Masterpiece

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am using Borbely all-fet series regulators in all my phono preamps, and they are outstanding subjectively. But they use out of production K170/J74 fets with idss greater than 9ma and which are becoming more expensive than gemstones.
I compared them to Jung SR , and Borbely sound was more neutral and dynamic.
Some high end designers prefer feedback free simple follower volt. regulators with CCS and RC filter for zener diode . Example are old Krell preamps, like PAM-3, KSA-5 etc., Aragon preamps, Coda 05 phono, Parasound JC-2 etc. Extremely well made in Italia AM audio preamps use the same type VR with eight paralleled bd139/bd 140 transistors to lower output impedance.
 
Member
Joined 2009
Paid Member
So you liked the Borbely better ?
Could you post a schematic ?
Did the batteries and the Borbely remove the remaining noise issue ?

As far as i can tell, Andrejs PSU uses a zener and two pass transistors at the output.

I could imagine that it is something like this but with stronger transistors :
Simple Voltage Regulators Part 2: Output Impedance

Yes I like better the Borbely.
As you know Borbely work has copyrights, could you ask him the schematic EB803/244 at Facebook.
Yes.
 
Member
Joined 2009
Paid Member
I am using Borbely all-fet series regulators in all my phono preamps, and they are outstanding subjectively. But they use out of production K170/J74 fets with idss greater than 9ma and which are becoming more expensive than gemstones.
I compared them to Jung SR , and Borbely sound was more neutral and dynamic.
Some high end designers prefer feedback free simple follower volt. regulators with CCS and RC filter for zener diode . Example are old Krell preamps, like PAM-3, KSA-5 etc., Aragon preamps, Coda 05 phono, Parasound JC-2 etc. Extremely well made in Italia AM audio preamps use the same type VR with eight paralleled bd139/bd 140 transistors to lower output impedance.

I have also EB-202/254 all-fet dual low-noise regulator this is a different Erno design using only fets, the EB-803/244 uses low noise opamp and mosfet as series element.
 
I would say no. SS shunts cab be made with output impedance in the mOhms up to high frequencies.
I think the answer lies more in the distortion spectrum at the output.
At least the guys that made the PSU test in Linear Audio found that out.
The PSUs that did best subjectively had the most benign output spectrum.

And I do agree to that. Current experience tells me that a PSU, first and for all, be a 'Low-pass-filter', this should take out all HF-junk. Secondly it should have a good line regulation (it's own PSRR) to be able to block all incoming junk (I think that the Paradise PSU is real good at this). Next, when powering an pre-amp, there should be a modest amount of load regulation, all other load regulation should be done by local decoupling.
 
Very interesting.
Yes, since i heard the tube supply i am convinced the PSU makes a big difference.
Of cause, everybody is saying so but it is more dramatic then i thought.

:nod::nod: I know that!! Remember when I said that some time ago I got attacked (just like for the V-Caps that turned out to be outstanding caps) told that the circuit doesn't have enough PSU Rejection?
:mischiev:

If you guys trust me a bit more ahahaha I can lead you into the holy of the audio...right there!! :D:D
:cool:

Anyway, I still think that shunts deserves attention, I feel there are potentials maybe combined with tube rectification.
I can't say that for sure as my first version of shunt sounded great with batteries not so much with regular tranny, but I have a revised version of that shunt which promises to behave much better so I will give it a try to that one too and see what happens.
 
And I do agree to that. Current experience tells me that a PSU, first and for all, be a 'Low-pass-filter', this should take out all HF-junk. Secondly it should have a good line regulation (it's own PSRR) to be able to block all incoming junk (I think that the Paradise PSU is real good at this). Next, when powering an pre-amp, there should be a modest amount of load regulation, all other load regulation should be done by local decoupling.

yeah but your thoughts still don't explain why your shunt sounds dull compared to the simple rectifier solution...and apparently simulates 1billions dB of attenuation :eek::eek: :D

:cool:
 
As far as I can see from the pictures Andrej's supply has tube rectifiers followed by a LC-filter (no CLC!) and paralleled zeners. This is still a (passive) shunt design!

Yes, this is still a passive design like you say.
This would prove that there is a way to make a shunt sound extremely good!
I don't know the internal contruction of these zeners but I think that putting an actual shunt on circuit have much more work to refine it the way you want it to sound compared to a simple zener...but it is doable!
 
Any regulation has a dynamic impedance in the time domain. Be it series, be it shunt, be it a mix.
I do not trust simple simulation in LT-Spice.
Where is the Monte Carlo Analisys ?
Where is thermal distortion ?
Where is the impuls response ?
etc.etc.

I have never ran such a detailed analysis on something other than when I was in school....I wouldn't know now where to begin nor understand the practicle sense of doing that since ultimate goal is sound quality and I don't know where these info would do for you.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I can not answer that as I do not have this experience. Basically the shunt is exactly the same as the Salas shunt, is the Salas shunt sounding dull? and/or attenuated (by 10^9 (units?))?

I did not see credit in the guide though.:p
The general architecture is following, especially in the parallel part. Which is different in open loop parameters and biasing. The CCS in yours is BJT CV+CC, i.e. much different.
Its very easy to change everything with just values, imagine what happens with secondary differentiations and other parts choices.
You can have the same reg with two different loads described harsh or dull. Difficult to strike balance for a general apps shunt reg. System integration is the key.;)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.