I wanna do the test analog vs analog but still I don't have the 2x18V tx to do the test, anyway I do the test digital vs digital, I mind good to know.
I read that it could be a good idea to use a series reg before the shunt, to act as prereg so both can compensate and work better in the high frequencies.... Any comments about this compensation ?
I would expect that attenuation of mains and harmonics and of spike interference would allow the shunt to perform better.
But how good are pre-regs at attenuating HF?
Can a 317 attenuate anything shorter than a few us pulse length?
But how good are pre-regs at attenuating HF?
Can a 317 attenuate anything shorter than a few us pulse length?
@AndrewT
are you familiar with http://www.waltjung.org/PDFs/Sources_101_Letter_Revisit_0409.pdf?
How much PSRR is needed/wanted?
are you familiar with http://www.waltjung.org/PDFs/Sources_101_Letter_Revisit_0409.pdf?
How much PSRR is needed/wanted?
Yes, I have worked through all of Walt's site.
You have linked to his paper on CCS, not his papers on voltage regulators, which include pre-reg.
You have linked to his paper on CCS, not his papers on voltage regulators, which include pre-reg.
Yes.
Even simple cascode as CCS part will give you over 120 db of line rejection over whole audio band. Even more line rejection at lower frequencies.
If this is not enough, how much PSRR is needed/wanted? Is pre-reg really necessary?
Even simple cascode as CCS part will give you over 120 db of line rejection over whole audio band. Even more line rejection at lower frequencies.
If this is not enough, how much PSRR is needed/wanted? Is pre-reg really necessary?
I do not think so but another menber is considering building a psu for a class D Amp and as he never experienced the benefits of these shunts, posted the possibility to use a prereg before the shunt ... that is why I started the discussion.
Reflektor 2nd opinion.
Using a small cap in the output zobel is benefitial. Better detail, more "snap", better transient response.
Bypassing the Vref big cap is different. I started with 680uF // .25r + 10u // .015u FT-1
Very extended highs promoting very wide sounstages but somehow loosing the "Full On" presentation of other builds. It seems to have a holow in the lower mids.
After conferencing with Salas, I reverted to 1000u EL alone... Not bad but lacking some speed.
So I decided to try the other way around... reducing the EL cap value, keeping the 0.25r + 10u MKP but without the FT-1. Now I am rewarded with a full scale presentation with very good definition, nice transient reproduction.
The lesson is: Do not overdo. Too many bypasses can be worst than none.
FT-1 is NOT like the FT-3... It has good resolution in the highs but is not even (to my hears).
Using a small cap in the output zobel is benefitial. Better detail, more "snap", better transient response.
Bypassing the Vref big cap is different. I started with 680uF // .25r + 10u // .015u FT-1
Very extended highs promoting very wide sounstages but somehow loosing the "Full On" presentation of other builds. It seems to have a holow in the lower mids.
After conferencing with Salas, I reverted to 1000u EL alone... Not bad but lacking some speed.
So I decided to try the other way around... reducing the EL cap value, keeping the 0.25r + 10u MKP but without the FT-1. Now I am rewarded with a full scale presentation with very good definition, nice transient reproduction.
The lesson is: Do not overdo. Too many bypasses can be worst than none.
FT-1 is NOT like the FT-3... It has good resolution in the highs but is not even (to my hears).
You should not reduce the polar cap too much though. Certainly not under 470uF for not upsetting spec. So in a nutshell what is the capacitors list configuration you feel best with? That could be a config to keep in mind although it has to be translating well in other systems and client circuits in the future on other possible attempts by members. I keep 2 basic conclusions. 1. Its always stable. 2. It ''listens'' to tweaks. Thanks for taking the plunge to do the first one 3rd party independent evaluation. Informative to all.
C1 = 0.047uF
C2 = 330uF
C3 = 10uF
R7 = 0.25r
R3 = 2r2
It is super stable and responds well to every tweak.
Now I have a Full presentation with a good balance between detail, extension and stage size.
In my case, increasing C1, looses detail and bloats bass.
Too big C2 and the bass drags a bit.
c3 can also be smaller, giving tigher bass... A question of system tunning.
What should I expect by increasing R7 to .4r ? (I still have the vitreous resistors merlin sent to me but they measure 0.39r instead of 0.22r)
C2 = 330uF
C3 = 10uF
R7 = 0.25r
R3 = 2r2
It is super stable and responds well to every tweak.
Now I have a Full presentation with a good balance between detail, extension and stage size.
In my case, increasing C1, looses detail and bloats bass.
Too big C2 and the bass drags a bit.
c3 can also be smaller, giving tigher bass... A question of system tunning.
What should I expect by increasing R7 to .4r ? (I still have the vitreous resistors merlin sent to me but they measure 0.39r instead of 0.22r)
C3 & R7 are just a bypass snubber, does not change the real electrical parameters of the reg. So you can do as you please. I would prefer you could be able to ''tune up'' your tone with a really big C2 and its bypass. Even up to 2200uF. That extends the really low bass spec. Anything negative towards that direction should be due to other reasons than value. Also see if you can make do with 100R shunt Mosfet gate stopper and still be stable with your 0.047u Zobel cap.
i finished the 1.0 and they sound a little harsh at times.
i know what they can do because i have built some before.
i ma using irf9540 until i receive irf9610 and i wonder if they
are oscilating.
unfotrunately i don`t have an osciloscope to test.
another reason could be long cables to the load 20cm.
what do you guys think?
i know what they can do because i have built some before.
i ma using irf9540 until i receive irf9610 and i wonder if they
are oscilating.
unfotrunately i don`t have an osciloscope to test.
another reason could be long cables to the load 20cm.
what do you guys think?
With 1330uF (1000u // 330u) it really has something special about it.
With 2200u // 330 the bass is also very good but as I am using a 2200u ZLH, I must wait until it settles because it is also affecting the trebble.
I agree that the best bass comes from bigger EL in the Vref... and I did not use anything in the zobel yet.
PS:
Do you mean replace both 220r gate stoppers by 100r ?
the one with pcb from quanghao group buy.
http://www.diyaudio.com/forums/group-buys/174421-salas-low-shunt-group-design-salas.html
p.s.i don`t use remote sensors either yet.
http://www.diyaudio.com/forums/group-buys/174421-salas-low-shunt-group-design-salas.html
p.s.i don`t use remote sensors either yet.
- Status
- Not open for further replies.