The simplistic Salas low voltage shunt regulator

Status
Not open for further replies.
Jump to Latest
#3,143
It is stable and very silent.... I am listening to the left channel only and somehow there seems to have some improvements in the bass weight.... I will now do the right channel shunt and report back.

PS: I mean jfet 2N5457... this and the bjt are really close
 
Last edited:
#3,145
I am only listening for 10 minutes now but already know we have a winner.

Now I have the "punch" and presence from the original v1, with the über detail from the v1.2

I am really pleased with the results and will report latter with more details but for now I must congratulate you for once more you disclosed your true genial posture 🙂
 
#3,150
This is major information !

After all the tonal presentation is dictated by the shunt layout. Inversely to popular belief, the amplifier is in reality modulating the signal coming from the PSU.

Why this change in behaviour ? V12R is quieter, more stable and easyer to implement and most of all, manages to produce the missing presence in upper low / lower mids...

Maybe that utmost clarity from the V12FF is missing but now Zappa sounds like Zappa again 🙂
 
#3,151
RCruz said:
This is major information !

After all the tonal presentation is dictated by the shunt layout. Inversely to popular belief, the amplifier is in reality modulating the signal coming from the PSU.

Why this change in behaviour ? V12R is quieter, more stable and easyer to implement and most of all, manages to produce the missing presence in upper low / lower mids...

Maybe that utmost clarity from the V12FF is missing but now Zappa sounds like Zappa again 🙂
Click to expand...


Play village of the sun for instance from his live with the Mothers, listen to the proximity effect and presence of the close up mike speaking. The primary detail in music is ought to be in mid bass, not in mid high. That one ices the cake. Any old British loudspeaker designer would attest. The clarity is there, but not spearheading like in an AKG 701. Its differences in the gain/phase contouring in the supersonic range I introduced to answer your question. I am seeing the reg as an audio power amplifier of high damping factor focused in 0 Hz. The audio amp is modulating the power ''quality'' available to it yes, not any signal must be coming from the PSU though.
 
#3,152
Hi Salas,

I've discussed this with you before, and I'd like to again, if I may:

I have just this evening finished my TDA1541A DAC, and it is now working fine, but I had a horrible problem with it earlier, and it was quite depressing for a while - I was getting MASSIVE distortion coming through. The reason, it turned out, was because the Salas shunt reg supplying the digital reclocker was going to a TL431A and the Salas shunt had dropped to 4.85v. Just bringing it back up to 5v made this HUGE distortion disappear, and it made me thinnk yet again about those TL431s and whether I should bin them. There are issues caused by having two shunts in series it seems clear to me. Oliver put them in there on the recommendation of the TDA1541 Guru Thorsten Loech, but I am still not so sure they help at all.

My question to you is this:

Can a Salas shunt reg, such as those Oliver DVBProjekt made, running at 200mA, supply two identical TDA1541A components, pulling identical current and voltage in perfect parallel, without a problem? I am guessing that there's no problem, and I should do without the TL431s, but last time you said something about nodes, and I didn't understand that.

Many thanks - sorry for rambling on.....
 
#3,153
Yes it can. Using the remote sensing makes it better for the different consumption nodes not to intermodulate, just keep short thick same length cabling in your non remote PCB. For digital I would prefer you use IRF9520 or 9620, 9610 for the CCS Mosfet.
 
#3,156
I was assuming that as they are two identical units pulling the same exact current for the exact same signal at the same time (pretty much down to the millisecond) that there would be little to no intermodulation. Is that a fair assumption?

If not, how do I implement remote sensing?
 
#3,157
Its a fair assumption. Still better decoupling isn't a bad idea. Why not two regs at a future point for instance. Anyway, you got to do your test with what you have, TLs out, and if its encouraging is all that matters for now.
 
#3,158
In order to replace the TL431 and keep cables short, ideally I would go into the TL431 output pin hole, which will only take a .6mm solid core wire. I have some pure silver .6mm solid core cable. I know you suggested thick cable, but is it any good to use this if kept to about 5cm max? In my experience silver is an amazing conductor (of heat it's incredibly fast, for example).
 
Status
Not open for further replies.
Top Bottom