It could be audible if you don't choose a low enough noise opamp for the I/V. Subjectively when AD8017 is used for I/V the music is noisier without the filter than with. The noise I'm hearing when no filter's in circuit must be IMD, its not there when paused.
You have no need to worry about 200nF on your TDA1543's output, they're current sources inside and happily drive into an AC short circuit.
You have no need to worry about 200nF on your TDA1543's output, they're current sources inside and happily drive into an AC short circuit.
I just had a quick look to ad815 and 8017 datasheet and they seem to have huge slew rates, just like my tpa6120 (ths-fails).Do you need this filter to tame the dac or to compensate your i/v op-amps or both?I remember the article writen by Walt Jung on AD811 as i/v and he clearly mentioned that it needs a filter on the input or a compensation as it gets to oscillate easily.
Last edited:
Very interesting. I noticed the same thing they described about DAC-3 and a lack of audibility or reduced audibility of low level reverb tails and ambiance. I noticed them more clear and present in Allo Katana dac when running it from linear supplies supplemented with film caps on the +-15v rails. I also hear the same low level reverb details with my modded dac during periods of time I was able to get jitter as low I ever was able to get it.
According to Allo, Katana measured lower in jitter than DAC-3 during tests performed by ASR. Katana, being a master mode dac, measured the lowest jitter ASR has found so far, according to Allo. That's a large part of the reason I am still working on reducing jitter.
Katana has ASRC and DPLL turned off, but my modded dac and DAC-3 has them turned on. I did have DPLL bandwidth adjusted to a very low value that I was able to keep stabilized for short periods of time.
Another thing I noticed is that when jitter is as low as possible on my modded dac and in normal operation of Katana dac that low level cymbal sound details approach more closely the almost perfect low level details of that sound exhibited by DAC-3.
Since I was using Katana with the 'thd' output stage, it did not have a pre I/V stage filter, nor does my modded dac. In contrast, I read somewhere that the designer of DAC-3 said there should be a filter cap before the I/V stages to filter out dac noise before the opamps. Unfortunately, I never saw a quote or blog of designer directly, so it was at best hearsay information.
Unfortunately, I did not have a dac to compare that I would consider to have a soft or warm sound. DAC-3 sounds less bright than both of the Q2M dacs I compared with it, but brightness does seem to fall off somewhat with improved jitter (per given interpolation filter). Most likely, the less bright sound of DAC-3 was mostly due to the proprietary external interpolation filter.
Lastly, I have heard what some would describe as hard and fast, or maybe cold and hard, sounding gear. I would not put DAC-3 in that class, but none of the Sabre dacs I have heard would be considered warm. I suspect that sound quality characteristic is related to its harmonic distortion profile acting on complex music signals to produce IMD. THD of the Sabre dac chips is rated down around -120dB (presumably with distortion compensation of H2 and H3 enabled). Some people don't believe distortion down at that level can be audible, but in my experience I believe it can be.
Last edited:
I mean that there's zero desire to listen to the unfiltered AD8017-based DAC. I always want to put the filter back in, without its just too unmusical.
Hey, thanks for the interesting discussion. I have chosen the PCM1794 after I read this article, which has in depth measurement/analysis/comparison of delta-sigma/multibit DACs and their differences. You may find it interesting: Какой звуковой ЦАП лучше? — Меандр — занимательная электроника. It is in russian so you should probably use Google Translate.
By the way, here is the whole setup in a temporary enclosure out of a Polo RL scarf
until I find someone with a CNC nice enough to make me two simillar enclosures.
The setup consists of:
- A DIY AC filter
- 2x15VAC+2x9VAC R-Core transformer feeding the DAC board
- 2x7,5VAC feeding the XMOS reclock and Raspberry Pi
- Raspberry Pi 3 running Volumio fed by a LT1083 + CRC filter PSU(on the black prototype PCB, painted black with a spray can)
- JLSounds XMOS, with it's reclock after galvanic barrier fed by an LM317 + CRC PSU(also on the black board)
- The modified chinese PCM1794 board
- An LCD display+MCU+rotary switch that came with the chinese board used to switch digital inputs and display current signal frequency
While still burning in, the sound of the PCM1794 is amazing. TBH I have never expected such a detailed sound that is still warm and pleasant to listen to from a delta-sigma DAC except for the most expensive Sabre pros. I don't believe the TDA1541 myths anymore.
The good thing is that the whole streamer/DAC costed less than 250$ including the trafos, raspberry, xmos, memory card, DAC board and all the components for the upgrade. It allows me to listen to music from online streaming services such as Spotify and Tidal, local HDD and I can still switch to the toslink coming from my TV with one click and use the DAC for Xbox(pass-through) and Netflix.
I still have to add the IR receiver for an Apple remote I have laying around but Volumio has a web interface so it's not mandatory.
By the way, here is the whole setup in a temporary enclosure out of a Polo RL scarf
until I find someone with a CNC nice enough to make me two simillar enclosures.
The setup consists of:
- A DIY AC filter
- 2x15VAC+2x9VAC R-Core transformer feeding the DAC board
- 2x7,5VAC feeding the XMOS reclock and Raspberry Pi
- Raspberry Pi 3 running Volumio fed by a LT1083 + CRC filter PSU(on the black prototype PCB, painted black with a spray can)
- JLSounds XMOS, with it's reclock after galvanic barrier fed by an LM317 + CRC PSU(also on the black board)
- The modified chinese PCM1794 board
- An LCD display+MCU+rotary switch that came with the chinese board used to switch digital inputs and display current signal frequency
While still burning in, the sound of the PCM1794 is amazing. TBH I have never expected such a detailed sound that is still warm and pleasant to listen to from a delta-sigma DAC except for the most expensive Sabre pros. I don't believe the TDA1541 myths anymore.
The good thing is that the whole streamer/DAC costed less than 250$ including the trafos, raspberry, xmos, memory card, DAC board and all the components for the upgrade. It allows me to listen to music from online streaming services such as Spotify and Tidal, local HDD and I can still switch to the toslink coming from my TV with one click and use the DAC for Xbox(pass-through) and Netflix.
I still have to add the IR receiver for an Apple remote I have laying around but Volumio has a web interface so it's not mandatory.
I had similar perceptions of "unmusical" devices with the most linear and non-oscillating ones...or with preamplifiers that had very low impedance feedback networks or filters in the path of a signal showing less perceivable noise.
The measurements on a scope showed nothing different in two different passive riaa networks up to 20khz square wave signals for an audio application i want to show you.
You have a similar statement of mine at the end of this post :
https://www.diyaudio.com/forums/analogue-source/255149-unu-pnono-riaa-mm-preamp.html#post3948497
I built a device these days and I have a similar problem with the break point of a chosen rc network where i can't easily make a choice between a noisier, but less aggressive with the highs content, opposed to a more effective network at cutting the highs.You can see here what is all about:
https://www.diyaudio.com/forums/dig...cd-players-enhancing-noise-4.html#post5664506
It seems that either the instability or the noise is preferable to our ears too.I realized that after a famous now video on youtube where a guy showed 4 or 5 versions of the same Pink Floyd album recordings and the reel to reel tape was preferred by most people and i personally think that was just because of the high content of noise on tape, not because it was a more accurate recording than the others.
The measurements on a scope showed nothing different in two different passive riaa networks up to 20khz square wave signals for an audio application i want to show you.
You have a similar statement of mine at the end of this post :
https://www.diyaudio.com/forums/analogue-source/255149-unu-pnono-riaa-mm-preamp.html#post3948497
I built a device these days and I have a similar problem with the break point of a chosen rc network where i can't easily make a choice between a noisier, but less aggressive with the highs content, opposed to a more effective network at cutting the highs.You can see here what is all about:
https://www.diyaudio.com/forums/dig...cd-players-enhancing-noise-4.html#post5664506
It seems that either the instability or the noise is preferable to our ears too.I realized that after a famous now video on youtube where a guy showed 4 or 5 versions of the same Pink Floyd album recordings and the reel to reel tape was preferred by most people and i personally think that was just because of the high content of noise on tape, not because it was a more accurate recording than the others.
Last edited:
ADA4898 has much lower noise than 5534(2) and is a very good choice for I-V.
Having said this I would certainly try ABX's pre opamp CLC filter.
T
Much lower voltage noise and much higher current noise. I heard ess9018 with ada 4898 and nothing made me think that this is a much better circuit than the older ones...
I might be wrong, but i also think that actually the output sheer power of an op-amp used in i/v circuits is often ignored in favor of higher slew rates .Using lower output drive with a high slew rate TIA won't give you what you need ... I saw that initially in purely analogue circuits where people considered that higher slew rates op-amps sound usually better , ignoring the fact that they also have greater output power and lower output impedance.It's the case with AD811 vs any audio op-amp in i/v stages where nobody sees it's HUGE quiescent current(needed for higher frequency response) and thus automatic greater output capability into lower impedance loads. That is why i tried paralleling audio op-amps in my i/v circuit and the subjective result is actually identical to the use of a higher slew rate op-amp ...
I know the whole theory backing the necessity of a higher slew rate TIA, BUT I DON'T AGREE with the need for higher slew rate for fighting ultrasonic or digital garbage on the grounds that actually the internal capacitance and miller effect of the input stage of any audio op-amp won't even allow this noise to be processed by the op-amp, while the capacitor paralleling the feedback resistor will actually be those to couple the input higher frequencies to the output bypassing the op-amp ,as the internals of an audio op-amp can't handle it, but that can be solved with an LC-RC combined filter for the higher frequency content at the op-amp output and that was done very well by a few manufacturers on 16 bit DAC's with very good results.You only need the op-amp to be able to drive the passive filter in the audio range.
There's no virtual ground input for the spurious higher frequency content with an audio op-amp, thus the RC network in a TIA is actually bypassing the op-amp.
So you don't need to raise the input noise putting an LRC filter at the i/v stage input if you can place it at the output of the op-amp with similar results.
Try paralleling op-amps in i/v stages followed by passive filtering and you'll hear something interesting that you never considered to be possible. Higher slew rate op-amps are simply useless if they can't drive the output load and the feedback network in the audio spectrum actually generating lots of IMD which sound pretty much like spurious high frequency content.
I might be wrong, but i also think that actually the output sheer power of an op-amp used in i/v circuits is often ignored in favor of higher slew rates .Using lower output drive with a high slew rate TIA won't give you what you need ... I saw that initially in purely analogue circuits where people considered that higher slew rates op-amps sound usually better , ignoring the fact that they also have greater output power and lower output impedance.It's the case with AD811 vs any audio op-amp in i/v stages where nobody sees it's HUGE quiescent current(needed for higher frequency response) and thus automatic greater output capability into lower impedance loads. That is why i tried paralleling audio op-amps in my i/v circuit and the subjective result is actually identical to the use of a higher slew rate op-amp ...
I know the whole theory backing the necessity of a higher slew rate TIA, BUT I DON'T AGREE with the need for higher slew rate for fighting ultrasonic or digital garbage on the grounds that actually the internal capacitance and miller effect of the input stage of any audio op-amp won't even allow this noise to be processed by the op-amp, while the capacitor paralleling the feedback resistor will actually be those to couple the input higher frequencies to the output bypassing the op-amp ,as the internals of an audio op-amp can't handle it, but that can be solved with an LC-RC combined filter for the higher frequency content at the op-amp output and that was done very well by a few manufacturers on 16 bit DAC's with very good results.You only need the op-amp to be able to drive the passive filter in the audio range.
There's no virtual ground input for the spurious higher frequency content with an audio op-amp, thus the RC network in a TIA is actually bypassing the op-amp.
So you don't need to raise the input noise putting an LRC filter at the i/v stage input if you can place it at the output of the op-amp with similar results.
Try paralleling op-amps in i/v stages followed by passive filtering and you'll hear something interesting that you never considered to be possible. Higher slew rate op-amps are simply useless if they can't drive the output load and the feedback network in the audio spectrum actually generating lots of IMD which sound pretty much like spurious high frequency content.
Last edited:
Video op-amps need to drive 50 ohms load ad 100m-1ghz, thus they need high output capabilities which are in favour of an I/V stage that needs to process the audio range with lower values of resistor feedback.
For some reason, most of the DAC manufacturers went on the high current dac output path for the IOUT DAC's so all the high slew rate, unity gain stable, bipolar op-amps developed nowadays are better only at V noise lower input impedance and output capability compared with the old ne5534 or some similar products.They need to drive lower resistor values and corespondent higher capacitors in the feedback network since TDA1541 to date.
Using higher slew rate and higher output op-amps, the TIA is processing all the input signals, meaning that they amplify the high frequency garbage too, then you need even a more aggressive filter for those components.
Some of the older DAC's would deliver lower currents and even lower spurious content which made possible for the op-amps to drive the feedback network easily while a few hundred pico farad capacitor on the input and a passive LC-RC filter at the output would kill the rest of high frequency content .
I have no idea why they kept that old technology only for v-out dac's while some of them have quite a good sound.
For some reason, most of the DAC manufacturers went on the high current dac output path for the IOUT DAC's so all the high slew rate, unity gain stable, bipolar op-amps developed nowadays are better only at V noise lower input impedance and output capability compared with the old ne5534 or some similar products.They need to drive lower resistor values and corespondent higher capacitors in the feedback network since TDA1541 to date.
Using higher slew rate and higher output op-amps, the TIA is processing all the input signals, meaning that they amplify the high frequency garbage too, then you need even a more aggressive filter for those components.
Some of the older DAC's would deliver lower currents and even lower spurious content which made possible for the op-amps to drive the feedback network easily while a few hundred pico farad capacitor on the input and a passive LC-RC filter at the output would kill the rest of high frequency content .
I have no idea why they kept that old technology only for v-out dac's while some of them have quite a good sound.
What is IMD using video opamps for I/V? I ask because Sabre dacs are rated for -120dB distortion, and the new AK4499 is what, -124dB? Suspect that opamps like OPA161x are used for I/V because of their low distortion. ESS uses them for the ES9038PRO evaluation boards, and has made pretty clear they are going for low distortion.
EDIT: I get the impression Abraxalito has taken to using fast opamps for his dac projects primarily because the dac chips are older designs with higher levels clock noise. In that case I could see that slower opamps might not be the best choice, especially since the dacs are typically 16-bit and trying achieve very low distortion with them is not so much of a consideration.
EDIT: I get the impression Abraxalito has taken to using fast opamps for his dac projects primarily because the dac chips are older designs with higher levels clock noise. In that case I could see that slower opamps might not be the best choice, especially since the dacs are typically 16-bit and trying achieve very low distortion with them is not so much of a consideration.
Last edited:
pcm1794 is 24 bit , 8x oversampling...even the 18 bit 4x oversampling TDA1549 needs no filter at the output... I have a tda1541 dac with a cxd1088 4x oversampling chip in a Sony cdp-750 and it has no output filter at all...it doesn't seem like it needs one either!
Last edited:
It's not...it's calculated.Measuring an ess9018 dem board with an AP system 2 there was no measurable distortion at 1khz...It was like there was nothing there...
I think that i listened to the transducer...some esl headphones plugged into an audio amplifier with 100x higher distortions than the DAC itself...So i assume that it was impossible for me to listen to the DAC distortions or noise...I'm more into masking techniques. I prefer to replace nonlinear distortion and patterned noise with harmonic distortion and unpatterned noise...
Okay, I understand that. Good for low cost dacs, but not what I like to do. Don't get me wrong though, I like low cost, but I also like to hear all the little details in a recording that are digitized into it. That means I prefer a lower distortion dac for my tastes, one that sounds great with all the detail, and that is not fatiguing at all. Hard to do right though.