Not a good Idea for DIY, since you then need to design an external I/V-Deglitcher instead of an ordinary I/V.
The hybrid topology of the D20400A (adding different MSB/LSB DACs of different settling speed together) calls for a Deglitcher, otherwise it will sound very very ugly.
Please find enclosed pictures of the labels of the two versions.
The hybrid topology of the D20400A (adding different MSB/LSB DACs of different settling speed together) calls for a Deglitcher, otherwise it will sound very very ugly.
Please find enclosed pictures of the labels of the two versions.
Attachments
The older one (Non A) uses slower and noisier Operational Amplifiers inside.
I would use it at 192 kHz NOS/ ASRC instead of 384 kHz, which gives it a little more margin to internally settle before the hold signal goes down.
If your sample rate converter allows to setup dithering, you should set it to 20bit. In this case you can listen to any hires material without any significant loss in resolution.
I listened recently to a AK4499, ES9038PRO (Two chips. dual mono), Lavry DA2002 and Classe DAC-1 (2xD20400A dual mono differential), all over the balanced output, using 24/48 material.
Surprisingly the Test had two winners:
- AK4499 (For its sheer neutral resolution) and Classe DAC-1 for its best overall presentation (Better soundstage than AK4499).
- Lavry DA2002 performed best regarding differentiation of the instruments in a complex orchestral mix, but also added a little of midrange grain (Tonally it is not really neutral).
- The overhyped-ES9038pro chip based DAC sounded just a little mushy/artificial, like a cheap toy (sorry).
Ultimately, with D20400 DIY, dual Mono (Two modules, one each channel, balanced output) is a must, if you want to attack the best DAC available today and careful 20b dither.
I would use it at 192 kHz NOS/ ASRC instead of 384 kHz, which gives it a little more margin to internally settle before the hold signal goes down.
If your sample rate converter allows to setup dithering, you should set it to 20bit. In this case you can listen to any hires material without any significant loss in resolution.
I listened recently to a AK4499, ES9038PRO (Two chips. dual mono), Lavry DA2002 and Classe DAC-1 (2xD20400A dual mono differential), all over the balanced output, using 24/48 material.
Surprisingly the Test had two winners:
- AK4499 (For its sheer neutral resolution) and Classe DAC-1 for its best overall presentation (Better soundstage than AK4499).
- Lavry DA2002 performed best regarding differentiation of the instruments in a complex orchestral mix, but also added a little of midrange grain (Tonally it is not really neutral).
- The overhyped-ES9038pro chip based DAC sounded just a little mushy/artificial, like a cheap toy (sorry).
Ultimately, with D20400 DIY, dual Mono (Two modules, one each channel, balanced output) is a must, if you want to attack the best DAC available today and careful 20b dither.
I've designed a few products using ESS DAC IC's and I must sadly agree with you.
Like Digital on the whole, ESS brought up the quality of cheap DAC's - but are not HiFi. I 100% agree with you that ESS sounds "artificial, like a cheap toy" sounds just about OK with simple tracks (but still with no real sound stage / fine detail) - but the second music gets complex ESS just collapses in a "Fart" and its all over...
Also, "artificial" just keeps going though ones mind when listening!... Not real instruments / vocals etc... does the crowd clapping sound like real clapping or pelting rain?
I designed an ES9038Q2M dac that skilled listeners favored over Benchmark DAC-3. It was a pleasure to listen to, but it was very hard to get it working that well. Almost gave up more than once.
It also would have been expensive to commercialize. Would have taken a lot more work to get the cost down.
It also would have been expensive to commercialize. Would have taken a lot more work to get the cost down.
Last edited:
D20400 vs. D20400A
I need to correct a comment, I provided here. I stated the following:
- D20400 is inferior to D20400A and uses slower Operational Amplifiers, which seems to be ENTIRELY WRONG
This was based on measurements and listening tests with a stock Audio Synthesis DSM DAC using a D20400 (Non A), a DF1700P Filter and CS8412 receiver and obviously wrong information/translation regarding AD825 and AD844 OPAs retrieved from Russian websites.
Obviously in the DSM DAC the sound quality was entirely dominated by the missing 20bit dither of the filter and the fact that no highres material could be used without heavy loss in sound quality.
In meantime I was able to exchange the DF1700P by a SM5847, configured for 4xFs and 20 bit dither and the CS8412 by a CS8414. This turned the DSM DAC into a completely different beast, because it now supported 24/96 material.
I also recently acquired two unused STAX branded D20400 (numbered) modules from 1989 and put one of them into the DSM DAC.
This little DAC using Reference Recording's HRX disc material now more or less blasted away every other DAC (Including the two Lavrys, the SFD-2, the DAC-1, a PCM63-K to D20400 Adapter in the DSM, PCM1704 in the Naim DAC and indeed the "super duper"-ESS), except the AK4499 DAC, which is on par with little more resolution but little less detailed sound stage.
I then inserted the two STAX D20400 into my SFD-2 (Replacing two D20400A) and got a significant improvement in measurement performance (DNL related measurements) and little better sound quality as well.
Regarding DNL measurements, the Stax D20400 provided close to 20bit DNL, other (used) D20400 modules about 19bit and the D20400A modules I have at hand are around 18bit, which is still very good, if you consider that all my PCM63-K (Japan and Korea) chips only provided 15..16bit DNL and even the PCM1704-K only shows a rather mediocre 16bit, which is the very same as a TDA1541A-S1.
Related to these measurements and also listening tests, I can state the following:
New Old Stock D20400 DAC modules meet their specified perfromance (still after 30 years !!) and seem to be superior to D20400A, PCM63-K and PCM1704-K
This being said, D20400A seems to be a cheaper, stripped down, mass producible "Lite" OEM version of the D20400 and the STAX D20400 could be a hand picked, numbered version of the standard D20400, but who knows ...
Does anybody have an original (OEM) D20400A datasheet to compare it to the D20400 datasheet, which is available here ?
I need to correct a comment, I provided here. I stated the following:
- D20400 is inferior to D20400A and uses slower Operational Amplifiers, which seems to be ENTIRELY WRONG
This was based on measurements and listening tests with a stock Audio Synthesis DSM DAC using a D20400 (Non A), a DF1700P Filter and CS8412 receiver and obviously wrong information/translation regarding AD825 and AD844 OPAs retrieved from Russian websites.
Obviously in the DSM DAC the sound quality was entirely dominated by the missing 20bit dither of the filter and the fact that no highres material could be used without heavy loss in sound quality.
In meantime I was able to exchange the DF1700P by a SM5847, configured for 4xFs and 20 bit dither and the CS8412 by a CS8414. This turned the DSM DAC into a completely different beast, because it now supported 24/96 material.
I also recently acquired two unused STAX branded D20400 (numbered) modules from 1989 and put one of them into the DSM DAC.
This little DAC using Reference Recording's HRX disc material now more or less blasted away every other DAC (Including the two Lavrys, the SFD-2, the DAC-1, a PCM63-K to D20400 Adapter in the DSM, PCM1704 in the Naim DAC and indeed the "super duper"-ESS), except the AK4499 DAC, which is on par with little more resolution but little less detailed sound stage.
I then inserted the two STAX D20400 into my SFD-2 (Replacing two D20400A) and got a significant improvement in measurement performance (DNL related measurements) and little better sound quality as well.
Regarding DNL measurements, the Stax D20400 provided close to 20bit DNL, other (used) D20400 modules about 19bit and the D20400A modules I have at hand are around 18bit, which is still very good, if you consider that all my PCM63-K (Japan and Korea) chips only provided 15..16bit DNL and even the PCM1704-K only shows a rather mediocre 16bit, which is the very same as a TDA1541A-S1.
Related to these measurements and also listening tests, I can state the following:
New Old Stock D20400 DAC modules meet their specified perfromance (still after 30 years !!) and seem to be superior to D20400A, PCM63-K and PCM1704-K
This being said, D20400A seems to be a cheaper, stripped down, mass producible "Lite" OEM version of the D20400 and the STAX D20400 could be a hand picked, numbered version of the standard D20400, but who knows ...
Does anybody have an original (OEM) D20400A datasheet to compare it to the D20400 datasheet, which is available here ?
Last edited:
maybe here some advices:
https://www.google.de/url?sa=t&rct=...d=1434311729&usg=AOvVaw2qfRuhuzkzikVos6sBr8ow
UltraAnalog D20400 DATA ! I have it all !! (PM to member "nat8808")
Info for UltraAnalog Jitter Analyzer (post #22)
@Jon Lord: check your mailbox
https://www.google.de/url?sa=t&rct=...d=1434311729&usg=AOvVaw2qfRuhuzkzikVos6sBr8ow
UltraAnalog D20400 DATA ! I have it all !! (PM to member "nat8808")
Info for UltraAnalog Jitter Analyzer (post #22)
@Jon Lord: check your mailbox
source of schematic are mentioned in post #22 under
https://www.diyaudio.com/community/threads/ultraanalog-d20400-data-i-have-it-all.254748/
https://www.diyaudio.com/community/threads/ultraanalog-d20400-data-i-have-it-all.254748/
I know this is a long shot, especially after reading this thread, but I am in need of a D20400A. I have a Sonic Frontiers SFCD1 player that has developed issues in the left channel. I have ascertained that the issue is with this module. Would anyone know where I could get one or is this a job for a qualified tech to pull my module apart and tinker? Is fixing the D20400A even possible?
