Salas SSLV1.3 UltraBiB shunt regulator

Jump to Latest
#4,101
IMG_1441.JPEG
 
#4,102
funhouse61 said:
Hi,
I have a problem on the negative side of the circuit. As soon as I turn on the power, small clicks appear and the printed circuit heats up at D4, you can see it in the photo, it becomes translucent.
My system must provide an analog output ian canada, the normal one, not the pro one, in 12v in 500ma, my transformer supplies in 16/0/16 and others in 9v (if it is important) and it does 80 va.
So I don't know if it's a connection error on my part, I'm not sure that's it but you can tell me, on the diagram that I made these are the 2 ways that I tried.

I had tested before without any load and I had no problems and the voltage was stable.
I also wanted to specify that the other loads of the transformer are not connected but are secure.View attachment 1399532
Click to expand...
500mA is too large for Ultrabib. May be better use L-Adaptger.
 
#4,103
bbest said:
It is all asbout your post #3546
Salas

Post in thread 'Salas SSLV1.3 UltraBiB shunt regulator'

Unpredictable in a way it will lose enough OLG from its normal spec and will also be easier to make oscillate. Because J3 will operate in a tiny DC voltage bracket. Making R8 5.6k could give some OLG back. C3 should be Low ESR. If easy to test you may give it a try. If no good result IRL530 also works in higher voltage range all the same, so not wasted.

Option two sounds like a proven wise scheme i.e. 5V shunt pre-reg + final low noise LDO closely located on R2R pcb. What is opa365 3.8V plan? Maybe LT3045?
  • Like
Now R8 = 5.6K as you advised. Left as is or change on 270?
Click to expand...
OK that was for a high Vgs J3 vs Vout. Depends on how much voltage you now got across R8 for deciding better. How much?
 
#4,104
funhouse61 said:
Hi,
I have a problem on the negative side of the circuit. As soon as I turn on the power, small clicks appear and the printed circuit heats up at D4, you can see it in the photo, it becomes translucent.
My system must provide an analog output ian canada, the normal one, not the pro one, in 12v in 500ma, my transformer supplies in 16/0/16 and others in 9v (if it is important) and it does 80 va.
So I don't know if it's a connection error on my part, I'm not sure that's it but you can tell me, on the diagram that I made these are the 2 ways that I tried.

I had tested before without any load and I had no problems and the voltage was stable.
I also wanted to specify that the other loads of the transformer are not connected but are secure.
Click to expand...
Hi, I suspect its because of sharing the transformer between two AC inputs. Ubibs on board rectification ask for their their own V-V each.
 
#4,107
R8 resistor don't change that voltage (Vgs) by itself only because its coming from inside the JFET. Goes along with Idss. Gives 1.5V/5.6k=0.26mA bias current in this case. Still works. That R8 high value was suggested to stiffen a high Vgs J3 as CCS for retrieving some lost OLG when used in a merely 3.8V Reg output TTL M2 situation.
In 5V or below Vout applications prefer smaller Idss/Vgs samples for that J3 current source to work best. Then smaller value R8 can be calculated.
Reason is Vgs eats from the already little available voltage across J3. Has to have more voltage than own Vgs across it to function properly.
 
  • Like
Reactions: bbest
#4,109
Salas said:
R8 resistor don't change that voltage (Vgs) by itself only because its coming from inside the JFET. Goes along with Idss. Gives 1.5V/5.6k=0.26mA bias current in this case. Still works. That R8 high value was suggested to stiffen a high Vgs J3 as CCS for retrieving some lost OLG when used in a merely 3.8V Reg output TTL M2 situation.
In 5V or below Vout applications prefer smaller Idss/Vgs samples for that J3 current source to work best. Then smaller value R8 can be calculated.
Reason is Vgs eats from the already little available voltage across J3. Has to have more voltage than own Vgs across it to function properly.
Click to expand...
Good (but it is a bit complicated for me 🙂 ), so left 5.6K for R8?
 
#4,112
funhouse61 said:
Hello, but I'm not sure I understand what transformer I would need or what kind of secondary, so.
Click to expand...
An alternative way (since you invested in an expensive three wire center tap Toroidy already) is to remove the diodes from those two pos/neg Ubibs and create an external shared bridge rectifier with center tap wire 0V configuration. Also measure the pulled diodes with your multimeter's diode check to make sure none is broken, you will need four. Follow this schematic and connect the external rectification bridge to the Ubibs C1 capacitors pins with wires as shown. Its a wise ancient schematic found on a parchment in a museum. Should work. 😎

External Common Bridge.jpg
 
  • Like
Reactions: myleftear
#4,115
If I understood your situation correctly, then yes you can use that modification diagram.
You are to link the two C1s for meeting the center tap zero Volt transformer wire at polarities shown. Zero line will be automatically carried over to the DC regs 0V output sides too through the boards.
 
#4,118
Air wiring with care, or you can do the external bridge on a small perforated board.
Anyway, remove all the existing rectifier diodes from the two boards and recycle four of them (tested first) for the external bridge.
This way you also automatically remove the glowing hot board point and the possibly compromised hot diode of your first test from the new situation.
 
You must log in or register to reply here.
Top Bottom