Voltage regulation and capacitor ground problem

Thread Starter

r92001r

Joined Oct 22, 2023
24
Hello- I’m hoping someone can help me figure out why my voltage regulation is not working. This has an “amplified zener” to drop 10vdc from 39vdc rectified supply (via a 58VCT transformer), and an LM350 to regulate part of the rail to 18vdc. In other words one rail at 18vdc and the other at around 29vdc. I can’t figure out why when the capacitors are connected to ground, the whole regulation stops working. Kind of a noob so probably some silly mistake I have. If anyone can assist that would be great.
69602F2E-14E0-45E4-96AF-B41954280E0A.png
 

dl324

Joined Mar 30, 2015
16,655
Welcome to AAC!
I can’t figure out why when the capacitors are connected to ground, the whole regulation stops working.
I suspect a wiring error.

A schematic would help. It would also be helpful to you if you used a schematic editor that includes a PCB layout tool so it can check your connectivity.
If anyone can assist that would be great.
Since you're a newbie, you need to be aware that there's an art to doing a "good" board layout. Align components, keep them at multiples of 90 degrees, etc.

This is a layout I did for a power supply for an EPROM programmer. It has 2 variable voltages (VPP and VCCp) and one fixed.
wilpwr.jpg1698163326375.png
 

Thread Starter

r92001r

Joined Oct 22, 2023
24
Welcome to AAC!
I suspect a wiring error.

A schematic would help. It would also be helpful to you if you used a schematic editor that includes a PCB layout tool so it can check your connectivity.
Since you're a newbie, you need to be aware that there's an art to doing a "good" board layout. Align components, keep them at multiples of 90 degrees, etc.

This is a layout I did for a power supply for an EPROM programmer. It has 2 variable voltages (VPP and VCCp) and one fixed.
View attachment 305761View attachment 305762
Indeed- sacrificing some art due to use of perf board and space. I would love to see your schematic though- maybe I’ll find a clue in there.
 

dl324

Joined Mar 30, 2015
16,655
I would love to see your schematic though- maybe I’ll find a clue in there.
My power supply takes DC input, so it won't help you.

I think I see a connection error on the bridge rectifier, but I'd prefer to see a schematic before attempting to trace your layout.
Indeed- sacrificing some art due to use of perf board and space.
The layout for that power supply board is mostly, or all (I can't remember), on a tenth inch grid.

I've done layouts that I implemented with perf board:
perfLayout.jpg
The board on the right didn't image correctly. The transfer media I used appeared to have shrunk and I couldn't use the 20 pin connector I had planned to use so I could use a disk drive cable for connections.
 

dl324

Joined Mar 30, 2015
16,655
This is a layout I did for 4 discrete R#S# flip flops. I modified it on-the-fly to do a 6 flip flop circuit on perf board to use up the extra width. The board was a grid or 2 too short, so I adapted the layout to accommodate that.
1698165632914.png1698165945142.png6FFcomponent.jpg6FFback.jpg
 

Ian0

Joined Aug 7, 2020
9,436
1. It's not going to work off a battery - the transformer needs AC.
2. That's a lot of rectifiers for 48V. There's enough there for 2000V (but that won't stop it working)
3. The zener is the wrong way round, and in the wrong place. It clamps the gate-source voltage to 0.6V so the MOSFET will never turn on.
4. There isn't enough current through the zener - a 1W zener will need 20mA to get anywhere near an accurate voltage.
5. Not sure how you hope to get 29V from a 3.9V zener.
 

Thread Starter

r92001r

Joined Oct 22, 2023
24
1. It's not going to work off a battery - the transformer needs AC.
2. That's a lot of rectifiers for 48V. There's enough there for 2000V (but that won't stop it working)
3. The zener is the wrong way round, and in the wrong place. It clamps the gate-source voltage to 0.6V so the MOSFET will never turn on.
4. There isn't enough current through the zener - a 1W zener will need 20mA to get anywhere near an accurate voltage.
5. Not sure how you hope to get 29V from a 3.9V zener.
Thanks for checking it out- yes it is a transformer not a battery, I just couldnt find the voltage symbol on my first go with kicad...
the original had the rectifier diodes doubled up like that for their use of 400-0-400 transformer- but yes I could drop the doubles.
The zener thing is an "Amplified Zener" to avoid using huge/expensive bolt zeners.
I am trying to combine a found circuit (would post author of it but the site I got it from didnt have the author either...) with the LM350 application from the datasheet...
Here is a patch together of 2 of the circuits (mine has a few value/specific mosfet changes; but it should work..or so I thought..).
For some reason it only works when the large caps are not grounded (which presumably means whatever follows wont work because stuff has to be grounded).
I'm just trying to make a project for myself (actually a preamp and Class D power amp in a vintage orange pleather suitcase that held reel to reel tapes)- Im not trying to work up a magic device to sell...
slicespower.jpg
 

Thread Starter

r92001r

Joined Oct 22, 2023
24
I may just have a soldering error. I just figured I would ask because it seemed like some basic issue with the combination someone might recognize ie “you need a coupling capacitor/diode/resisto/something between the two circuits for the ground scheme to work”… I’m learning … slowly but surely…
 

Thread Starter

r92001r

Joined Oct 22, 2023
24
Always draw a schematic before you build the circuit.
Doing a layout from the design in your head is a recipe for errors.
Good advice. This power supply is kind of a remix of two schematics (below) so I assumed the jump to layout wouldn’t be too crazy…
 

Ian0

Joined Aug 7, 2020
9,436
You will see voltage on the output when it is not loaded as current will go through T1 and the zener (which is forward biassed) and to the output capacitor. However, the forward biased zener is clamping the Gate voltage of the MOSFET to 0.6V, so it can never turn on.
 

Thread Starter

r92001r

Joined Oct 22, 2023
24
If you have it the wrong way round, it will be 0.7V, regardless of its rated voltage.
Thank you!!! This was the problem- so the found amplified zener schem was incorrect. Reverted to the normal mosfet follies version which had the zener correct. I will fix the schematic and post later this evening in case it might help anyone else. Interestingly with the new circuit I’m getting 27.5v and 17.5v using a 3.9v zener- which will work splendidly - and the mosfet is just warm so I’ll probably overdo it and install a heat sink. The LM350 is not even warm.
 
Top