resarch or interest.
Following that simple minded philosophie wie would still fight pest and cholera, use in best case horses for transport, and would kill everyone who only think about using wheels as a nonbeliver.
No matter what question or topic is brought up in the digital section of diyaudio, someone will say - all crap, use a TDA154x ... This can´t be the final answer for someone who wants to do experiments and not belive in some obscure kind of resistor sound guru religion.
I ´m really interested if that CDP uses the DACs really in a each one halfwave configuartion, or if its parallel or balanced configuration.
I'll save you the posts/links.
I am not sure if you're speaking to me. If so,please say my name in you post. It is so stupid to refer to someone without adressing to the person in question IMO. Anyhow, just measuring harmonic distortion does not cut it as I tried to say.
FYI : PCM56 is almost or just as old as TDA1541. One of the first DAC chips in mass production. Maybe it measures better but it does not sound as well as its 20 year old opponent.
I hope you will see that we still are talking/discussing second generation DAC chips in *2004*. If all modern things would be better we would be talking PCM1738 etc.
I do not adress specific to you, as i got lots a valueable informatiopn from you. But i´m sure you will see this phenomenum also here.
I allready wrote what i see in measurements. I measure to see if it works like it should, it help me to find mistakes. For example i go to much THD from one of my D1 stages, what dissappeared after change of one of the IRF510s.
I allready wrote what i see in measurements. I measure to see if it works like it should, it help me to find mistakes. For example i go to much THD from one of my D1 stages, what dissappeared after change of one of the IRF510s.
jean-paul said:
You're a lucky guy, Jean-Paul... I was accused to go off-topic just by saying that.
I post this schematic taken from a BB datasheet.
Good old times...
The engineer who made this datasheet was surely an audiophile.
Not anymore at TI.
Look at the op-amps they recommend.
NE5534 now?!
Attachments
jean-paul said:
I think it's even older.
It's from the time Philips had only 14bit dacs.
I have a very old CDP (Telefunken
) with led display, Sony mechanism and one PCM56 for the two channels, with a switching circuit. Doesn't sound bad at all, with OPA2132 on the analog stage.
But ultimately, the op-amps don't deserve the CDP.
This thing is just for playing some music when I'm fiddling with something on the basement.
jean-paul said:
* If it measures better it'll sound more like the real thing.
It is clear that some amount of the right kind of distortion is pleasant to the ear.
* You can read that even in application notes.
The Technics is no 1bit player.
* I wrote that. 6dB better noise and distortion.
As long as I can see most of the harmonics, my instrument is ok.
* I have also PCM56 in Denon DCD3300 and it is noisy and has distortion like 1541 in nonos and addidtional nonharmonics.
But I like the sound.
Sorry, off topic again....
* If it measures better it'll sound more like the real thing.
It is clear that some amount of the right kind of distortion is pleasant to the ear.
Measuring only harmonic distortion is not enough is what I try to make clear for the nth time. I know that some amount of the right kind of distortion is pleasant to the ear, that is not the point.
* You can read that even in application notes.
The Technics is no 1bit player.
I know that, I was talking harmonic distortion here, your point of interest. As in: a lot of modern chips have better distortion figures but are dull sounding. As in: not sounding like the real thing. Don't think I don't know my cdplayers.
* I wrote that. 6dB better noise and distortion.
As long as I can see most of the harmonics, my instrument is ok.
With best instruments I mean your ears if that wasn't clear. They do the ultimate selection with normal human beings. Having heard very good gear that did not grab my attention for more than a few minutes I think it is sane to say that listening is the best test after measuring ( or before ).
* I have also PCM56 in Denon DCD3300 and it is noisy and has distortion like 1541 in nonos and additional nonharmonics.
But I like the sound.
I like sound in general as well ( not the sound of DCD3300, I've heard to many Denons that sounded like crap and failed more than they should so I skipped this one ), what is relevant is if it sounds better than the other ones you mentioned. PCM56 is a mediocre performer that was less in ( subjective ) quality than TDA1541A in its days. Now it still is AFAIK. I don't even care to look up its datasheet now. You'll rarely catch me on : "been there, done that" but in the case of PCM56 it is : "been there, done that" for me. If I can make you happy with it I can send you some selected ones or its better 18 bit equivalent: AD1860.
* Nobody knows if this clock is better.
????????? Then do your measurements with a reference that is known to be good instead of fiddling around with cutting canned oscillators open with unknown jitter specs. You can only compare when you have sources that are of known origine/quality. Otherwise I don't see the point of comparing when the modified/new clock can be as good or bad as the original one.
* You say it yourself. You prefer 1543 which has the highest distortion.
Where did you get that from ? Selective dyslexia ? I tried more DAC chips than I would like I had tried. TDA1543 is just one that is difficult to screw up because of the inherent simplicity of its application. That's what I said and I still think it is like that. PCM63,TDA1541A, AD1865 and TDA1543 are the ones that I happen to like most where TDA1543 ( TDA1541A clearly is not ) is the one that needs least energy to sound right. Very practical and almost guaranteed good results ( soundwise that is, I did not like much what I measured ). Very practical for starters.
Also it is quite strange to judge a chip in non os when you've never heard or measured it in that application.
I definitely don't like TDA1547 and Sony current pulse DAC chips or any 45 MHz clock driven DAC chip ( too hard to get silent ). Write that down so you have something to respond on next time.
I think we'll never agree on anything I'm afraid. If you want the chips you can send me a PM if you like.
* If it measures better it'll sound more like the real thing.
It is clear that some amount of the right kind of distortion is pleasant to the ear.
Measuring only harmonic distortion is not enough is what I try to make clear for the nth time. I know that some amount of the right kind of distortion is pleasant to the ear, that is not the point.
* You can read that even in application notes.
The Technics is no 1bit player.
I know that, I was talking harmonic distortion here, your point of interest. As in: a lot of modern chips have better distortion figures but are dull sounding. As in: not sounding like the real thing. Don't think I don't know my cdplayers.
* I wrote that. 6dB better noise and distortion.
As long as I can see most of the harmonics, my instrument is ok.
With best instruments I mean your ears if that wasn't clear. They do the ultimate selection with normal human beings. Having heard very good gear that did not grab my attention for more than a few minutes I think it is sane to say that listening is the best test after measuring ( or before ).
* I have also PCM56 in Denon DCD3300 and it is noisy and has distortion like 1541 in nonos and additional nonharmonics.
But I like the sound.
I like sound in general as well ( not the sound of DCD3300, I've heard to many Denons that sounded like crap and failed more than they should so I skipped this one ), what is relevant is if it sounds better than the other ones you mentioned. PCM56 is a mediocre performer that was less in ( subjective ) quality than TDA1541A in its days. Now it still is AFAIK. I don't even care to look up its datasheet now. You'll rarely catch me on : "been there, done that" but in the case of PCM56 it is : "been there, done that" for me. If I can make you happy with it I can send you some selected ones or its better 18 bit equivalent: AD1860.
* Nobody knows if this clock is better.
????????? Then do your measurements with a reference that is known to be good instead of fiddling around with cutting canned oscillators open with unknown jitter specs. You can only compare when you have sources that are of known origine/quality. Otherwise I don't see the point of comparing when the modified/new clock can be as good or bad as the original one.
* You say it yourself. You prefer 1543 which has the highest distortion.
Where did you get that from ? Selective dyslexia ? I tried more DAC chips than I would like I had tried. TDA1543 is just one that is difficult to screw up because of the inherent simplicity of its application. That's what I said and I still think it is like that. PCM63,TDA1541A, AD1865 and TDA1543 are the ones that I happen to like most where TDA1543 ( TDA1541A clearly is not ) is the one that needs least energy to sound right. Very practical and almost guaranteed good results ( soundwise that is, I did not like much what I measured ). Very practical for starters.
Also it is quite strange to judge a chip in non os when you've never heard or measured it in that application.
I definitely don't like TDA1547 and Sony current pulse DAC chips or any 45 MHz clock driven DAC chip ( too hard to get silent ). Write that down so you have something to respond on next time.
I think we'll never agree on anything I'm afraid. If you want the chips you can send me a PM if you like.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.