OK, glad that's clear.
It would be nice if a purely passive vref was possible, there is peace of the mind that the dam regs are purpose built and built well (on engineering terms at least).
I have been wanting to try super cap supply for 5V input but dont really like the idea of using series regulator for charging, seems to defeat the purpose.
getting within 4.5-5.5v range doesnt seem impossible by purely passive means,
its mostly down to finding a suitable trafo.
depending on Vf of chosen bridge diodes a 4 to 5VAC trafo would be needed,
it doesnt seem like this range is available in any standard mains trafo, the results are dominated with some other type of low voltage power transformers.
It would be nice if a purely passive vref was possible, there is peace of the mind that the dam regs are purpose built and built well (on engineering terms at least).
I have been wanting to try super cap supply for 5V input but dont really like the idea of using series regulator for charging, seems to defeat the purpose.
getting within 4.5-5.5v range doesnt seem impossible by purely passive means,
its mostly down to finding a suitable trafo.
depending on Vf of chosen bridge diodes a 4 to 5VAC trafo would be needed,
it doesnt seem like this range is available in any standard mains trafo, the results are dominated with some other type of low voltage power transformers.
May be worth reading CanadaIan's response here: Develop ultra capacitor power supply and LiFePO4 battery power supply
Yes. Post 41 onwards. However the entire thread covers the evolution.
Caps still have time to start delivering current - larger the cap the more time it takes typically. I would assume having varying sizes in parallel would help cover the ripple.
Switching at 3MHz means there's pulses of on/off that the PSU needs to cope with so the faster the switch rate the faster the power needs to respond to the demand for current.
Ah so big piece of important info I missed is that Ian's ultra cap design is not required to be connected to reg at all times, with 300F it has enough charge to supply on its own for a period. That obviously will offer the best SQ aswell.
Though if it's an impractically short length of time for a load like the dam, i.e a few minutes, that wont be much good.
Though if it's an impractically short length of time for a load like the dam, i.e a few minutes, that wont be much good.
Last edited:
Got another question.
If I'd feed 352k8/384 material - the first stage target rates, will the first 1941 sampling/filter stage get bypassed?
Which in turn would make all (?) this "filter brewing" task obsolete!?!?
BTW: A friend of mine just ordered a 1941 DAC. I'll than have a chance to take a closer look at the SW part.
While - without having any specifics - keeping this in mind:
If I'd feed 352k8/384 material - the first stage target rates, will the first 1941 sampling/filter stage get bypassed?
Which in turn would make all (?) this "filter brewing" task obsolete!?!?
BTW: A friend of mine just ordered a 1941 DAC. I'll than have a chance to take a closer look at the SW part.
While - without having any specifics - keeping this in mind:
Last edited:
Got another question.
If I'd feed 352k8/384 material - the first stage target rates, will the first 1941 sampling/filter stage get bypassed?
Which in turn would make all (?) this "filter brewing" task obsolete!?!?
BTW: A friend of mine just ordered a 1941 DAC. I'll than have a chance to take a closer look at the SW part.
While - without having any specifics - keeping this in mind:
The filter configuration you can upload is configured with a bypass - technically you could put a filter in the configuration.
if you are feeding 352/384 content that either means the recordings were recorded at those sample rates so the X8 filtering is naturally not needed, though massive downsides are extremely limited content and huge file size at these rates, the whole DAC filtering thing exists from the limitations of red book sample rate, or you are externally upsampling, in both cases bypassing internal filter is desired.Got another question.
If I'd feed 352k8/384 material - the first stage target rates, will the first 1941 sampling/filter stage get bypassed?
Which in turn would make all (?) this "filter brewing" task obsolete!?!?
BTW: A friend of mine just ordered a 1941 DAC. I'll than have a chance to take a closer look at the SW part.
While - without having any specifics - keeping this in mind:
If you cant or dont want to externally filter then the internal filter brewing is really valuable.
Im am using SD card reader for my I2S source so I cant, but this sounds much better than typical PC>USB>I2S chain, better than the small (if any) gains from external PC upsampling.
IMO PC upsampling is a gimmick, even when I used PC>USB>I2S I actually preferred the sound of my DS DACs internal filters than HQPlayer filters, with either DSD or PCM. There was certainly audible differences, but I never felt you could call it a ''improvement'', like what you can get from a hardware tweak.
Last edited:
I did that kind of 352k8/384 resampling with my PCM51xx RPi HATs all along.
With these DACs the internal filters also got bypassed (? - perhaps better - neutralized) at these FS.
It seems there's always space for improvement on these internal DAC filters.
If you read the filter-brewing thread you'll read more than once that it is considered
a "must" to run custom filters for best performance.
AND. PC based DSP is not a gimmick. Usually you can achieve better qualities on a PC.
And beside that you'll gain a lot more flexibility and control over your data.
Just to mention it: ESS Sabre DACs also come with "OSF Bypass".
You'll also loose the onDAC volume control by running it. To me this Sabre function looks
like a real DSP bypass though. The whole function block seems to get bypassed.
Just setting all coefficients (filter and VC) to 1, I don't consider a real "bypass" btw.
I am wondering if that was meant by
What configuration?? Do I need to upload a specific filter set??
Of course running the filters and VC in "neutral " mode, would be a nice starting point.
Thx
With these DACs the internal filters also got bypassed (? - perhaps better - neutralized) at these FS.
It seems there's always space for improvement on these internal DAC filters.
If you read the filter-brewing thread you'll read more than once that it is considered
a "must" to run custom filters for best performance.
AND. PC based DSP is not a gimmick. Usually you can achieve better qualities on a PC.
And beside that you'll gain a lot more flexibility and control over your data.
Just to mention it: ESS Sabre DACs also come with "OSF Bypass".
You'll also loose the onDAC volume control by running it. To me this Sabre function looks
like a real DSP bypass though. The whole function block seems to get bypassed.
Just setting all coefficients (filter and VC) to 1, I don't consider a real "bypass" btw.
I am wondering if that was meant by
Code:
...The filter configuration you can upload is configured with a bypass...Of course running the filters and VC in "neutral " mode, would be a nice starting point.
Thx
Last edited:
Me too, but i realise this is a minority viewpoint. The real time PCM>DSD conversion craze via HQPlayer is another impenetrable mystery to me.
There are plenty of minorities out there. You're not alone.
The whole DSP subject is a huge annoyance. Simply because (certain) DSP is a necessary evil.
And this discussion proves it once more, that there simply won't ever be consensus about it. And that's great. That keeps us going.
However. My question simply relates to an option to increase the/my flexibility with this DAC.
THX
For DSP outside of filtering (room correction, digital XO etc.) PC is invaluable of course, Im headphone user so I often overlook that aspect.I did that kind of 352k8/384 resampling with my PCM51xx RPi HATs all along.
With these DACs the internal filters also got bypassed (? - perhaps better - neutralized) at these FS.
It seems there's always space for improvement on these internal DAC filters.
If you read the filter-brewing thread you'll read more than once that it is considered
a "must" to run custom filters for best performance.
AND. PC based DSP is not a gimmick. Usually you can achieve better qualities on a PC.
And beside that you'll gain a lot more flexibility and control over your data.
Just to mention it: ESS Sabre DACs also come with "OSF Bypass".
You'll also loose the onDAC volume control by running it. To me this Sabre function looks
like a real DSP bypass though. The whole function block seems to get bypassed.
Just setting all coefficients (filter and VC) to 1, I don't consider a real "bypass" btw.
I am wondering if that was meant by
What configuration?? Do I need to upload a specific filter set??Code:...The filter configuration you can upload is configured with a bypass...
Of course running the filters and VC in "neutral " mode, would be a nice starting point.
Thx
Only IME was PC filtering not all its chalked up to be,YMMV.
I think filter design is the real important part, not necessarily large computing power that PC offers (remember PCM upsampling is very light on CPU in HQP, only DSD conversion is the CPU really needed.), luckily we have custom filters with dam and skilled members designing them
Most importantly typical PC and USB is really poor source, alternatives to PC are very limited currently so most people might not realise that.
external filtering or not there is no contest between PC and SD player.
There are ''filter'' files that do as you say, set coefficient to 1 for ''NOS''.
I also wonder with external 352/384 input if it does the same or it's real bypass?
That makes no sense to me. FIRs are involved in either application. It's the same situation.
Except.
These onDAC filters are "necessary" evil and room correction is (IMO) evil (been there, done that).
Enjoy.
That makes no sense to me. FIRs are involved in either application. It's the same situation.
Except.
These onDAC filters are "necessary" evil and room correction is (IMO) evil (been there, done that).
Enjoy.
Not sure what you mean, you said PC based DSP is not a gimmick.
I was only referring to filters, and I assumed more complex DSP was not possible on-board with dam ( it is?) so a PC would be useful in that case.
Its surprising because you would think the idea of multibit DS tech i.e combining elements of R2R tech with DS, would appeal to audiophiles.
yet they prefer to use these dacs more like old 1 bit DS dacs which everyone seemed to hate so much