The simplistic Salas low voltage shunt regulator

[jwb]

Originally posted by Salas


I preferred the simple and widely available Zener solution. Somewhere I have read that it is even better, but I have not tested.

Its a finished project as much as I am concerned. Many people are very happy with it. But of course its up to you guys if you want to test it with more elegant additions or performance boosting twists.

I think you'd have to be crazy to use a zener above 6.3V. A red LED makes so much less noise, and adding noise sources together is sublinear, so you can add as many of them as you like. I use four red LEDs in my 9V version.

Zeners less than 6.8V are far more quiet than their high-voltage counterparts.

Has anyone (besides me) done a PCB layout of this regulator? My layout can be seen in this PDF but I'd love to see some others. In mine, I put the current source at the load point. I tried to sensibly use separate ground tracks where I felt the ground currents might interfere with one another.

[jameshillj]

'evening to all, just up on a lazy day!

Okay - we have lined up the horses and jockeys ...

1. +ve supply layout for now (not much difference anyway)
2. Hence - (been saving that one!) j1 = k170 only, j2 = room for oyjer pins later.
3. R10 <-> j3 plus options space
4. Include R with radial. axial C1 (and the HFZ size)
5. leave our extra R at j1 (maybe space?)
6. I like this 0.1R so can directly measure Shunt current - but as tracking here at M2 is critical, possibly add the extra surface mount on top of main track for optional cutting of ... I'll try it and see.

I see a mention of a "pre filter" input cap - perhaps a 1,000uF radial option? Plus the usual 5mm c/c terminal blocks? And perhaps to see if can squeeze in some closer pads for the IRF9610s without trouble - will use the usual offset pin scheme for power M1,2 - look a bit strange at first.

That about covers most everything, so I'll get started and see what it looks like - then we can get stuck into all the usual errors and mods. We've got the new Xmen movie on tonight - great brain food, but the usual ghastly sound!

[Salas]

Originally posted by jwb
I think you'd have to be crazy to use a zener above 6.3V.

Naah, not that dramatic noise from a cap bypassed and locally ccs'd Zener. If it was so noisy it would hiss. Some who made it in the 15-25 of Volt, report they thought it was switched off when used on headphone amps. The 8.2V Zener value is said to be specifically noisier of all. Of course if someone wants to make a lengthy LED string for over 24V regs, it will look very cool, but I doubt if there will be some appreciable practical difference, since the leds will become enough by then.

[Salas]

Originally posted by jameshillj
'evening to all, just up on a lazy day!

Okay - we have lined up the horses and jockeys ...

1. +ve supply layout for now (not much difference anyway)
2. Hence - (been saving that one!) j1 = k170 only, j2 = room for oyjer pins later.
3. R10 <-> j3 plus options space
4. Include R with radial. axial C1 (and the HFZ size)
5. leave our extra R at j1 (maybe space?)
6. I like this 0.1R so can directly measure Shunt current - but as tracking here at M2 is critical, possibly add the extra surface mount on top of main track for optional cutting of ... I'll try it and see.

I see a mention of a "pre filter" input cap - perhaps a 1,000uF radial option? Plus the usual 5mm c/c terminal blocks? And perhaps to see if can squeeze in some closer pads for the IRF9610s without trouble - will use the usual offset pin scheme for power M1,2 - look a bit strange at first.

That about covers most everything, so I'll get started and see what it looks like - then we can get stuck into all the usual errors and mods. We've got the new Xmen movie on tonight - great brain food, but the usual ghastly sound!

G'day!

Good luck! Waiting for news!

[Lumba Ogir]

Salas,
sorry for being unclear, the source resistor should be a fixed value, I just could not specify it.

[jameshillj]

I just had to sneak a look at your "little pdf" over a cuppa, jwb, and there appeared this* "masterpeice".*** " just knocked it up over a lunch break, eh!!*** Congratulations.*** That's is indeed a nice peice of work and a real example of a different approach.*** I think this is the same circuit as your post #283 - I'll* print off this and put it beside your pcb - a rather* different orientation with that CCS down the bottom as T55, C105, with T53 as the driver smd, etc,** I see those 4 little surface mount leds across C97 on the RHS Reg tht you mentioned* ....will print off the cct to follow the different orientation - a nice puzzle working it backwards - this must be what the* China Copy House does all the time -* it's a whole lot easier to copy than to actually design.

*** Incidently, noticed your raw supply R-C-R-C ripple reduction.* I have given up on trying to get folks to try this - seems to have gone out of fashion or something - noticed Denis Moorecroft made a point about it on his DNM Audio site where he explains the T-Network caps.

[jameshillj]

Sorry about that - hit a wrong button or something strange.

I replaced the 8v2 zener in the headamp with 4 basic reds and a green to get about 8.2v also, and yes it was a lot quieter.
Must add that when I changed out the bipass cap, this also changed things a bit - nothing is ever just simple!

[Salas]

For 9V and especially substituting the 8V2Z is easy. But I use 36V most of the time...

[homemodder]

Originally posted by jwb


I think you'd have to be crazy to use a zener above 6.3V. A red LED makes so much less noise, and adding noise sources together is sublinear, so you can add as many of them as you like. I use four red LEDs in my 9V version.

Zeners less than 6.8V are far more quiet than their high-voltage counterparts.

Has anyone (besides me) done a PCB layout of this regulator? My layout can be seen in this PDF but I'd love to see some others. In mine, I put the current source at the load point. I tried to sensibly use separate ground tracks where I felt the ground currents might interfere with one another.

This is not entirely correct, zeners can have as low noise as LED, in certain conditions even lower noise than LED. Zeners below approximately 6V have true zener breakdown and above this they have avalanche breakdown. The important distinction seems not to be which type of breakdown they exhibit, but how close to
from the transition they are. Zeners close to the transition are noisy as hell, between 6 volts and 7.5 volts being the worst. Low voltage zeners have moderate noise, higher than LED but a 12v zener will have similar noise to a single LED but lower noise than a string of LED for the same voltage. Temperature compensated zeners have dramatically lower noise than regular zeners too and one temp compensated zener at around 8.2v is likely to have similar low noise than a string of LED. Above this voltage the LED will lose out in a big way.

[jwb]

Can you post some representative part numbers? In my experience the 8.2V zeners are by far the worst. I measured a 1N4738 to be (at least) 30dB noisier than a red LED. I can add four red LEDs together and it only costs me 9dB, so I'm still ahead by 21dB.