Low DC Voltage Output On U-Frame Power Supply

[partyanimallighting]

Hi wizards! Me again. With another power supply problem. I've been very quiet for a while because of the pandemic but I recently started tinkering again so here's my new problem. I have two identical power supplies that have two different problems. The power supplies are 90~130VAC input and two DC outputs, one 390VDC to an HID lamp ballast and one 24VDC to drive other electronics. These power supplies are very old but, having nothing but time on my hands, I decided to change out all the main components (fast recovery diode, MOSFETS etc.) These units were dead before but after changing out these components, they sprung back to life. HOWEVER, I'm getting a low reading on the 390VDC side, fluctuating around 345~355VDC (the 24VDC leg is fine there, no fluctuations). I'm wondering if the fluctuations would be caused by the main capacitor (450V 100mf) being weak. I pulled the cap in question and tested it and it read around 90mf which I think is a little low. Would this cap cause the fluctuating low voltage? I'd be grateful for any advice which will help me to add to my limited knowledge base.

 

[Harald Kapp]

I pulled the cap in question and tested it and it read around 90mf which I think is a little low.

I assume this is supposed to be a 100 µF (not mF!) capacitor, so 90 µF is still within tolerances (usually -20 %/+ 50 % or so). However, ESR may be too high. You'll need an instrument capable of measuring the ESR to verify that. Or simply tray a new capacitor.
Another issue could be that the 380 V output needs a minimum load for stable regulation. Some SMPS will not generate a stable output without load. What happens when you attach the lamp as intended to this output?

 

[partyanimallighting]

Hi Harald. Thanks for the input. Firstly, let me first say that I just noticed the μ symbol in the header above but I used mf to represent microfarads, my apologies. Based on what I've learnt here in previous posts I know that the 90 μf reading is OK but as far as measuring the ESR, I think that will not be a possibility, instrument wise as I don't have anything available. I'm in a small island in the West Indies so I have not been able to source a new 100 μf cap so I'll have to order a couple online and wait the few weeks for them to arrive so I'll try the PS in a unit soon to see if the lamp strikes and holds output and report back on the post. One other question....I know from previous posts that the lamp ballast outputs a very high voltage so I can't test ballast output to the lamp but can I test the 390V output from the PS to the ballast without frying my meter?

 

[Harald Kapp]

can I test the 390V output from the PS to the ballast without frying my meter?

If your meter has a suitable range (e.g. 600 V): yes.
You could also build a resistive 1:10 divider to scale down the 390 V to e.g. 39 V and attach the meter to that lower voltage. The resistors would have to be rated for the voltages present across them. A typical leaded 1/4 W resistor is good for ~200 V. Use a few of them in series and you're good.

 

[partyanimallighting]

Well I ran into another problem, albeit somewhat hilarious . I hooked up the power supply to the unit to test the lamp strike and switched on and there was a loud sparking sound like fireworks coming from the area of the ballast so I immediately switched off to check the cause. Now, here's the interesting thing. Upon checking, I saw a bit of moisture on the top of the ballast and I believe, as the unit was scrapped open and on the floor of the workshop that "maybe" the cat may have peed on the unit, totally unbeknowst to me, hence the moisture on the ballast. Mind you, he has never done this before so I was not checking for any type of moisture. So, now the ballast is fried and has to be replaced. Lesson learnt. Check all units thoroughly before switching on to test. Now, I'm not totally sure the moisture was cat pee but he seems to be the most likely culprit. I'll keep the thread updated with my next step but I'm not too sure what that will be because I do not want to blow out another ballast.