Also I went back to check on the source of noise.
It turns out my 5V supply is very malfunctioning (30-50mV AC can you believe that?).
I swapped in a SSLV. Problem solved.
Congrats on your successful build!!!
I just posted on finding something like this DAC design on Computer Audio
The minimum DSD DAC - Page 3
Still don't know anything about how it works but would want to try building one though I have not done any active circuits before. What is your cost estimate?
Have you tried to compare the DAC to a commercial one yet?
How is the output implemented? Is there any DC on it?
First of all, it's a DIY product so the expectations are it outperforms off-the-shelf products in the same price range.Congrats on your successful build!!!
I just posted on finding something like this DAC design on Computer Audio
The minimum DSD DAC - Page 3
Still don't know anything about how it works but would want to try building one though I have not done any active circuits before. What is your cost estimate?
Have you tried to compare the DAC to a commercial one yet?
How is the output implemented? Is there any DC on it?
as i mentioned here: http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1.html#post4099916
The cost of it should be something ballpark $300.
Sound?
LeonBernieniv describes it as clear.
Im more leaning towards "real". DACs Ive owned are all in the range of $500-800, I ve never experienced this kind of sound signature, smaller sound stage, very clear picture, every note is presented with well defined edges.
First of all, it's a DIY product so the expectations are it outperforms off-the-shelf products in the same price range.
as i mentioned here: http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1.html#post4099916
The cost of it should be something ballpark $300.
Sound?
LeonBernieniv describes it as clear.
Im more leaning towards "real". DACs Ive owned are all in the range of $500-800, I ve never experienced this kind of sound signature, smaller sound stage, very clear picture, every note is presented with well defined edges.
The output stage if i understand it correctly is AD844 for i/v then lme49720 for LPF then lme49710 for gain.
DC in the output? I've got a few mV which shouldn't be a concern in any system.
First of all, it's a DIY product so the expectations are it outperforms off-the-shelf products in the same price range.
as i mentioned here: http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1.html#post4099916
The cost of it should be something ballpark $300.
Sound?
LeonBernieniv describes it as clear.
Im more leaning towards "real". DACs Ive owned are all in the range of $500-800, I ve never experienced this kind of sound signature, smaller sound stage, very clear picture, every note is presented with well defined edges.
Thank you.
I had two things in my system that have caused soundstages to shrink and move backwards, which in a magnepan Tympani setup is a real oddity. One is a bad match of power cord and amp. The other is Salen-Key filters - much worse while they break-in, which is apparently what Miska put in the DSC1.
In the case of the Fosgate amp with the Bryston power cable from the Bryston amp I switched it to the fat JPS power cable that was on the Classe DR9 and the problem was solved. The Bryston cable was just not a good match for the Fosgate.
You might get a kick up with replacing big film caps (I saw a few 2.2uF on the list) with cheap Dayton foil caps from Parts Express. I use them extensively in my crossovers. One of the reasons I gave up on my Marchand XM44 crossover was that there was not enough space in it for foil caps bigger than 0.01uF.
The other is Salen-Key filters - much worse while they break-in, which is apparently what Miska put in the DSC1.
There are many ways to design active Sallen-Key filters and the performance depends on the filter parameters and component selection. The ones in DSC1 are designed the way they are for a reason... So I don't think it is possible to make a blanket statement about SK filter sound.
I was also thinking about using MFB topology I used in my earlier DAC, but it is more demanding for the opamps used, so I settled for SK in this case. LME49713 would fit for MFB.
There are many ways to design active Sallen-Key filters and the performance depends on the filter parameters and component selection. The ones in DSC1 are designed the way they are for a reason... So I don't think it is possible to make a blanket statement about SK filter sound.
I was also thinking about using MFB topology I used in my earlier DAC, but it is more demanding for the opamps used, so I settled for SK in this case. LME49713 would fit for MFB.
I don't know enough to distinguish a well designed SK from a bad design or a compromised one. Like feedback, can be done right or can make a gain circuit sound like a rotary tool. I would not know how to tell the designs apart. So I like to avoid both.
What would the benefit of an MFB filter be over a SK? A quick read suggests using them where component variance is an issue, besides that it just makes you use another resistor you wouldn't have otherwise. Is there a time domain/waveform preservation benefit?
What about a passive filter with make up gain instead? Like Lampis? It makes a difference in RIAA eq. and crossovers, perhaps here too?
What is the LP filter alignment you use?
Is there a time domain/waveform preservation benefit?
No, there's none. The current one design is already optimized for both time and frequency domain performance, 7 kHz square wave performance is very good.
An externally hosted image should be here but it was not working when we last tested it.
What about a passive filter with make up gain instead?
I don't like the idea and I certainly don't want to have any coils on the audio path. But of course everyone is free to experiment and come up with mods they like... 🙂
Large part of the resulting sonics depend on used oversampling filter and delta-sigma modulator. This design is intended to match well with HQPlayer's filters and modulators.
To me the beauty of an open chipless design like this is we can tweak more stuffs than just changing capacitors and adding coils.
I did try to use a pair of Cinemag input transformers in the amp section to get balanced output, that doesnt improve sound stage very much.
On top of my head right now lots things can be improved without changing design principals, e.g. better PS (tps7a4700s instead of LMs,LTs) more accurate resistor networks.
Also any news about DSC2?
I did try to use a pair of Cinemag input transformers in the amp section to get balanced output, that doesnt improve sound stage very much.
On top of my head right now lots things can be improved without changing design principals, e.g. better PS (tps7a4700s instead of LMs,LTs) more accurate resistor networks.
Also any news about DSC2?
Im thinking about upgrading my i3 computer so I can enjoy DSD512 or upper without clipping with 10 tabs opened in chrome.
My current setup glitches while im opening big tab pages in chrome and playing DSD256.
A question to Miska:
which is more important spec number of cores? cpu clock? size or speed of mem?
My current setup glitches while im opening big tab pages in chrome and playing DSD256.
A question to Miska:
which is more important spec number of cores? cpu clock? size or speed of mem?
No, there's none. The current one design is already optimized for both time and frequency domain performance, 7 kHz square wave performance is very good.
I don't like the idea and I certainly don't want to have any coils on the audio path. But of course everyone is free to experiment and come up with mods they like... 🙂
Large part of the resulting sonics depend on used oversampling filter and delta-sigma modulator. This design is intended to match well with HQPlayer's filters and modulators.
The SK circuits for the filters should not be difficult to duplicate with RC LP filters, you may end up having one more gain stage depending on the number of poles and the Q of the filters. But then you are possibly high up enough in voltage to forego a real output stage.
One of the gain stages can be a transformer that can serve as a filter pole. That was what Ted Smith did. He wanted to forgo another active stage and RC filter and spec'd a custom wound trannie to work as a filter at his specified fc, and serve to be an output stage and decouple DC.
I have to say that as a result there is a hint of "transformer sound" to the PS Audio PW DSD. I think you can pick a transformer that is naturally (i.e. not by deliberate spec) rolling off at one of your preferred filter pole's fc and not necessarily use it for output.
A successful implementation of this passive RC + makeup gain is in Nelson Pass' First Watt B4 crossover. It can do 4th order filters with 2 gain stages and IIRC there might be a buffer at the output. I think the schematics were offered online at his DIY forum, might still be there.
The SK circuits for the filters should not be difficult to duplicate with RC LP filters, you may end up having one more gain stage depending on the number of poles and the Q of the filters. But then you are possibly high up enough in voltage to forego a real output stage.
One of the gain stages can be a transformer that can serve as a filter pole. That was what Ted Smith did. He wanted to forgo another active stage and RC filter and spec'd a custom wound trannie to work as a filter at his specified fc, and serve to be an output stage and decouple DC.
I have to say that as a result there is a hint of "transformer sound" to the PS Audio PW DSD. I think you can pick a transformer that is naturally (i.e. not by deliberate spec) rolling off at one of your preferred filter pole's fc and not necessarily use it for output.
A successful implementation of this passive RC + makeup gain is in Nelson Pass' First Watt B4 crossover. It can do 4th order filters with 2 gain stages and IIRC there might be a buffer at the output. I think the schematics were offered online at his DIY forum, might still be there.
I see where you going with this, a step up transformers followed by opamp or buffer as i/v+lpf?
I am already using this type of configuration:
HQplayer >= DSD128 --> Amanero --> passive filter --> volume transformer (isolation, +6 db gain, additional filtering) --> power amplifier. The transformer really is this one:
SACThailand --> search for STA-522A and look for frequency response. This has been also checked in a laboratory here in Italy and
output at the frequencies of DSD signal is virtually non existant.
Some discussions has been made about a PWM "DSD gain stage" between Amanero and passive filter (note: because of the low output and impulse response a 1st order filter has been used) but the influence of this on jitter could be a problem. The signal maybe should be reclocked after the gain stage and before the filter.
In general I am satisfied with the result., so I agree with the old idea of Ted Smith, that can be applied also to other type of interfaces.
HQplayer >= DSD128 --> Amanero --> passive filter --> volume transformer (isolation, +6 db gain, additional filtering) --> power amplifier. The transformer really is this one:
SACThailand --> search for STA-522A and look for frequency response. This has been also checked in a laboratory here in Italy and
output at the frequencies of DSD signal is virtually non existant.
Some discussions has been made about a PWM "DSD gain stage" between Amanero and passive filter (note: because of the low output and impulse response a 1st order filter has been used) but the influence of this on jitter could be a problem. The signal maybe should be reclocked after the gain stage and before the filter.
In general I am satisfied with the result., so I agree with the old idea of Ted Smith, that can be applied also to other type of interfaces.
Amanero Combo384+CR LPF DSD256 unnecessary DSD DACI am already using this type of configuration:
HQplayer >= DSD128 --> Amanero --> passive filter --> volume transformer (isolation, +6 db gain, additional filtering) --> power amplifier. The transformer really is this one:
SACThailand --> search for STA-522A and look for frequency response. This has been also checked in a laboratory here in Italy and
output at the frequencies of DSD signal is virtually non existant.
Some discussions has been made about a PWM "DSD gain stage" between Amanero and passive filter (note: because of the low output and impulse response a 1st order filter has been used) but the influence of this on jitter could be a problem. The signal maybe should be reclocked after the gain stage and before the filter.
In general I am satisfied with the result., so I agree with the old idea of Ted Smith, that can be applied also to other type of interfaces.
Are you using this design here?
What's the sound signature like?
At the output of Amanero I have only one 33 Ohm resistance in series, a capacitor in parallel and the transformer. I don't need a coupling capacitor because of transformer. I don't need a higher resistance (this also because of transformer).
The isolation from Usb ground is made by a Gefen USB extender. Power amplifier ground is isolated with the output transformer.
The S/N is not exceptional, should be around 85/90 dB, but this configuration has been made for a tube power amplifier whose real S/N could be around the same. So no record dynamic range. Moreover the output signal is lower than 2 V
The output transformers of tube power amplifier will work also as a final filter before the signal is sent to the speakers (Audiostatic eletctrosatic modified).
The sound signature is of a very transparent sound. The sound is the opposite of dark but is not bright, because high frequencies are smooth and "liquid". There is an impression (as in all DSD reproduction I have tried) of a bass a little less present.
Confronting to iFI DSD micro in DSD there is no story: this configuration is a lot better, immediately more transparent. iFi idsd micro (to my knowledge) should use the analogic internal filter of the DAC chip and an active buffer. In my case on the signal path there are not transistors, chips, no feedback, only one resistance, one capacitor in parallel, one transformer.
The isolation from Usb ground is made by a Gefen USB extender. Power amplifier ground is isolated with the output transformer.
The S/N is not exceptional, should be around 85/90 dB, but this configuration has been made for a tube power amplifier whose real S/N could be around the same. So no record dynamic range. Moreover the output signal is lower than 2 V
The output transformers of tube power amplifier will work also as a final filter before the signal is sent to the speakers (Audiostatic eletctrosatic modified).
The sound signature is of a very transparent sound. The sound is the opposite of dark but is not bright, because high frequencies are smooth and "liquid". There is an impression (as in all DSD reproduction I have tried) of a bass a little less present.
Confronting to iFI DSD micro in DSD there is no story: this configuration is a lot better, immediately more transparent. iFi idsd micro (to my knowledge) should use the analogic internal filter of the DAC chip and an active buffer. In my case on the signal path there are not transistors, chips, no feedback, only one resistance, one capacitor in parallel, one transformer.
I recently finished my DSC1 build.
Removed unused 3.3v regulator (it's the faulty one that causing the noise)
Replaced LT1086 with tps7a4700 (much less ripple)
Removed unused 3.3v regulator (it's the faulty one that causing the noise)
Replaced LT1086 with tps7a4700 (much less ripple)
For those who asking for boards, here are the projects on OSHPark I've created and shared.
Disclaimer: I didnt get my boards from OSHPark, but they are considered reliable and good quality.
DAC section:
https://oshpark.com/shared_projects/Dcm5mq1P
($160 for 3 boards)
PSU section:
https://oshpark.com/shared_projects/Zx4dezQf
($159 for 3 boards)
After all it might be cheaper to get PSU section by using kits from other source e.g. diyinhk: 4.17uV Ultralow noise DAC power supply regulator +-12/15V 1A - DIYINHK
Disclaimer: I didnt get my boards from OSHPark, but they are considered reliable and good quality.
DAC section:
https://oshpark.com/shared_projects/Dcm5mq1P
($160 for 3 boards)
PSU section:
https://oshpark.com/shared_projects/Zx4dezQf
($159 for 3 boards)
After all it might be cheaper to get PSU section by using kits from other source e.g. diyinhk: 4.17uV Ultralow noise DAC power supply regulator +-12/15V 1A - DIYINHK
Last edited:
I see where you going with this, a step up transformers followed by opamp or buffer as i/v+lpf?
Sorry I missed your comment earlier.
Yes, that is one option.
But look at how a good quality transformer behaves in the freq and time (phase) domain and it may work as the entirety of the filter stage while being the I/V stage too.
SAC Thailand
There is a version of this transformer without the multiple taps a 1:2.2
Unfortunately, this also has consequences in deep bass THD:
SAC Thailand
This interstage transformer is actually better than the TVC transformer above, having a top end rolloff at >60khz instead of 35khz, but they didn't publish THD info for it.
We were looking at the PW DSD from PS Audio which has an output transformer as the key filter and integral filter, and its bass is weak and woolly - at least before the last firmware upgrade which seemed to have found a compensation in the software domain. We found a partial solution by lowering the input to the filter/circuit by lowering the digital volume by 10db - which still leaves you at bitperfect up to dsd128 and PCM192. But takes you away from core saturation and lowers THD in the bottom octave in a big way.
So implementing the idea here, there would still need to be a drop from the Amanero or XMOS board's DSD on I2S from its 0.5-1v to <0.1V or so, or we would need to go to a very large (power amp output sized - or an actual power amp output transformer used in reverse) transformer.
In the case that we take the voltage down, there would need to be an active output stage. Perhaps something like this: http://www.amazon.com/preamplifier-...ie=UTF8&qid=1420697653&sr=8-1&keywords=ge5670
where you can remove the volume control.
For those who asking for boards, here are the projects on OSHPark I've created and shared.
Disclaimer: I didnt get my boards from OSHPark, but they are considered reliable and good quality.
DAC section:
https://oshpark.com/shared_projects/Dcm5mq1P
($160 for 3 boards)
PSU section:
https://oshpark.com/shared_projects/Zx4dezQf
($159 for 3 boards)
After all it might be cheaper to get PSU section by using kits from other source e.g. diyinhk: 4.17uV Ultralow noise DAC power supply regulator +-12/15V 1A - DIYINHK
I like DIYinHK XMOS USB board and high precision clocks and low ripple regulators. (1uV) the XMOS USB board is supposed to have isolation features and allows you to run the XMOS chip off of onboard power rather than source (PC) USB power so the grounds should be (but I am not certain) isolated.
Sorry I missed your comment earlier.
Yes, that is one option.
But look at how a good quality transformer behaves in the freq and time (phase) domain and it may work as the entirety of the filter stage while being the I/V stage too.
SAC Thailand
There is a version of this transformer without the multiple taps a 1:2.2
Unfortunately, this also has consequences in deep bass THD:
SAC Thailand
This interstage transformer is actually better than the TVC transformer above, having a top end rolloff at >60khz instead of 35khz, but they didn't publish THD info for it.
We were looking at the PW DSD from PS Audio which has an output transformer as the key filter and integral filter, and its bass is weak and woolly - at least before the last firmware upgrade which seemed to have found a compensation in the software domain. We found a partial solution by lowering the input to the filter/circuit by lowering the digital volume by 10db - which still leaves you at bitperfect up to dsd128 and PCM192. But takes you away from core saturation and lowers THD in the bottom octave in a big way.
So implementing the idea here, there would still need to be a drop from the Amanero or XMOS board's DSD on I2S from its 0.5-1v to <0.1V or so, or we would need to go to a very large (power amp output sized - or an actual power amp output transformer used in reverse) transformer.
In the case that we take the voltage down, there would need to be an active output stage. Perhaps something like this: Amazon.com: Class A GE 5670 tube valve preamplifier preamp amplifier include transformer 110V: Electronics
where you can remove the volume control.
Appalling distorsion figures for those trafos....
//
Appalling distorsion figures for those trafos....
//
That is at rated power - they could have done like Audio Note and used their high nickel permaloy cores to lower THD but chose instead to lower the size and cost of the trannies. The Audionote 80's % nickel core transformers cost more than 10X.
If you go down in current the THD values drop off more rapidly than the current. Going down in signal by 10 db would lower THD by 15-16 db.
- Home
- Source & Line
- Digital Line Level
- Signalyst DSC1