# Problems with Commodore 1084S-P VDE PSU

#### rosaage

##### Newbie level 4
I have a Commodore 1084S-P VDE PSU (seperate board from the rest of the circuit) with some problems. If I connect it to 230V input it starts making a high pitched noise, measuring with a multimeter on the output pins gives me ~17V on pin 3 (C146) and 0V on the 25V and 125V rails. I have a spare working monitor with an Identical PSU, so I confirmed the monitor is working fine.
Since one psu was working I started moving over parts from the working psu to try and locate where the fault was as I was not able to meassure anything seemingly wrong. I have replaced:
Transformer T101, Input coil S102, All electrolytic caps except C146, C116 and C122. All transistors, Output diodes (D141,D142) isolator (IC101) and all red capacitors (Film?)
When I power it at lower voltages (<50V) all 3 outputs appear to charge up to the correct voltages, If I then increase the voltage it starts clicking slowly and the scope shows shorts on the output caps (C144). Increasing above ~200V it stays at 0V and the noise is high pitched.
One thing I find strange is that after letting it sit for some days and coming back to it when I power on the circuit it sounds completly normal untill my test leads touch the output caps, then it starts making the noise. Is the Multimeter load responsible for triggering something?

##### Super Moderator
Staff member
If I then increase the voltage it starts clicking slowly and the scope shows shorts on the output caps (C144).

Can you track down the clicking sound? Is it inside the caps? Inside a transformer?
It might be arcing. Over the years some component could start losing its ability to withstand 230V (peak=330). So it works okay when you run at a lower voltage, but goes faulty as you increase voltage.

#### rosaage

##### Newbie level 4
When I powered it on today it worked for some seconds and then I took the voltage up to 200V and the fuse blew. After replacing it the amp needle on my transformer goes to max right away, so one of the components has probably failed harder now. I'll probe around and see if I find where the short is coming from.

#### rosaage

##### Newbie level 4

This schematic is for a slightly different version as mine is just a power board and not one single board. MY board stops at the 1M2, 3M2, 2M2 point and brings them to a connector together with a ground pin, but the main layout and idea is the same.

I had another go at this today, have been a bit busy with school work and life recently.
I haven't managed to pinpoint the short yet, but it goes away with the main transformer desoldered (T101), I have tried with IC101 (optocoupler) and Caps C103 and C104 (left of bridge rectifier) removed (only T101 left) as those are the only components directly connected from the mains side to the lower voltage side. With the transformer removed the main cap charged fine up to 280V+ (only tested just above 200V input) and didn't immediately discharge as earlier. Tried swapping transformers but it didn't help.
I think my next step will be to remove the diodes/caps (141-143,151) and see if it runs with just the transformer, unless someone else has a good tip to try.

#### Relayer

##### Full Member level 5
Hello rosaage,
Setting your multimeter on low ohms, measure between D141 cathode to ground,
same for D142 and D143.
Did you try measuring TS152 for opens or shorts?
Please let us know how you get on.
Regards,
Relayer

#### rosaage

##### Newbie level 4
The diodes measured 10k ohm and up in circuit, starting on overload. TS152 (SCR?) seems fine, my tester reports it as a transistor with correct pinout if it would have been a transistor. I managed to find the short now, it was TS132. 0 Ohms between all 3 pins, and replacing it seems to work. But now I am stuck with 0V on all output caps. I will measure a bit around and see what I find.

#### rosaage

##### Newbie level 4
I did some probing around and noted down some of the difference between my board and the schematic. The source pin of the mosfet is directly connected to the negative on C112 "HOT EARTH" is noted on the pcb. C112 charges fine now, and does not get discharged right away.
As far as I can tell from the datasheet $$V_{GS(th)}$$ is 2.5V min so the mosfet should be off. From what I can tell This mosfet controls what goes through pin 2 to 4 in the transformer? And this mosfet is controlled by the smaller transistor TS121? The mosfet is clipped to the only heatsink on the pcb.

##### Super Moderator
Staff member
Below C132 you drew a red ground icon. Although it makes sense for a ground connection to be there, C132 absolutely must be healthy not leaky. There's a chance it provides a snubber effect, in order to substitute for the DCR arrangement D133 C133 R133 which is often installed as a snubbing network (but which you marked through with a red X which implies it's absent).

A flyback primary (T101) commonly generates high voltage spikes at shutoff. To absorb these spikes a snubber is normally installed. However if snubbing is inadequate then the spikes can destroy neighboring components. Example, C132 or TS132 where you reported abnormal volt readings.

This makes a reason to examine C132 because if it's not healthy it can cause problems especially as you increase power. C132 must maintain a proper Farad value, tolerance to high voltage, with no leakage meaning its DC ohm reading should be infinite.

And it may help if you install D133 C133 R133 as seen in your schematic as well as numerous other flyback schematics.

#### rosaage

##### Newbie level 4
Below C132 you drew a red ground icon. Although it makes sense for a ground connection to be there, C132 absolutely must be healthy not leaky. There's a chance it provides a snubber effect, in order to substitute for the DCR arrangement D133 C133 R133 which is often installed as a snubbing network (but which you marked through with a red X which implies it's absent).

A flyback primary (T101) commonly generates high voltage spikes at shutoff. To absorb these spikes a snubber is normally installed. However if snubbing is inadequate then the spikes can destroy neighboring components. Example, C132 or TS132 where you reported abnormal volt readings.

This makes a reason to examine C132 because if it's not healthy it can cause problems especially as you increase power. C132 must maintain a proper Farad value, tolerance to high voltage, with no leakage meaning its DC ohm reading should be infinite.

And it may help if you install D133 C133 R133 as seen in your schematic as well as numerous other flyback schematics.
I have now replaced C132 with a new part, it has a slightly higher value (new: 80nF, old: 2.2nF). I tried running it directly on mains now (~230V) and the main cap charged to just above 300V, the gate pin on TS132 is now ~1.5V.

I managed to find some better documents as well, the Philips PM8833 is apparently the same monitor and the manual for that one was a lot clearer, it also looks closer to my monitor than the old schematic.

##### Super Moderator
Staff member
I tried running it directly on mains now (~230V) and the main cap charged to just above 300V.

Do you still get the clicking noise reported in your initial post?

If you made progress then it's time to check your output voltages. If they are not normal then the supply isn't fixed, and you should avoid running it at full house voltage longer than the time it takes to test voltages.

the gate pin on TS132 is now ~1.5V

The schematic shows -4.1. Since that section of circuitry has high supply voltage, it's hard to be sure whether the readings of low voltage are an average due to switching on-&-off, and whether 1.5V is close enough to call it normal or whether it ought to be closer to -4.1V.

#### rosaage

##### Newbie level 4
Do you still get the clicking noise reported in your initial post?

If you made progress then it's time to check your output voltages. If they are not normal then the supply isn't fixed, and you should avoid running it at full house voltage longer than the time it takes to test voltages.

The schematic shows -4.1. Since that section of circuitry has high supply voltage, it's hard to be sure whether the readings of low voltage are an average due to switching on-&-off, and whether 1.5V is close enough to call it normal or whether it ought to be closer to -4.1V.
Seems like I forgot to include that part,
The clicking is gone now, but all 3 output caps measure 0V (The clicking came as the one of the output caps was shorted, so I assume this might be related to why it is gone now), Yes I have only powered it for a few seconds at a time to measure certain points, and then I drain the main cap with a 100Ohm resistor.
I have tried measuring directly on the transformer (before the diodes), but both AC and DC mode shows 0V as well as the scope showing no voltage.

My multimeter (HP 34401) shows 0Hz when I measure the the frequency on the Gain pin, I'm pretty sure this means it is 1.5V dc but I can try hooking up the scope and take a good look at the signal there.

I can try replacing all signal diodes (I have 4148 diodes on hand) and measure the zeners and see If that helps. I think they are the only parts now that I haven't tried replacing yet. I have BC547 as well I can try to swap, and I'll take a look if I have anything higher power to replace TS121 as well

##### Super Moderator
Staff member
It doesn't sound as though transformer T101 is receiving an oscillating waveform. If it were then some output voltage would appear.
It's not clear how the transistors are driven to oscillate. The windings at 6-7-8 are no doubt in the transformer. Some influence comes from the optoisolator which usually has to do with voltage regulation. However optoisolators have a tendency to go bad over time.

You may need to devise a way to give oscillations a 'kick' to get them going. Or you may need to build a circuit to generate oscillations.

I can't say whether your new value for C132 (forty times the original value 2.2nF) enhances operation or not.