JDSU 2110U-SLCPEB died. "Dim" LEDs, slow fan.

Well, I was going to ask about strange behavior in one of my ArIons, but looks like this overtook it.

Posting here, to see if LSRFAQ or anyone else may have any ideas what I may be able to do to troubleshoot this.

About an hour ago, my JDSU 2110U-SLCPEB power supply was chugging along fine, running my lasers on their monthly run-in. Brought some of them up to 10A (this supply was the only one of my many that would do over 8a out), no problems.

Was doing the run on one of my multilines, got called in for dinner. Set the power to idle for a few minutes, then turned the tube current off, turned the interlock off, and unplugged the power supply.

Came back an hour later, and now, while the red interlock light is bright red , the green interlock LED is dim; the fan turns slowly, the JDSU interface is lit up dimly and the numeric display on the interface reads "7.89" even if tube current is set to off.

Needless to say it won't start any lasers anymore.

The interface is fine, as it works with all of my other JDSU 2110 supplies (the 'gold box' ones); lights up brighter and no problems at all.

There's no 220/110 switch on this; it uses the 20a cord style, [-_-] and for my ham radio stuff out here, I *do* have 20a service and 20a outlets. So current draw shouldn't be the problem.

Just not sure why it was running happily an hour ago, and sometime between unplug and plug-back-in, now its all dim/slow fan/no start. Since nobody on earth seems to have another one of these (except on ebay, from sellers that are charging $1500 for EOL tubes as well) - I was wondering if anyone had any idea where to even start to troubleshoot this and hopefully repair it. I spoke to Dave, who I got it from originally, and he says it's beyond his expertise. (and, he doesn't have any more...)

Any info at all would be greatly appreciated...

Well, have you opened it up? Some pictures would be helpful

_usually_ the fans in these power supplies run directly from mains power. So I'm not sure if the fan running slowly would be the result of the problem, it could even be the problem itself/ But like I said, video's and pictures would be good, if you can.

Check voltage everywhere that you know what the voltage should be.

And no storms, power bumps or surges during the time that you can speak of?

How many fuses on the input side? If both legs are fused and one blew, you could be drawing half voltage in and returning through ground?

If the LEDs are dim, I would definitely agree with checking the voltages, I'd guess you have a 5V supply in there to power all the logic and control circuitry, that's a good place to start. If it's a switchmode power supply then electrolytic capacitors are a prime suspect. Look for any that have bulging tops or leaking electrolyte. Look for the filter cap on the 5V supply, it'll probably be a few hundred to a couple thousand uF at 10V or so.

Thanks guys, I'll look into it some more here, and maybe get some pics though it's pretty much layer-upon-layer.

It is acting like exactly that: like it's getting 60v instead of 120v.

I suppose it's possible that there was some kind of surge while I was out of the 'lab' for an hour, but the skies were clear at the time, and none of my UPSes or anything showed anything funny.

-Each leg of the mains IS fused, and that's what I figured it might be, so I checked continuity on the two primary fuses and they were both fine...

-The fan only turns on when I select 'tube current on', on the control panel, so there is some logic behind it.

As far as checking voltage where I know voltage should be, a bit tough since I'm sure no schematics exist for this in any publically accessible way.

I'll check the logic board voltage here shortly and see what I get.. and take some pics in the meantime. This one is more complex than all of my other JDSU supplies (and smaller and lighter -- switching supply maybe.)

Also the only one with a 20A mains cord.

Look for one or two large high value electrolytic capacitors near the input. These will be the bulk filter caps after the rectifier. Measure the voltage across those, it's possible that you have a bad rectifier or a bad connection somewhere.

Guys -- I want to thank you -- as you'll see, sometimes are "simpler than they need to be", but your comments, about half the power returning through ground, resulting in half power (slow fan, etc) is exactly what the behavior looked like.

So I was thinking about this overnight, and logically I figured, how could it fail like this so suddenly. Well, time to take it apart:



Here's the PSU, 2110-SLCPEB before opening it up. As you can see it's a bit smaller, thinner (and thankfully lighter) than most of the normal 2110 "gold brick" (and one black) JDSU Argon Ion PSUs I have. It also uses a 20a plug, versus the normal 15a 110vac mains cord.



So, opened up, you can see a power board, with what appears (to me) to be fuses and toroids, where the power first passed through. Last night, I checked the continuity of these fuses, and checked the connections (even though 'tab1' in the lower right looks loose there) - with no change in behavior.



Removed that power (filtering? maybe someone knows betteR) board; checked all of those components. They all seemed pristine, and at the very least, had continuity. I especially wanted to check that bank of what looks like they're either ceramic caps or MOVs (not sure) - to make sure they weren't open.



Then, to the side, I removed a board which was labeled "Current Loop". I worried that this may be suspect.. So I checked the readings across all of the resistors, the diodes for continuity, etc. Since I wasn't sure what the ICs did, I didn't monkey with those, but all of the passives seemed fine.



Here's the main board. I expected to see a charred part or something here, or a bulging or leaking cap. No such thing. Checked the large 470uF capacitors, they seemed ok; Discharged, but would taper off after a charge, so it looked like they were charging and discharging correctly.



Now here's when things got a little bit "huh.." -- This is the front panel board. It pretty much just acts as a 'jumper' as far as power is concerned, but, right away I noticed that the green and yellow wire (with blue plug caps) - the end going to the board, was barely on. I figured it was a long shot, so I made sure it was on there firmly, and reassembled. Could that have been all it was?



Hot damn... The LEDs aren't dim.. The fan's running at full speed. The interface is fully lit up and not displaying "7.89" for some reason, but the normal '0.04' it would show (filament current maybe?) when tube current wasn't on...

So.. next... the moment of truth...

... pressing 'TUBE CURRENT -- ON' and waiting...

... And waiting ...

... Nothing ...

... I guess that ground jumper wasn't the pr -- OH WAIT.. Notice in the picture above, upper left... No interlock key! ... Whoops... helps to have that.. Got too excited.

... Insert key ...
... Turn key ...
... TUBE CURRENT -- ON ...

... waiting ...

... BAM!



4A idle current. Tube lights. Hot damn! Bring up the power?



Don't mind if I do!



Nice. A visible beam in daylight. Multiline, at that.


First time I've been able to run this multiline at its full potential and...





What a potential it is. 115mW of multiline. Sweet. Clean beam, no artifacts, solid and steady. (Note, this isn't the 'strange' laser that was peaking and dropping that I posted in the other thread... This one is rock solid 113-115mW ~ 10a).

Anyways, thanks guys for at least getting me in the right troubleshooting mindset for this. Immediately often what I do is jump to 'worst case', and sometimes it really is simething simple, easily explained by observed behavior.

Hopefully other folks can take that advice too when something stops working -- stop and think, listen to experienced folks around here, and take the next logical steps.

It's nice when things work out. Thanks guys. And +'s.

Glad you got it working!

The 0.04 is probably just a minor calibration error. The filament current is way more than 40mA. More like 25A if I remember correctly

I thought it measured the tube current, not filament current?

Anyway, glad you got it working with minimal difficulty! Switchmode power supplies are possibly one of the hardest things to troubleshoot, especially when all the boards are packed in like that! (I have a slightly different model)

Kind of interesting how a jumper could just wiggle it's way loose.

Oh, well, yeah it probably doesn't (measure filament current that is) - I was just taking a wild guess on why it read 0.04 without current on. I'm sure Cyparagon is right that it's just a calibration mismatch.

Hey awesome job getting that going! It feels good to have poked your way through the problem until you've found that its only a simple issue. With a loose wire there, it could've been a little more disastrous! A loose connection can get hot enough to melt insulation, especially at tube level currents.