This is how my ESS9018K2 AVCC analog rails are generated. Is my understanding right ?
Are these kind of caps suitable, they have a big esr 10ohms.
EECF5R5H104 PANASONIC - Capacitor: electrolytic | TME - Electronic components
An externally hosted image should be here but it was not working when we last tested it.
Thanks four your support Joe !
I assume you expect an answer from Joe, but I will come with some suggestions here, as one who initiate the large decoupling capacities approach from quite long time now, and encouraged the use if such caps for DAC chip decoupling power rails, as for the post DAC processing stage.
You may first have in mind that very large capacities, it will cause a quite high current pick at power on (its charging sequence). You must have a power supply (regulators) which can handle/limit such high current, these caps as usually may need for charging.
I`m not very sure your AVCC power supply it may tolerate a big current pick with these caps on its output.
Also the energy accumulated in these caps have to be discharged accordingly when power off. Else, the targeted devices it may rest powered, while the power from the main rails of the system are gone. However ES9018 it tolerate very well to be powered on its analogue power rails, while the rest of system is powerless.
Not all these types supercaps are suitable for decoupling use. You may of course chose the lowest possible ESR. 10 ohm is just huge! Such low ESR supercaps are enough expensive, but one can find too some suitable for decoupling use at a reasonable price.
You may find here also my AVCC PSU, based on such very large capacities:
http://www.diyaudio.com/forums/digi...-s-bdp105-discussions-upgrading-mods-124.html
But as this subject it may be out of topic here, I may suggest the eventual further discussion about these aspects to happen into an more appropriate thread....
Last edited:
Everything needs a power supply, don't they? 😀
In this instance I just wanted him to get the best possible sound out of his Sabre DAC in the configuration he revealed.
.
Actually, maybe that is too low. This is not about getting low ESR. I think the one I used were about 50R and even higher.
The high ESR needs to work in your favour as you need to look at the in-rush current. Coris goes much more extremo than what I am suggesting to you - and he has to deal with large inrush and I see he now uses relays? What I am suggesting is simple and yet still beneficial.
The question is this: If you have an instant 3.3V and the ESR = 10R is the max load on in-rush, then it become 0.33A worst case. Chances are it won't be that high. I now use some 0.33F righ now that are about 50R and the chip regulator used does not freak out. So maybe use a 0.1F to 0.33F with higher ESR is the safer way.
If you are using the output of LM4562 follower to supply power? That may be problematic. I would honestly by-pass then and add the 0.1F.
But maybe first, just stick with fitting the passive 3R3/3R3/0.82uF filter on the Sabre DAC.
Cheers, Joe
PS: I note the ADP151 is rated up to 200mA. So get a value ESR that won't go over that, no lower than ESR 20R, and chances are the ADP151 has some inbuilt protection from over current - bottom line, by-pass LM4562.
.
Last edited:
As we all hopefully learned in high school lab tech, if you change two things at once, you won't know which of the two actually caused the improvements/differences.
Thank you Joe for your advices ! I will try to apply your advice but first i need to find a way to cut the traces in a safer way in order to install those 3r3 resistors.
After this will come the 0.1Farad mod.
Regards,
Adrian
Hi Adrian
Yes, be careful and deliberate to cut those traces. In some cases cutting a track and baring copper, using SMD resistors can be one way to do it.
And to satisfy Julf, try the post-DAC filter first and have a listen and let us know. Then afterward do the power supply.
Cheers, Joe
PS: Maybe Julf might one day be a doer as you are at least trying.
.
Two of the relays used in my AVCC PSU it have the only function to discharge the large capacities at the power off. The third one is used for signalizing purposes (only). The caps charging process at power on it is handled very well by the included regulators.
As these caps it have more the decoupling role/function at the chip power rails, I think high ESR is not just a good thing.
I don't actually think the primary role here is decoupling. I believe it is about storing energy and hence the equivalent to mass/inertia. In Utapi's instance, where he is using existing limited power supply and not one dedicated to loading up storing large amounts of energy, so I have kept and suggested his approach for now to be as simple, call it KISS if you like, so that he can try it in his existing DAC configuration. That should not be beyond him and can advance it further later if he likes. Start is simple for the first time.
The delta-sigma modulator is basically a "clipping engine" and one leading (and intelligent) manufacturer actually adds a 1 Hertz filter on the power to the modulator, one that is passive and not exactly low Z. Noise on the rail gets converted into jitter the moment the rail is 'clipped'. While that is his solution, a 1 Herz filter, the supercaps is simply another way to do it, or else low Z chip regulators would sound just fine on their own. But they don't. So the caps don't need to have absolutely low ESR. The stored energy is what does the trick.
Cheers, Joe
.
I may agree with this statement...
Actually the huge stored/available energy it play the role of a filter in this case, as it attenuate quite much the disturbances induced by the modulation, switching mechanism and so (chip side) and that which may come from outside (power system side).
We see things subtly different - and yet similarly. But the highish ESR means that low Z alone doesn't explain it all.
Cheers, Joe
.
After all we have gone through together... You are a cruel and heartless man!
Ah, that's what I was doing wrong!
An externally hosted image should be here but it was not working when we last tested it.
Julf surrender, your picture reminded me of an incident years ago where my tech put his soldering iron out a hole into the hallway left by a contractor, someone immediately grabbed it.
LOL ! Can't count the times I was busy with SMD work (staring in a magnifying glass) and grabbed the tool without looking carefully. Got wiser though.
Or shaking off excess solder only to see it land on your trousers 😉
Or shaking off excess solder only to see it land on your trousers 😉
Last edited:
I am glad that never happens to me 🙂
You should have seen the carpet of my apartment back in my university days. I guess all the metal in it made it a good antistatic mat...
OT ad personal posts removed.
- Status
- Not open for further replies.