• Please review our updated Terms and Rules here

DEC H7441 5V regulator repair

nater1217

New Member
Joined
Aug 9, 2022
Messages
9
Hello,
I'm hoping to receive some advice about repairing my PDP 11/34a's 5V power supply. I've made some progress, but I'm still having issues.
When I received the machine, the power supply's fuse would blow immediately. I traced this to a shorted D12 and Q2.
I replaced D12 with a ues2605r, the parts list calls for a ues602r. From what I can tell, the replacement has a higher PIV at 200v instead of 100v, and a slightly higher forward voltage at 1.25 vs 930mv of the original (assuming UES602R == UES2602R). Would this be an issue?
I replace Q1 and Q2 with NOS ASI 2N5302 transistors. I also replaced output capacitors C11 and C16. I, replaced input cap C1 with the cap from a working supply.
This solved the fuse blowing, and the power supply worked with no load, and with a 1ohm load on it.

However, with a .33 ohm load, the voltage would start slowly dropping. I turn it off before it gets below 3.9V.
This seems to be related to heat, as the starting voltage will only start near 5V after the supply has been turned off for a while. Knowing this, I started looking for components that were hot. I found Q5, which wasn't hot, but was getting warm. To test, I sprayed Q5 with a can of compressed air upside down to cool it down, and immediately the voltage went back up. So, I ordered a NOS D44C12 to replace it.

This did not solve my problem, unless the new part is also bad. It seems to behave the same way as before. When I spray it, the voltage comes back up.

Any advice? I'm not familiar with power supplies, so I don't know where to go from here. I have an oscilloscope, so I can check things if anyone has suggestions. I also have a working H7441 to compare to.

Also, here is a link to the schematic on page 9 in FMPS MP00271. H7441 Field maintenance

Thanks
 
What kind of temperature on Q5 are we talking about?

Sometimes it is hard to target one single component by spraying it. It might be some other component close to it that is sensitive.

There is a voltage regulator circuit Q4, D2 and R1. Check the voltage on the emitter of Q4. Is it constant or change over time?

Check pin 3 on the NE555 with the scope. Does the signal remain the same or change somehow?
 
What kind of temperature on Q5 are we talking about?

Sometimes it is hard to target one single component by spraying it. It might be some other component close to it that is sensitive.

There is a voltage regulator circuit Q4, D2 and R1. Check the voltage on the emitter of Q4. Is it constant or change over time?

Check pin 3 on the NE555 with the scope. Does the signal remain the same or change somehow?
I measured 54 C on Q5 when I turned it off at 3.9V output. The working supply's Q5 does not get hot even with a .33 ohm load.
Both resistors that are close to it measure good. Well, .3 ohm instead of .25 ohm for R5. The good supply also measures that way.

Q4 emitter starts at 11.6 V then gradually rises to 11.75 V. That stays the same loaded and unloaded. The good supply measures 12.5 V

Pin 3 on E1 is 18.2 kHz and 27.4% duty cycle. Seems to stay the same. Looks the same as the good supply.

Pin 3 on E2 is 640 kHz. The duty cycle falls as the voltage drops. The signal looks the same as the good supply.
 
Difficult one. The lower voltage on Q4 emitter may indicate higher regulated voltage consumption. But there are variations on the zener diode voltage.

To be honest I am not sure why it has two 555 chips there. 640kHz switching frequency sounds very high for a PSU this old. Maybe you could measure the signal on the base of Q5 and see if that differs from the working one. Also check the signal on Q5 collector and compare it with the working one.

It seems like the bad supply consumes to much power driving the switch transistors. Check all the diodes D4-D8 if they are OK.
 
Yes. Indeed quite complex.

I am just thinking out loud now: if the Q5 transistor is not driven into saturation it might get hot. Then it will not be able to drive the following stages properly either perhaps causing a lower output. The error amplifier will then cause the duty cycle to change to try to increase the output.

Check and compare levels on the base of Q5 and the small signal diodes around it. Check signal on Q6.

I have seen leaky SCRs causing problems. Check the crowbar circuit. D21, D22. Perhaps temporarily remove crowbar just to check.
 
Thanks for the replies. I just got a chance to work on this tonight.
Check all the diodes D4-D8 if they are OK
I was checking Diode D7 and if I'm remembering correctly, it was measuring a pretty low voltage on diode test mode. I wanted to remove it to test it out of circuit. I put a little pressure on the lead to pick it up while heating the solder and the diode shattered. I hope I'm remembering correctly that it was that diode that measure weird and I found an issue. All the other diodes seem to test okay.
I didn't have a chance to check anything else, so now I am waiting for parts again.

A114B's aren't easily available, so I was going to get a modern part and replace both D7 and D8. Would something like this be okay? 1N4935-T
I'm not sure how critical all the specs are.

Did you happen to see Jerry Walker's youtube series on the 11/34?
I did. Unfortunately for me, his supplies all worked, so he didn't go very in depth.
 
Well, I got my replacement diodes in the mail and... IT WORKS!
Thank you @MattisLind for pointing me in the right direction!

To recap incase anyone in the future has this issue:
My H7441 power supply would blow the fuse as soon as it was turned on. I found D12 and Q2 were both shorted.
I replaced D12, Q1, and Q2. I also replaced C11 and C16 since they were leaking.

That brought back 5V, but only at low load. With a heavier load, the supply voltage would drop gradually from 5V. Q5 would get warm.
I found that D7 was shorted. I replaced D7 and D8 and that fixed the supply completely.

I also replaced Q5 because I thought that that was the problem, but it wasn't.

Thanks all!
 
Back
Top