Salas SSLV1.3 UltraBiB shunt regulator

Member
Joined 2006
Paid Member
Hi Salas

First of all congrats on the new upgraded design that I will build soon.

Now, as you know I like to understand the circuits I build and usually you take the time to write a description of the functions of the various components in your designs. Would you care to do the same here ?

I am particularly puzzled with the norton vref that seems to be working without a dedicated CCS as in the other versions.

Why did you choose to cascade Q3 with Q2 ?

Best

Ricardo
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I reduced the current source to inherent Vbe/R (Q2,R5) because missing in-production low pinch off low noise high output conductance JFET to nest under one Vbe. Inherent also is the Vbe's negative thermal drift until warmed up. On the other hand yet another active noise source has been eliminated. You could put back one though if you must use a Zener instead of VRR when deleting VR1 (Vout=Vz+4.7V). That would solve the warm up drift if its crucial not to have it in some application. But better avoided if not a must. Also mentioned in the guide.

The rest of circuitry goes on same principles as you know in the SSLV series. In this iteration its a Vbe controlled MOSFET current source and a CFP (BJT+MOS) current shunt with the unity gain passive Vref we discussed sandwiched in-between.
 
You can use a lower voltage capacitor for C1. However, I think 5V is the minimal output voltage so you cannot do -3V...

I was under the impression it was symmetrical on the negative board and if necessary to use an LDO as in my application (Hagerman piccolo mc step up) it wouldn’t detract too much from the lower noise levels because it was on the negative rail.
 
Here is my little power station... 3,3V x2, 5V x2 , 10V x2. All grounds only meet at the obscure object of desire!
 

Attachments

  • a.jpg
    a.jpg
    535.5 KB · Views: 885
  • Like
Reactions: 1 user
diyAudio Chief Moderator
Joined 2002
Paid Member
Looking at schemantic on build guide, I think it's a mistake:
''For CC <1A M1 IRF9610...''
Instead of less ''<'', does should not be more ''>''?

The IRF9610/610 at M1 position are trusty enough as alternatives to the listed Fairchilds for below 1A constant current at about 5V across them. Still their core to case thermal resistance is significantly higher and their positive/negative type curves are not as close but they develop less VGS voltage while being easier to find and as fast or little faster for parasitic capacitance.
 
Congratulations for this new reg. Salas, thanks for share and help.
I was confused, excited to find out, I read this thread on one sitting.
Now I learned that there are TWO substitutes for M1, for less than 1A. IRF9610. And then for high current, IRF9640. Ok. Ok.
Other than to power a DCB-03 preamplifier, and a 14v. Tripath chip amp., my plan is also use one UltraBiB as a prereg. set it at 7v. and then chain it to a several 500mA. modules ultralownoise 0.8mV. postregulators LT3405. These I got plenty and cheap on Taobao site. Are claimed a flexible voltage dropout between 1V. and 4V. So theoretically I could power many 3,3v and 5v. gizmos I have in my USB chain: Filters, hubs, memory sticks, clocks, etc.
 

Attachments

  • 2018_0723_23053700-2592x1458.jpg
    2018_0723_23053700-2592x1458.jpg
    580 KB · Views: 846