The simplistic Salas low voltage shunt regulator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
stormsonic

stormsonic

Member
Joined 2008
pidesd said:
with the clock i was hearing weird kind of echoes and for the receiver nothing was wrong but the performance was clearly better with the sigma or flea. i ve checked my wiring and everything seems fine. i m using bc560c for bc857, 2sk170bl for jbf862 and a mpsa18 for bfr92p. also i omitted c3. anyway, i ll do more testing.
Click to expand...

it's purpose is to output 3.3V from for well regulated 5V input voltage. A single transistor is used as CCS, not very good PSRR.
If you substituted BF862 with 2SK170 you should check biasing resistors and change them accordingly.
If you hear strange echoes, and those echoes is kind of "metallic" or "glassy" sound, then regulator is probably oscillating, check with your scope.

For low 1/f noise (DAC, clock,...) you should use opamp based regulators, like PCM Flea or Walt Jung designs.
 
pidesd

pidesd

Member
Joined 2008
oh...i didnt know it needed a good psu. luckily, the reg is fed by a sigma11.

i ve retested it with a small film cap bypassing c2(dont know if it did anything) and now it works ok. it s a good idea to check the biasing resistor, i forgot about that. it may not hurt indeed.

yes i am kind of sceptic about using the shunt regs you guys are doing with digital stuff,
because on the sim graphs the noise increases with frequency and digital chips operate at higher frequency than 20khz. am i not right?

that said, if find the sound a bit smoother with the shunt reg feeding the clock than with the flea, but that is about the only improvement i can hear.
 
A

AndrewT

R.I.P.
Joined 2004
pidesd said:
.............yes i am kind of sceptic about using the shunt regs you guys are doing with digital stuff,
because on the sim graphs the noise increases with frequency and digital chips operate at higher frequency than 20khz. am i not right?
Click to expand...
it's at the higher than audio frequencies where a discrete shunt reg is so much better than an IC series reg and really shows where to put the technology to help the digital side.
 
F

fred_com

Member
Joined 2009
pidesd said:
i just wish a 3.3v SS1.2 could be made. i know someone earlier said he was going to experiment on this (fred?). maybe he could join in?
Click to expand...

Sorry, 3.3V reg is not ready yet (v1.2), but I will do it eventually (probably next week), and will post results here.

Salas, I have a noob question somewhat: do I need to change value of R1 for 5V reg ?

Thanks,
Fedor
 
Last edited:
K

Kuro

Member
Joined 2008
Hi guys, pardon for my ignorance. If I were to build the Salas shunt regulator, what sort of noise figure can I expect? My requirement is +5V and nearly 3A current load.

The linear PS I built using LT1085 is giving me <2mV noise on the output with full load. Can the Salas shunt regulator do better than this?
 
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
ikoflexer said:
I built a few prototypes that ran at about 2A and the noise was less than 100uV p-p ripple. I agree with Salas, if I was to build another high current version v1.0 would be my choice.

What needs 3A? What's the circuit you want to power?
Click to expand...

Thanks for measuring. That would be 35uV RMS. Used the HP analyzer for that?

P.S. Was it a V1.0 with Zener or Norton Vref?
 
K

Kuro

Member
Joined 2008
ikoflexer said:
I built a few prototypes that ran at about 2A and the noise was less than 100uV p-p ripple. I agree with Salas, if I was to build another high current version v1.0 would be my choice.

What needs 3A? What's the circuit you want to power?
Click to expand...

Is the 100uV p-p noise with 2A loaded or unloaded?

My current design is using LT1085 with bypass cap on the ADJ pin and it has 200uV p-p noise when there is no load. When loaded, the noise is a little less than 2mV.

I'm trying to power my entire digital room correction preamp with it. Despite all being digital, some chips are still behaving like analog, such as the ASRC chip. I hear a big sonic improvement with the LPS I just built. I'm wondering if there is any quietier supply out there.
 
iko

iko

Ex-Moderator
Joined 2008
Salas said:
Thanks for measuring. That would be 35uV RMS. Used the HP analyzer for that?

P.S. Was it a V1.0 with Zener or Norton Vref?
Click to expand...

Oops, didn't have a look at this thread lately.

It was a while ago when I measured it, before I had the HP. I used my old Tek scope that goes down to 100uV/cm resolution. Take the measurement only as an indication, not the absolute truth. The actual value was smaller than what I reported.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom