Salas SSLV1.3 UltraBiB shunt regulator

3V Vin-Vout is normally sufficient for this LM three pin fixed reg series. Since their datasheet states 2V minimum at 1A 25C. The reg chip's temperature is a factor. You will measure the whole rail system and make sure in practice anyway. When a reg drops out stops fixing Vout and just follows Vin dropping. You avoid that region. You will also see about max thermal drift of your own UBiB installation to factor in which depends on box ambient temp, the CCS setting, the sinking.
 
  • Thank You
Reactions: MEPER
That is at 20mA. They will basically work for us at 2mA also but they have somewhat shallower IF/VF curve at low mA than our legacy red (someone's bought Mouser's full WP2773ID stock during Christmas). Closer to it is the orange version. We wouldn't care much for such details for the whole voltage output range but at circa 5Vout its the two LEDS and the two transistors (Q2,Q3 Vbe) which are the main Vref contributors. An interesting type is the Vishay TLDR4400 deep red round one (Mouser) which has low Vf again but very steep curve at low mA in its datasheet. I hope its chemistry isn't more noisy than other reds for some reason.
Hello!
Is it possible not to use an LED?
 
Like this

Jump Leds.png
 
  • Like
Reactions: NikNet
I finished the UltraBIB for my streamer clocks,tweaked with a Flexy for Use a Mundorf 4 pole capacitor,This time I used a magnificent 100VA O Core transformer from James transformer.
The UltraBIB gives a completely different dimension to my clocks,thanks Salas.
IMG_9014.JPG
IMG_9018.JPG
IMG_9019.JPG
IMG_9020.JPG

wiring is not over
IMG_9016.JPG


The first version for my dac clocks

IMG_8896.JPG
IMG_8900.JPG
IMG_8904.JPG
 
  • Like
Reactions: Sorenm and Salas
To find out R1’s value you need to take the maximum current that your load draws(150mA in your case) to which you need to sum the spare current of the regulator(100mA). The total here is 250mA
You will need to divide 630mV to 250ma which will give you the value of R1 in ohms. In this case it will be 2.52R

R1 = 630mV / ((shunt spare current mA) + (load maximum current mA))
 
What is the output voltage are you aiming for? What is your application?
For c2 bigger capacitance gives lower output noise and longer startup times. If something digital on the rails maybe you don’t want long startup times..
For c3 you have to select your cap as suggested by Salas(it needs to be in a certain impedance range 0.12-0.22r @ 100khz).
You can play around with c2 but not with c3 unless you have an oscilloscope and know what you are doing.
Unfortunately using black gates doesn’t guarantee you a trouble free ending.
 
So for the analog side of your pcm or for its digital side?
I guess for analog. In this case you can use for c2 220uf16v.
Do you have a datasheet for the 47uf and 100uf caps? You can find there what is their actual impedance and would be easier to see which can be used for c3.
 
  • Like
Reactions: NikNet
You can try the same 220uf but 100uf will be preferred. With 100uf you are sure the reg for digital starts sooner than the one for analog.
Beside this the digital tolerates noise a bit more than the analog so lower capacitance for c2 shouldn’t be a problem.

Edit: if you want the lowest noise possible use 220uf for c2 on the digital reg and 220uf || 100uf(parallel them for 320uf) on the analog reg.
 
  • Like
Reactions: NikNet
You will need to divide 630mV to 250ma which will give you the value of R1 in ohms. In this case it will be 2.52R
Rather 0.58V +/- 3% (560mV-600mV) Vbe I have seen from my BC3XX Q1 transistor samples at the current bias level we use them in this circuit (1-2mA). Small practical division result difference when preferring a 580mV median or a max 600mV to pre-calculate R1. But if you do get 630mV from your samples its more precise to use that value of course.
 
  • Like
Reactions: schultzsch
Salas I want to thank you for sharing this great regulator with us.

I made a small smd regulator pcb for my diy dac. The sound is absolutely fantastic. Better in every aspect than my current super regulator (which i maybe didn't tune very well i guess).

For now i feed it with smps wallwart, but I will buy suitable transformer soon.

19VDC smps input (will change to toroidal transformer 12VAC soon)
10,5VDC output
CCS setting-700mA
M1-IRF9610
M2-IRF530
BC859C and BC807-40 transistors
SUSUMU RG 0603 resistors 10-25ppm
2SK209-GR jfet-s for CCS-s (135R degeneration resistors, 1.8mA current)
Lite-On RED leds - will change that to HLMP-6000
C1-2x Nichicon 4700uF, 35V
C2-Nichicon ES 470uF, 35V
C3-SMD PANASONIC FC 100uF, 35V (0.15ohm ESR)

PCB is 4 layer of which 3 layers are GND. I wanted to have top and bottom layer separated by just 0.2mm
Out of 700mA CCS setting i have 500mA constant load (4 mini local shunts) 100-120mA spare current and the rest is used by digital portion

Is the ccs jfet suitable for my build? Is it not degrading the regulator performance?


1701980771638.jpg
1701980771645.jpg
1701980771651.jpg
 
@juliobozek

"2SK209-GR jfet-s for CCS-s (135R degeneration resistors, 1.8mA current)"
"Is the ccs jfet suitable for my build? Is it not degrading the regulator performance?"

-The 2SK209 is essentially a 2SK117 in SMD form. So its a good low noise Jfet suitable for the purpose.

"The sound is absolutely fantastic. Better in every aspect than my current super regulator"

- I am glad that you like it. Congratulations for your build and enjoy the music.
 
  • Like
Reactions: juliobozek