MGE Pulsar ES4 UPS , not passing wall voltage

[cashkennedy]

I have a MGE pulsar es4 battery back up / UPS from like '93 ish. Infact I have 4 of these and I consider them VASTLY superior to anything apc has ever made for the consumer / small pc market. No large iron core transformer in these or shitty basic square waves. (maybe ill spark some fights on this)

But anyways this one that's no longer working doesn't pass 120v to the rear outlets / or charge the battery, but it can run off of battery (these units start without being plugged in / "cold start" , something APC has never mastered as far as I know...). There is a large OMRON relay in the center that is not activating, It requires 12v to activate. There is a small transformer that appears to create the 12v then go into part 3 (a diode) and then part 2 (a fet) and then part 1 (a startup? cap) that is a 470uf 50V cap. (This unit is mostly jamicon caps, with a few UCC) the area I labeled 4 is where I believe 4 diodes are creating a bridge rectifier.

When I measure the voltage after the cap I get ~ 11.4 vDC to the wall neutral at first, but over time that degraded down to 10vDC and kept decreasing. I also get a buzzing from the small transformer.

Would yall replace the small 50v cap between the transformer and the relay? Or is the transformer more likely to be going bad? Is measuring the voltage vs the wall neutral even the proper way to determine the voltage reaching the relay's coil?

[Khron]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

That area looks like the "5v stand-by" that you find in ATX power supplies. U008 is the controller chip, and it's driving the FET right below it, next to that white ceramic power resistor.

First step i'd take would be replacing the small cap right next to U008 - that is the actual start-up cap.

My guess is, the part you marked as "2" (which is actually a bipolar transistor, going by the BCE markings on the silkscreen) is acting like a discrete linear regulator, since i don't see any optocoupler near that transformer (that would give a feedback voltage to U008).

Linear regulators need the input voltage to be higher than the output - in this case, 12v. If the PWM that's supplying the input voltage to that, is wonky, you get no relay triggering.

Try replacing the cap right next to U008 and see what happens If at all possible, use a low-ESR one from one of the (few) reputable brands

[cashkennedy]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

Replaced that cap with a 50v 47uF panny FR series cap.

Didn't fix the problem. I noticed that the cap I marked 2 in the picture is still getting lower in voltage. And when I unplug the unit the cap actually goes to 17vdc , (this is with no batteries / or wall plugged in) ...

Is there any way to test the small transformers output before the bipolar transistor? Or further test the "SNPS" chip is operating

[Khron]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

What voltages are you getting across the cap you just replaced?

And what markings are there, on the top of that U008 chip?

[cashkennedy]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

From what I can tell the UCC cap that is located next to the bridge diodes is the main DC filter cap.

So for a reading of the startup cap I assume I should take a DC reading of the + of startup cap vs the - of the main DC filter cap (Larger UCC cap).

For that reading I get 22 vDC for half a second then I get 17vDC stably after that.

The part number of U008 is motorola UC3842AN

[cashkennedy]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

http://www.alldatasheet.com/datashee.../UC3842AN.html Based on the data sheet I believe I need 10-15 v for VCC

Should the ouput on pin 6 be greater then 12v? Is that what you were trying to say about the bipolar transistor? because on the output pin 6, I get 1.8V , reference voltage I get 5.02 V (both are dc relative to negative of main filter cap)

Oddly when I connect the batteries to the device I get 90vdc on the coil of the relay / and on base and emitter of the biopolar transistor that I marked 2 in the original picture.

Did I mention that after replacing the startup cap the high pitched noise went away? Doesn't work though yet of course.

[Khron]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

The bipolar transistor i mentioned, is right next to the relay (the one with transparent casing), and it's on the secondary side of that transformer (ie. galvanically isolated, no DC path between them).

The output on pin 6 is actually a square wave (which you won't be able to measure with a multimeter).

The "more than 12v" should be on the cathode (silver band) end of the big fat D015 diode.

[ben7]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

Well, do you know if anything else runs off the same transformer as the relay? Arethe rectifire diodes getting warm? What about the transformer?

[cashkennedy]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

I did some further testing including forcing the relay to switch with an external 12v power supply.

Also read some French forums where people fix these all the time apparently (MGE was a French company). And the first recommendation is always to check the startup cap that Khron had me replace, or the main filter cap.

I went on to test it with the batteries again and realized that it doesn't "flip" the relay till it is turned on, I was under the impression that it would flip the relay just from being plugged in and realizing it needed to pass mains to the back and start charging from mains. But that wasn't logical because the front switch controls the back plugs obviously, and the charging circuit must not be related to the relay at all.

So perhaps the unit works now, I do get power from the back outlets now when it's on, which i'm quite sure I wasn't getting before replacing the startup cap. I will have to drain the batteries though and make sure it is recharging them though.

[Davsta]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

good information thank you from someone trying to preserve his ES4 as well

[kc8adu]

Re: MGE Pulsar ES4 UPS , not passing wall voltage

looks like many small lytics near 1w resistors.change em all.bad design to place them near hot parts.
btw most ups wont do anything without good batteries.