TNT, I went back to partypak0.99 eqv5 again. It seems to have brought back the clarity but retained the naturalness and definitely improvements in bass from the caps mod. The change can perhaps be regarded as drastic. I would encourage anyone to try both and see the difference. I was running 1.19firmware 1.20FPGA.
Last edited:
Cool! Did you try my 4k filters I posted in the Brewing thread the other day?
https://www.diyaudio.com/forums/dig...ilter-brewing-soekris-r2r-52.html#post5658075
//
https://www.diyaudio.com/forums/dig...ilter-brewing-soekris-r2r-52.html#post5658075
//
If you're using windows try the Resampler-V plugin for foobar.
Let the foobar plugin do the upsampling and bypass the DAM.
The Resampler-V plugin has a very nice interface that let you adjust about everything.
Let the foobar plugin do the upsampling and bypass the DAM.
The Resampler-V plugin has a very nice interface that let you adjust about everything.
Attachments
Last edited:
Did I insult you ? I do listen to my DAC's, procured a Sennheiser HD800S for that purpose. But I'm not designing only by listening, otherwise I would never get anything done. I'm drawing on my 40+ years experience designing computers and audio, simulate it, build breadboards as I don't trust simulations very much, measure it, and off course, listen to it.... And I do listen to live music for reference.
You've moved the goalposts a bit higher, haven't you? Do you have any sense of fair play?
Yes, it will be quite an improvement over stock. But even more capacitance will still make a difference.
Yes. At that capacity you would definitely want to go for polymer electrolytic caps for the low ESR. We still don't know what size is best yet. If Paul's original simulations are right, if we adjust the series resistor to 0.01R we would benefit greatly up to around 4800uF, at which point there would be no peaks.
Edit: Sorry didn't see your latest post. 1000uF *3 would not be too much. I have 1500uF * 3 and it works great now. I've yet to see evidence that more capacitance would hurt things. Just don't use the latest firmware.
Last edited:
Does it help? Isn't the FPGA upsampling supposed to be as good as any other software implementation?... This might let it run cooler though I guess.
Soren could you also reply to my earlier request for help? It's quite passive-aggressive of you to choose to ignore certain questions. There is nothing insulting in androa76's claim. It was indeed exaggerated as we all know you do listening tests. But to get the facts right you don't have to insult him back.
I still believe that the vref caps make a huge difference on rev4/5, though I'm not sure if lowering the series resistor to 0.01R would help further as is shown in Paul's original simulations. It would've helped a great deal and very easy on your part to simply explain that you think simulations are not always reliable. But are Paul's simulations trustworthy? If so, is 0.01R indeed better than 0.1R if we have 5000+uF per rail?
I'm also having big problems with the latest firmware but I'll post about it in more detail later. Again thanks for all your help!
Last edited:
That's seems like a good show. If he ever comes to Chicago I'd very much like to meet him in person
The presentation probably would've been better if it was in his native language but I'm an engineer too and giving presentations isn't my favorite either...
Last edited:
100uF Samsung I believe. There are also smaller ones: 47uF per rail (plus change?).
1330uF should be good. I went for 1500uF because it was very cheap and basically as large as small footprint polymers can get. If you're ordering from digikey/mouser then maybe it's worth going for the 1800uF option directly for not much more money. Or even 2200uF if you have more vertical space.
you can also go for single large electrolytics directly as they have low ESR too. Panasonic offers a good selection
Upgrade to 1.19 fw again. Sound wasn't distorted and too soft like 1.20 FPGA. What I can say is that compared to EQHQv5 apo the new 4k linear has noticeably more bass and is perhaps softer in the treble. No idea if that's good or bad.
Baffling that I didn't notice the soft sound with 1.20 FPGA before the caps mod - I listened to it for quite a while. Regardless, the vref caps still added to the bass significantly, and most likely dynamics too. With Paul's filter, bass is neither boomy nor loose. The 4k linear is fine too. I don't know which is supposed to be better, but I'm happy with Paul's filter that is known to be technically superior with 2k taps constraint. The sound could perhaps be characterized as harsh, but I'm not audiophile enough to be able to tell if it's something that can be fixed without coloration. Maybe a 4k filter implemented with the same principles as EQHQv5 would make it smoother. What a nice personal project it could be to an electrical engineer with interests in signal processing, or a hobbyist with time to spare
Soren, since the vref cap mod clearly made a positive difference to many of us, could you advise on whether the series resistor value change would benefit those of us with added capacitance?... Thank you!
Baffling that I didn't notice the soft sound with 1.20 FPGA before the caps mod - I listened to it for quite a while. Regardless, the vref caps still added to the bass significantly, and most likely dynamics too. With Paul's filter, bass is neither boomy nor loose. The 4k linear is fine too. I don't know which is supposed to be better, but I'm happy with Paul's filter that is known to be technically superior with 2k taps constraint. The sound could perhaps be characterized as harsh, but I'm not audiophile enough to be able to tell if it's something that can be fixed without coloration. Maybe a 4k filter implemented with the same principles as EQHQv5 would make it smoother. What a nice personal project it could be to an electrical engineer with interests in signal processing, or a hobbyist with time to spare
Soren, since the vref cap mod clearly made a positive difference to many of us, could you advise on whether the series resistor value change would benefit those of us with added capacitance?... Thank you!
Last edited:
Soekris... No way. I wtote it is fair dac for that price, but it is maybe not optimized in software region enough. Maybe test on bigger system will not be bad.
Analog_sa.. Ok i will be more direct from other side. It is more artiffical than my 4490 and 4495 tube out dac es9038q2m with transformer out and similar to modified Allo boss dac (that cost... )
Analog_sa.. Ok i will be more direct from other side. It is more artiffical than my 4490 and 4495 tube out dac es9038q2m with transformer out and similar to modified Allo boss dac (that cost... )
What mods do you have and what's your implementation? Mostly just curious.