I would also suggest you use non oversampling. It can be easily tried and if not satisfied you can always return to regular setup (oversampling). Heres the link: http://members.xoom.virgilio.it/hi_fi/com.htm
http://members.xoom.virgilio.it/hi_fi/cd960.htm
I actually didn't cut any traces, just the three of the I2S carrying wires and soldered them directly to decoder's corresponding pins. But your lay out may be different.
http://members.xoom.virgilio.it/hi_fi/cd960.htm
I actually didn't cut any traces, just the three of the I2S carrying wires and soldered them directly to decoder's corresponding pins. But your lay out may be different.
Even though I will try non oversampling when the time comes I'ts my gut feeling that is better a digital filter than an analog low pass filter.
I'll do the non oversampling after I get the analog portion working and then figure out how to do the analog low pass.
What's your feeling regarding this two (different) filter approach?
I'll do the non oversampling after I get the analog portion working and then figure out how to do the analog low pass.
What's your feeling regarding this two (different) filter approach?
Hi Apassgear,
Thanks for your info and pics. I am not a real-pro-experienced-diy-freak like Peter but I would like to share my experience with you:
I have wonderfull results with a CS8412 / TDA1541A / single BTJ I/V /capacitor / to amplifier or direct (4mApp) into headphone. There's only 1 cap in the signal path, no low-pass filter, no oversampling. There's a snap, rythm and ambiance to the bass I've never heard before in my other DACs. The CS8412 has it's own regulator, like the TDA1541A, but the I/V stage (the single BJT) shares its +/-15V power supply with the TDA1541A, decoupled with 100hm/1000uF, running 13mA Iq each channel.
This setup has a assymetrical slewrate but I don't know wheter that actually mattters. Anayways, the discrete single BJT I/V stage can be made small enough to fit a 8 pin DIP package, just to give you another option.
Goodluck, regards,
Thijs
Thanks for your info and pics. I am not a real-pro-experienced-diy-freak like Peter but I would like to share my experience with you:
I have wonderfull results with a CS8412 / TDA1541A / single BTJ I/V /capacitor / to amplifier or direct (4mApp) into headphone. There's only 1 cap in the signal path, no low-pass filter, no oversampling. There's a snap, rythm and ambiance to the bass I've never heard before in my other DACs. The CS8412 has it's own regulator, like the TDA1541A, but the I/V stage (the single BJT) shares its +/-15V power supply with the TDA1541A, decoupled with 100hm/1000uF, running 13mA Iq each channel.
This setup has a assymetrical slewrate but I don't know wheter that actually mattters. Anayways, the discrete single BJT I/V stage can be made small enough to fit a 8 pin DIP package, just to give you another option.
Goodluck, regards,
Thijs
Thijs,
Well that's a very minimalist TDA1541 and very atractive proposal for me and specially if it sounds as you say, this is what I'm looking at the end.
Did you make a dedicated PCB or is it a P2P setup?
Just to get an idea, is the BTJ connected similarly to Rbroer shema without the opamp? Grounded gate?
Why did you select a CS8412?
Thanks for your input hope to hear more from you.
Well that's a very minimalist TDA1541 and very atractive proposal for me and specially if it sounds as you say, this is what I'm looking at the end.
Did you make a dedicated PCB or is it a P2P setup?
Just to get an idea, is the BTJ connected similarly to Rbroer shema without the opamp? Grounded gate?
Why did you select a CS8412?
Thanks for your input hope to hear more from you.
Coming back to my LM6181 I/V "stage".
I removed the ceramic bypass caps and connected the DAC directly to the inverting input of the opamp still with the 1K FB resistor. I did not remove the 0.01u cap across both rails though.
To my surprize I got some AC noise through the speakers which was not present before and this is without any signal going through the CDP!!!
Feeding signal I got same noise as before.
Beats me!
Any suggestion?
I removed the ceramic bypass caps and connected the DAC directly to the inverting input of the opamp still with the 1K FB resistor. I did not remove the 0.01u cap across both rails though.
To my surprize I got some AC noise through the speakers which was not present before and this is without any signal going through the CDP!!!
Feeding signal I got same noise as before.
Beats me!
Any suggestion?
I think it was Phred who liked the sound of Scott Nixon's DAC. As far as I know, it doesn't use low pass filter.
And I remember when you said that both TDA1541/43 are ancient chips and non-oversampling can't be any good. It happens that I just tried both of them without digital filter and they are not any worse than my PCM1704 DACs. I can only blame myself for not trying it earlier
But, I also don't measure anything and rely only on my (and some friend's) ears. Maybe it's good, maybe not.
And I remember when you said that both TDA1541/43 are ancient chips and non-oversampling can't be any good. It happens that I just tried both of them without digital filter and they are not any worse than my PCM1704 DACs. I can only blame myself for not trying it earlier
But, I also don't measure anything and rely only on my (and some friend's) ears. Maybe it's good, maybe not.
More speculation...........
I have a low pass filter on mine. I had a talk with Scott and think you might get away without one in some cases. The high output impedance of the DAC's I to V stage (a resistor in this case) acts as a filter with the cable impedance to provide some filtering. Many amps and preamps also contain low pass filters at the inputs. In light of this you actually do have a low pass filter but at a pretty high frequency. I put low pass filters on the inputs of most analog circuits to suppress RFI noise at the input. I do it because it sounds better (and I know the engineering reasons why) So much for the objective versus subjetive debate on this one.
"But, I also don't measure anything and rely only on my (and some friend's) ears. Maybe it's good, maybe not."
Jocko and I listen to everything we design as well as circuits designed by others. That's how you make correlations between good sonics and good engineering pratices. Often, one hears sonic differences and then goes back to measure things to see what the technical reason might be. You can then pay attention to this part of the design on the next go around. It's kind of like the joke about the guys who goes to the doctor and says" It hurts when I do this." The doctor tells him "Then don't do that!"
I have a low pass filter on mine. I had a talk with Scott and think you might get away without one in some cases. The high output impedance of the DAC's I to V stage (a resistor in this case) acts as a filter with the cable impedance to provide some filtering. Many amps and preamps also contain low pass filters at the inputs. In light of this you actually do have a low pass filter but at a pretty high frequency. I put low pass filters on the inputs of most analog circuits to suppress RFI noise at the input. I do it because it sounds better (and I know the engineering reasons why) So much for the objective versus subjetive debate on this one.
"But, I also don't measure anything and rely only on my (and some friend's) ears. Maybe it's good, maybe not."
Jocko and I listen to everything we design as well as circuits designed by others. That's how you make correlations between good sonics and good engineering pratices. Often, one hears sonic differences and then goes back to measure things to see what the technical reason might be. You can then pay attention to this part of the design on the next go around. It's kind of like the joke about the guys who goes to the doctor and says" It hurts when I do this." The doctor tells him "Then don't do that!"
Peter Daniel said:I will simply remove low pass analog filter. I don't think it's needed
Playing with fire AFAIK. I don't have sophisticated equipment but my scope reveals RF garbage coming out with the filter removed. The filter is needed, especially when using SS amps without input filtering. In the case of tube gear the necessity is probably arguable. A friend of mine used the non os TDA1543 DAC with and without filtering on his 300B SE amp and wasn't able to hear any difference but kept the filter "just in case".
In my setup the difference is audible so I use filtering. I use 1 % polystyrene caps for this purpose as tolerance is critical in this setup. That also counts for the resistors used for I/V.
Hi Assapgear, hi all, nice to hear from everybody again.
I also realize that NonOversampling could mean intermodulation distortion from the high frequency quantisation noise. Yet is does sound OK to me. I also build Scots DAC, but the TDA1543 looks bad on my scope. This picture of the noisy low-level signal on my scope keeps popping up when I trying to listen. I put 10nF (?) parralel to the output resistors to tame the HF-noise.
The IV stage is jocko's published simple IV-stage but using resistors instead of CCS. 1 BJT, 1 diode, 3 resistors. Im sure Jocko's version will outperform mine. Both are build within a hour.
Regrads,
Thijs
I also realize that NonOversampling could mean intermodulation distortion from the high frequency quantisation noise. Yet is does sound OK to me. I also build Scots DAC, but the TDA1543 looks bad on my scope. This picture of the noisy low-level signal on my scope keeps popping up when I trying to listen. I put 10nF (?) parralel to the output resistors to tame the HF-noise.
Hmmmm that's not a compliment is it? I wasn't trying to be mininmalistic. The DAC-board has 6 regulators and 16.000uF total capacitance on board, excluding the I/Vstage. I hope this makes it a little less minimalistic. However it is P2P wired.
I got them cheap, they are easy to solder (P2P) and I use only a 44.1/16bit CD source so I don't need the CS8414.
The IV stage is jocko's published simple IV-stage but using resistors instead of CCS. 1 BJT, 1 diode, 3 resistors. Im sure Jocko's version will outperform mine. Both are build within a hour.
Regrads,
Thijs
Seriously Elso ?
Well, I know the feeling.
I just moved and overall my soundstage improved, except for the lowest bass, it's missing now
Last week I listened to a Sony DVP9000ES SACD player with the Vakuumstate board and low jitter clock upgrade.
It was very clear redbook is flawed; lacking too much microdetail in comparison.
Each time going back from SACD to the same album on redbook, it's quite obvious how much it's missing; goodies like ambiance, CD sounds it's in a hurry, soundstage, sweetness of treble, decay...
I'll keep my non-os TDA1541A-S1, Kwak-clock, synchronous reclocked, Jocko's common base (soon to be the "better" one as he hints and I've verified it with him) player, but know there's better out there, not in the circuits, but just more info in the data available to start with...
The end of the day, SACD has more potential. Note this V.S.E. upgrade takes the raw DSD signal stream, and it will not be available in all SACD players.
I also took the liberty (hope you don't mind Allan) to reverse engineer the V.S.E. circuit; it's very good.
Discrete Pooge regulators, but setup as shunt regulators (similar to LCaudio's zapfilter), fed by current sources with current sources for the LED references, fet buffer with dmos cascode and complementary darlington. DC servo to finish.
Cheer up Elso,
Maybe you should upgrade your turnable ?
Re: Oh brother.......where art thou???
But I think he did. And I think I know why he did it. I think I also do say the same.
For those interested, I did measured TDA1543 DAC, powered by 6V and I/V converted by 1K resistors. THD is 0.08%, and IMD is 0.05%. As seen, figures are quite decent, but still notably worse (let us say 10x to 30x) than almost any commercial CDP. That is about distortion measurements. Was funny to make them, but more than it said something about the measured units, it said something about THD and IMD measurements themselves (again). Sorry, I do not have a way to measure TIM, but it anyway concerns unit following DAC and not the DAC itself.
Then:
Whatever you do with non-o/s you can not filter it “redbook successfully” unless you practically completely destroy the sound. So for the “legitimate engineers” it will always be incorrect (they will always prefer what they see at their spectrum analyzers to what they hear, if hear anything).
And when I am at TIM, my experience with low passing of the buffered gainclone reminds me it might be better to deal with the amp’s ability to resist to garbage (hmm, GC is SS?) than to bother with the redbook principles of filtering in digital.
For those still interested in non-o/s filtering may I suggest to look at the five filtering options for the non-o/s, five quite marginally visited pages at my site.
Pedja
Ps: …I saw the light!
Jocko Homo said:
But I think he did. And I think I know why he did it. I think I also do say the same.
For those interested, I did measured TDA1543 DAC, powered by 6V and I/V converted by 1K resistors. THD is 0.08%, and IMD is 0.05%. As seen, figures are quite decent, but still notably worse (let us say 10x to 30x) than almost any commercial CDP. That is about distortion measurements. Was funny to make them, but more than it said something about the measured units, it said something about THD and IMD measurements themselves (again). Sorry, I do not have a way to measure TIM, but it anyway concerns unit following DAC and not the DAC itself.
Then:
Whatever you do with non-o/s you can not filter it “redbook successfully” unless you practically completely destroy the sound. So for the “legitimate engineers” it will always be incorrect (they will always prefer what they see at their spectrum analyzers to what they hear, if hear anything).
And when I am at TIM, my experience with low passing of the buffered gainclone reminds me it might be better to deal with the amp’s ability to resist to garbage (hmm, GC is SS?) than to bother with the redbook principles of filtering in digital.
For those still interested in non-o/s filtering may I suggest to look at the five filtering options for the non-o/s, five quite marginally visited pages at my site.
Pedja
Ps: …I saw the light!
Attachments
Re: In case you guys don't know......
I tried transformers at the output of TDA1543. I didn't have Jensens, but Lundalh LL1570XL were within reach. Some people in my are do serious business by installing them at the output of CDPs for bettter sound. But when I instaled them in my DAC, it was not an improvement. A bit of the real essence of music was gone and I will never put them back again.
I also tried MSR860 diode in the PS (DAC circuit) of my Marantz CD94. And here I was for another surprise. While both MUR and MSR 860 worked great in the gainclone, in a CD player they simply messed up everything. The soundstage was gone, the top end dissapeared and everything was flat. I quicly reinstalled the original Philips supplied diodes.
Jocko Homo said:
I tried transformers at the output of TDA1543. I didn't have Jensens, but Lundalh LL1570XL were within reach. Some people in my are do serious business by installing them at the output of CDPs for bettter sound. But when I instaled them in my DAC, it was not an improvement. A bit of the real essence of music was gone and I will never put them back again.
I also tried MSR860 diode in the PS (DAC circuit) of my Marantz CD94. And here I was for another surprise. While both MUR and MSR 860 worked great in the gainclone, in a CD player they simply messed up everything. The soundstage was gone, the top end dissapeared and everything was flat. I quicly reinstalled the original Philips supplied diodes.
Thanks Jan for your input. Very welcomed.
If you read my previous posts you will notice that the ceramic caps were added to a 1uf film cap, which is arguably low in value but did sound apparently OK. Leads from the PSU are rather short, about 10 Cm (4"). The PSU +- 15V is dedicated to this stage and is the Jung/Sulzer type w/o sensing option. You can have a partial look at the posted pics.
So removing the cer caps I only went back to what I had which did not produce the humm, but...
Also you will read that when I added the cer caps I also added a MKS 4 0.01u across the both rails at the chips and this one was not removed. So this is the suspect now.
I have been concerned about DAC damage, you know, working with little experiance in this area, though I strap myself to ground, but testing some not well design output analog stages, so the question....
How easy is to damage the DAC through the output pins?
And taking this opportunity I have a silly question concerning the 14 bit caps that go along with the '1541 DAC. Are these reverse polarity?
Meaning if the leg that goes to ground is positive in relation to the connection to the pin?
If you read my previous posts you will notice that the ceramic caps were added to a 1uf film cap, which is arguably low in value but did sound apparently OK. Leads from the PSU are rather short, about 10 Cm (4"). The PSU +- 15V is dedicated to this stage and is the Jung/Sulzer type w/o sensing option. You can have a partial look at the posted pics.
So removing the cer caps I only went back to what I had which did not produce the humm, but...
Also you will read that when I added the cer caps I also added a MKS 4 0.01u across the both rails at the chips and this one was not removed. So this is the suspect now.
I have been concerned about DAC damage, you know, working with little experiance in this area, though I strap myself to ground, but testing some not well design output analog stages, so the question....
How easy is to damage the DAC through the output pins?
And taking this opportunity I have a silly question concerning the 14 bit caps that go along with the '1541 DAC. Are these reverse polarity?
Meaning if the leg that goes to ground is positive in relation to the connection to the pin?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.