Hi Ryan,
Thanks, I am glad I have found it. It is pretty frustrating when your hard work breaks down after 2hours of playing.
The post from EC designs is very strange and I would judge is as a wrong conclusion.
Based on the datasheet of the UKA capacitors:
The leakage is calculated according to the datasheet and is bigger if the rated voltage is bigger so 16v would perform better than 25v.
I will try your suggestion with -60db recording (just hope my amplifier gain is enough to put out anything audible at this level
)
I am still curious to learn how to measure the DEM is running on 50Hz, can you advise on this?
Thanks, I am glad I have found it. It is pretty frustrating when your hard work breaks down after 2hours of playing.
The post from EC designs is very strange and I would judge is as a wrong conclusion.
Based on the datasheet of the UKA capacitors:
The leakage is calculated according to the datasheet and is bigger if the rated voltage is bigger so 16v would perform better than 25v.
I will try your suggestion with -60db recording (just hope my amplifier gain is enough to put out anything audible at this level
)I am still curious to learn how to measure the DEM is running on 50Hz, can you advise on this?
Attachments
Good morning!
Yes I see what you mean, you have a valid point. The way I understand it is that the higher rated caps would have a higher insulation resistance. According to the data sheet they should have the same leakage- this doesnt make a whole lot of sense to me. Maybe my logic doesnt apply in this case, im really not too sure. Maybe someone out there that isn't a high end solder slinger like myself can enlighten us.
You may need to put your ear a little closer to the speaker than normal to hear the -60dB recording.
To check the DEM, probe either of the DEM pins with reference to GND.
Yes I see what you mean, you have a valid point. The way I understand it is that the higher rated caps would have a higher insulation resistance. According to the data sheet they should have the same leakage- this doesnt make a whole lot of sense to me. Maybe my logic doesnt apply in this case, im really not too sure. Maybe someone out there that isn't a high end solder slinger like myself can enlighten us.
You may need to put your ear a little closer to the speaker than normal to hear the -60dB recording.
To check the DEM, probe either of the DEM pins with reference to GND.
Yes higher rated caps need thicker insulation/dielectric but in turn also need a bigger capacitor surface to reach the same capacitance.
So regarding leakage the bigger surface will probably compensate for the thicker dielectric.
I am listening to the differential setup during running in, and it is an increasingly satisfying experience!
The sound is developing into more color, increasing in dynamics, micro-dynamics and detail. It is extremely transparant and natural sounding.
I have been trying to find a better dac design than my beloved dddac and have failed several times with other designs but this one will surpass it in naturalness, musical flow and dynamics.
Your I2S-sim + Balanced D3 design seems to distill the best out of this old dac chip.
So regarding leakage the bigger surface will probably compensate for the thicker dielectric.
I am listening to the differential setup during running in, and it is an increasingly satisfying experience!
The sound is developing into more color, increasing in dynamics, micro-dynamics and detail. It is extremely transparant and natural sounding.
I have been trying to find a better dac design than my beloved dddac and have failed several times with other designs but this one will surpass it in naturalness, musical flow and dynamics.
Your I2S-sim + Balanced D3 design seems to distill the best out of this old dac chip.
Sounds like the cap issue isn't worth worrying about then!
Your description is pretty much exactly what I noticed going from SE to balanced. Nice to know you got similar results to myself because I have read that a few people have not liked the result going from SE to Bal. Glad you're satisfied with the performace so far.
Well done getting it all working, its not exactly an easy project.
Your description is pretty much exactly what I noticed going from SE to balanced. Nice to know you got similar results to myself because I have read that a few people have not liked the result going from SE to Bal. Glad you're satisfied with the performace so far.
Well done getting it all working, its not exactly an easy project.
Yes, after the 4th or 5th one I started getting higher success rate!
Im not totally convinced doubling up is worth the investment - I noticed slightly more dynamics and I was hoping the the extra current drive would be enough to drive my F5 directly given that it has a 100k input impedance but I was not happy with the result and I still needed the DCB1 after the trans to properly drive the F5, without it the bass suffered a lot. I have not yet tried this approach with the Sowter 1465s, I may have to revisit this sometime soon.
Im not totally convinced doubling up is worth the investment - I noticed slightly more dynamics and I was hoping the the extra current drive would be enough to drive my F5 directly given that it has a 100k input impedance but I was not happy with the result and I still needed the DCB1 after the trans to properly drive the F5, without it the bass suffered a lot. I have not yet tried this approach with the Sowter 1465s, I may have to revisit this sometime soon.
Finally I managed to actually give my D3 a proper housing. It deserves it well, just listened a/b with my Soekris DAM1021 which I really like, and the D3 is more pleasing to listen to. I am glad I went the D3 route. My nephew worked at the Philips department where his colleagues where designing the TDA series, so a bit of nostalgia is involved… never mind…
The setup included TDA1541A S1 (thanks super surfer), the Ryan, Ifi and Salas power supply’s, JLsounds in direct Simultaneous mode with upgraded crystals, rpi4 with Moode audio, Abbas I/V germanium converter, 30R rhopoint I/V resistors.
Next is the balanced D3 setup with the Sowters, all the sub components are already on the wooden test board, look forward to finish that build,
Thanks Ryan for the D3 service.
The setup included TDA1541A S1 (thanks super surfer), the Ryan, Ifi and Salas power supply’s, JLsounds in direct Simultaneous mode with upgraded crystals, rpi4 with Moode audio, Abbas I/V germanium converter, 30R rhopoint I/V resistors.
Next is the balanced D3 setup with the Sowters, all the sub components are already on the wooden test board, look forward to finish that build,
Thanks Ryan for the D3 service.
Attachments
Im going to give it another go in the next week or two, just waiting on some parts on order. It may be the case that my 2 spare populated D3s will be put to better use in a separate DAC, or 2 SE DACs.
Yeah I've read good things about his work, seems to be fairly popular among diyers, im guessing it will sound excellent. It will be interesting to hear your comparison.
Last edited:
Looks great asanden, your very welcome, thanks for sharing.
Ill be interested to know how you go with your balanced setup.
Interesting about your nephew! I wonder if the designers knew their design would still be used nearly 40 years later, not very often this happens in the world of electronics.
Hi Albert,
Nice to see your dac in a fine housing. I am afraid I will never arrive at this stage
Looking forward to your experience with the balanced boards!
One tip if I may: feeding your rpi with a linear power supply will clean up the sound. The most smps have quite a lot of hf noise that pollutes the system.
Good luck with your next build!
Groeten,
Nice to see your dac in a fine housing. I am afraid I will never arrive at this stage
Looking forward to your experience with the balanced boards!
One tip if I may: feeding your rpi with a linear power supply will clean up the sound. The most smps have quite a lot of hf noise that pollutes the system.
Good luck with your next build!
Groeten,
Hi Stefan ,
Thanks, normally I like to use the Salas L-dapter for powering micro computers, but I ran out of space, so I had to use the ifi power X, which has very low noise.
I will report on the new build
, Thanks, normally I like to use the Salas L-dapter for powering micro computers, but I ran out of space, so I had to use the ifi power X, which has very low noise.
I will report on the new build
Hi Albert,
Nice to see your dac in a fine housing. I am afraid I will never arrive at this stage
Looking forward to your experience with the balanced boards!
One tip if I may: feeding your rpi with a linear power supply will clean up the sound. The most smps have quite a lot of hf noise that pollutes the system.
Good luck with your next build!
Groeten,
Hi,
The D3 dac boards had some time to run in last week. Now it is time to start sound tweaking.
The sound could do with some more dynamics and a bit more clarity in my setup.
In my experience Oscon polymer caps are a perfect choice for digital power supply lines but not for analog.
So I changed the capacitors in the -15v line for some Black Gates I had on stock.
C22 is populated with a BG FK series 100uf/25v
C19 is populated with a BG S series 100uf/50v
This sounded immediately more pulsed and open (the caps were already run in).
I also have some BG non polars that I will try on C22 in anti parallel configuration to see if it will better the FK.
Maybe the termination resistors on the Sowter output can also be set a bit higher for more dynamics bit at the expense of more distortion. Anyone with experience with this?
The D3 dac boards had some time to run in last week. Now it is time to start sound tweaking.
The sound could do with some more dynamics and a bit more clarity in my setup.
In my experience Oscon polymer caps are a perfect choice for digital power supply lines but not for analog.
So I changed the capacitors in the -15v line for some Black Gates I had on stock.
C22 is populated with a BG FK series 100uf/25v
C19 is populated with a BG S series 100uf/50v
This sounded immediately more pulsed and open (the caps were already run in).
I also have some BG non polars that I will try on C22 in anti parallel configuration to see if it will better the FK.
Maybe the termination resistors on the Sowter output can also be set a bit higher for more dynamics bit at the expense of more distortion. Anyone with experience with this?
That is pretty interesting. You are likely on to something special.
I run the original D1. Its a very different thing. FWIW, I settled on 100u FKs on +5 -5, a pair of N in super e config on the -15 and the nx hi-q near the chip. The change was like night and day. Big deep sound stage and deep bass punch vs the OSCONs they replaced.
I will follow your experiments in case I ever get around to putting D3 in.
I run the original D1. Its a very different thing. FWIW, I settled on 100u FKs on +5 -5, a pair of N in super e config on the -15 and the nx hi-q near the chip. The change was like night and day. Big deep sound stage and deep bass punch vs the OSCONs they replaced.
I will follow your experiments in case I ever get around to putting D3 in.
Interesting. I found that with the on board shunt regs the difference in cap sound was negligable. Are there any modern day caps that can compare to the black gates?
IMHO many claim to have found a cap they think is nice and can be an alternative, but to date no one can duplicate them.
Audio Note UK bought up the bulk of the existing inventory and used them in their products now for years. Their founder has been trying to restart production now for years, but the tooling was lost and they cannot get the carbon impregnated paper right. They have a line of caps they claim is close but misses that key ingredient.
Audio Note UK bought up the bulk of the existing inventory and used them in their products now for years. Their founder has been trying to restart production now for years, but the tooling was lost and they cannot get the carbon impregnated paper right. They have a line of caps they claim is close but misses that key ingredient.
Thats really interesting Walter.
Im going to have to retract my last statement, because I just remembered I have tried some black gates on a shunt reg for my CEN output stage and the result was very good. The only black gates I have are ones ive pulled out of an old RCD855 but I may have enough to give them a go on the D3. Thanks for the recomendation Supersurfer.
Hows the rest of your tweaking going Superfurfer?
Im going to have to retract my last statement, because I just remembered I have tried some black gates on a shunt reg for my CEN output stage and the result was very good. The only black gates I have are ones ive pulled out of an old RCD855 but I may have enough to give them a go on the D3. Thanks for the recomendation Supersurfer.
Hows the rest of your tweaking going Superfurfer?
Hi Ryan,
As Walter wrote: unfortunately there is no alternative for Black Gate. I have some BG stock and use them only in designs I think are worth it.
Another very good electrolytic capacitor is the Elna Silmic. I even like them over the BG in some areas: f.i. as kathode capacitors in my tube amps.
They are very good for analog ps use! So maybe this can be an alternative for you.
I have not made other changes yet, it is Friday afternoon here and I will spent some tweaking time this weekend.
As you already mentioned: shunts are transparant so everything that is used downstream of the shunt will be audible.
As Walter wrote: unfortunately there is no alternative for Black Gate. I have some BG stock and use them only in designs I think are worth it.
Another very good electrolytic capacitor is the Elna Silmic. I even like them over the BG in some areas: f.i. as kathode capacitors in my tube amps.
They are very good for analog ps use! So maybe this can be an alternative for you.
I have not made other changes yet, it is Friday afternoon here and I will spent some tweaking time this weekend.
As you already mentioned: shunts are transparant so everything that is used downstream of the shunt will be audible.