| Electronics > Repair |
| HP E3610A Power Supply - Current not displaying / Voltage not changing |
| << < (5/5) |
| F4:
--- Quote from: pqass on December 21, 2024, 04:11:43 pm --- For now, we need to work from Q1,Q3 thru Q2 to U1 output. Upstream of U1 can wait until the driver transistors are confirmed working first. a. Start by confirming R3,Q1,Q3 are working by pulling-off Q2 from the PCB (watch out for the ferrite bead on the emitter leg and mark the orientation on the PCB). You should see 26V on the output terminals. If not, STOP here. --- End quote --- I had spent some of yesterday pulling the unit apart, including the back heatsync, etc, to give a clean and check Q1 and Q2. They were fine, I changed them at the beginning of the project. I pulled several resistors along the path and checked them. Managed to break CR5, so I needed to order some diodes too. It did not affect my ability to test. None of the stuff I looked at failed. I pulled Q2, which is also changed from the beginning and it still checks out. The other thing I found is somewhere along the way I mixed up the chips so put them in their right places, I thought for a while this was the root problem, but the unit behaved the same when I put it back together, U6 still heating up. There is more responsiveness from the dials now, on the bright side. Pulling Q2 resulted at 0V at the output so I stopped there, and U6 no longer heats up. I am going to test the 3 chips from a known good unit against this one again and see if mixing the order ruined that swap test last time and one of the ICs is in fact bad. You assistance is appreciated, I would have likely given up by now from frustration, and I have learned a lot sticking with it. EDIT: I made a mistake, I am pulling 12v at the output on the previous test. |
| pqass:
--- Quote ---EDIT: I made a mistake, I am pulling 12v at the output on the previous test. --- End quote --- You mean, pulling Q2 off the PCB gives 12V at the output terminals? If so, it would appear that Q1,Q3 are being partially turned on but not fully turned on. See attached. This is a problem. You must correct this before worrying about other components. It could be one or both (Q1 or Q3) are bad. Yes, you've tested them but have you tried a simple switch test? See attached. If you find one power transistor works but the other doesn't, pull the bad one and move on (with testing) down the loop (from b. in my reply#17; adding back Q2, CR5 leg lifted, etc.). While one (Q1 or Q3) is removed, don't pull too much current (<0.5A) in your testing until you replace it. If both are bad, then you probably can get by with any spare beefy NPN darlington you may have on-hand. Again, don't pull too much current in your testing until they can be replaced with equivalents. Although, finding TO-3 cased darlingtons may be expensive. You may have to improvise with TO-247 packaged equivalents. Try not to be frustrated. Take baby steps from known good working parts slowly solving each problem in the loop toward U1 and R37 |
| garrettm:
I've "fixed" over 10 of these older HP/Agilent PSUs. Every one of them had oxidized contacts on the front panel push-button switches. And if there are slide switches on the backside to allow for tracking and multi-unit connection ("master"/"slave" etc.), then check them too. I had to remove all the rear switches on 4 different units. Ended up replacing with wire jumpers because new switches were too expensive (replacements would have cost more than the PSU!). The front panel switches cannot be fully disassembled (unlike older switches of the same type), so you will likely have to unsolder and ultrasonicate in a dispersant (to remove the silicone lubricant) and then ultrasonicate with a silver cleaning solution, rinse in distilled water, dry with compressed air, and then do a final spray with fresh silicone lubricant to keep the contacts from oxidizing and lower friction of the plastic parts. |
| Navigation |
| Message Index |
| Previous page |