Darell's dead power supply - LIVES!

[Darell]

Darell\'s dead power supply - LIVES!

Well, the last "dead power supply" thread inspired me to kill my own.

Mine is a stand-alone Tripp-Lite PR-10b (10A, 12VDC regulated power supply). I use it for everything from testing LED assemblies to running my ham radio equipment. The other day I noticedf it wasn't doing its thing, after working fine for three years. It has an external 4A fuse which test fine (though I shocked myself between it and the case when I reinserted it with the unit off, but plugged in - any clue?)

I'm about as bright as an indicator-grade LED when it comes to electronics, so I know there are LOTS of you out there who know far more than I can ever hope to know.

Here is all I can tell you (until you as me for other specifics):

When plugged in, but off, I measure about 4.4mv (which is lower than the reading I get by shorting my two test leads together). With the power ON, the output "jumps" to 5.5mv, and climbs by 0.5mv every second. I let it run for about a minute with no sign of that changing.

I've taken the case off, and see nothing amiss or burned up. When the power is switched on, the unit hums and vibrates just like it always did. The power is there somewhere, it just isn't coming out the leads! That inspired me to perform one more test - I hooked it up to my watt meter, and it draws zero when off, and about 20watts when switched on (with no load, obviously).

A call to Tripp Lite technical support gets me: "We don't repair them since it is cheaper to replace it."

If anybody can offer any sort of suggestion, I'm all ears.

 

[star882]

Re: Darell\'s dead power supply - LIVES!

Start at the primary of the transformer when testing.
Go closer and closer to the output step by step until the voltage disappears.
The stage just before where the voltage disappears is the culprit.
Always unplug the unit when changing fuses.
The fuse is the first item in the primary circuit so it will protect the supply even if the switch shorts.

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

Thanks Star -

That certainly makes logical sense, and I'm sure it is really the only way to find the problem. BUT, the big issue of:

I'm about as bright as an indicator-grade LED when it comes to electronics

<font size="2" face="Verdana, Arial">is still at play here, unfortunately. If I sit and stare at it long enough, I can probably figure it out.

Whatever is wrong with it is probably pretty darn simple. Anybody interested in actually owning this fine piece of electronic wonder?

 

[Saaby]

Re: Darell\'s dead power supply - LIVES!

Ok Darell...put on Monsters Inc. for Kyra, stick that Arc in the Freezer, and start staring

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

Surprise! The Arc is already in the freezer as we speak. I figure I've got plenty of time before the battery dies this evening.

Ok, I'll get get the tester back in here....

 

[snakebite]

Re: Darell\'s dead power supply - LIVES!

i bet you had an accident and opened the series pass transistor.
if you have 20somthing volts at the case of this transistor(its the big one on the heatsink)it is likely open or the lm723 is dead.
pull it out and measure the voltage at the base with the power on.use the neg output terminal for the neg of you meter.just measure the 2 sockets the transister plugs into.one should have 14v or so.if so go to radioshack and get a new 2n3055 or 3771.

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

Ok, Ok.

Two pieces of good news:

1. I'm still alive
2. There are TWO items stuck to the outside of my heatsink with "2n3055" printed on them.

What should I pull out and measure? Pretend I'm in third grade. Really, I'm not proud.

If I don't get this thing fixed, it'll end up on your doorstep (after reading your sig again)

Just took some pictures. I'll put them up next...

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

And for your viewing pleasure, here's exactly what we're dealing with. Feel free to describe any items with shape, color, position, etc.

Third grade.

http://darelldd.com/led/powersupply1.jpg
http://darelldd.com/led/powersupply2.jpg

 

[Graham]

Re: Darell\'s dead power supply - LIVES!

Hmm, looks like a pretty simple unit.

Those two silver jobbies on the big heatsink will be the power transistors, used for the regulator.
There's a chance one of those might be dead..

For starters, those 4 black cylindrical parts on the circuit board are power diodes which together make up the rectifier (the thing which gives you a mostly positive Vout rather than pure AC), and the big black cylindrical part standing vertically on the board is probably the main filter capacitor. Any one of those parts might be the problem, but kind of hard to tell.
If you have a diode tester on your meter, you might be able to check the diodes, but since they are in a rectifier circuit it can be tricky to test unless you know/understand how they are connected to each other.

Umm, can't think of any easy way to figure it out. If I was there I could probaly work it out easily enough...

Oh yeah. And I don't have to mention (do I?) that it should be turned OFF and unplugged from the mains before touching anything in there..unless, of course, you know with certainty where to put your probes..

Graham

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

Originally posted by Graham:
Oh yeah. And I don't have to mention (do I?) that it should be turned OFF and unplugged from the mains before touching anything in there..unless, of course, you know with certainty where to put your probes..

<font size="2" face="Verdana, Arial">This is, in fact, why I'm still alive. See above.

Thanks Graham - even though it doesn't really help me breath life into the thing, at least this confirms that you truly understood my 3rd grade request. Perfect!

Any way to test the power transistors safely? They're easy to get at.

 

[lux0]

Re: Darell\'s dead power supply - LIVES!

"Any way to test the power transistors safely? They're easy to get at."

use your continuity tester, ohm meter or diode check to measure the transistors.
You should see them conduct in one direction only, with about a .6V drop. If so, they're good. If they read open or short, they're bad.

 

[star882]

Re: Darell\'s dead power supply - LIVES!

Try
http://www.ibiblio.org/obp/electricCircuits/ for lessons in electronics.

 

[bmsmith]

Re: Darell\'s dead power supply - LIVES!

I recently repaired my father's 12 volt power supply that also used a single 3N3055 transistor. In my case I was a dummy and assumed that the reason the supply was dead was because the 3055 had died and I had ordered a replacement ($0.60, so not big deal). Come to find out the 4 diodes that were acting as the rectifier had gone to diode heaven. A new full wave rectifier (instead of 4 separate diodes) fixed the problem.

In your case, and I'm not 100% positive since I can't see where the AC power is going into your unit, it looks like the transformer is a standard step-down type (120 V AC --> 13.5 V AC or so). Then the output is put to the 4 diodes where it is changed to DC and then sent to the regulator circuit which contains the 3055's. If you look at the circuit board where the diodes are (the 4 black cylinders in between the red wires in your first photo), you should see that the 2 red wires (the output of the transformer) connect to the diodes.
You can check the AC output of the transformer by using your multimeter set to AC Voltage. You should see something around 12-14 Volts AC when the unit is powered on. I tried to make an ASCII circuit diagram, but this BBS software compresses spaces and uses a variable width font, so I cannot.

If you see voltage on the red wires, then the next step is to see if the diodes are shot. So you'd need to switch your meter to DC Voltage and put the meter's leads on the diode leads that are NOT connected to the red wires. It does not matter if you connect it backwards; your meter will just say -12 Volts instead of 12 Volts (or if your meter is analog, the needle will try to go backwards, but it's nothing terrible if it happens. Just switch the meter's leads). If the meter doesn't read 12 Volts DC, then most likely one or more of the diodes are dead.

Of course you should be careful, but don't be too afraid of it. It's only 12 volts where you are testing. Use alligator clips (with rubber shielding) or hook clips to clip onto the diode leads and you are safe. Since you are measuring voltage with your meter, you cannot hurt the meter or the circuit by clipping your meter leads to the wrong component leads. It'll just give you strange readings.

I hope this helps.

- Brian

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

Hey Brian -

That helps a bunch, sounds easy enough, and even makes sense to me! All good things. I'm not scared of this stuff at all. I know how to keep safe. This is one of those areas where I may be a danger to the device, but I'm never a danger to myself.

Stand by.

 

[star882]

Re: Darell\'s dead power supply - LIVES!

Try http://repairfaq.org/REPAIR/F_smpsfaq.html for some info on repair.

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

Preliminary Brian report -

I have 34VAC between the two red wires, and 17VAC between the black wire and either of the red wires. Doesn't make sense to me. Make sense to you?

The 110V does indeed go directly into the transformer after only passing through the fuse and the switch. The Tripp Lite tech guys are actually supposed to be FAXing me the schematic, simple as it is. Heck, a photo of this thing is about as good as a schematic.

 

[star882]

Re: Darell\'s dead power supply - LIVES!

Here are some common causes for a dead power supply:
bad startup circuit (open startup resistors)
open fusable resistors (due to shorted semiconductors)
bad controller components
open semiconductors
not plugged in(don't laugh, some experts have been screwed by this)
open transformer
open connection
shutdown activated(overvoltage/overcurrent or faulty protection circuit)

 

[bmsmith]

Re: Darell\'s dead power supply - LIVES!

Darell,

Interesting. Honestly I missed the black wire coming out of the transformer in your picture. I wasn't looking or expecting it. I'm not sure how or why Tripp Lite is using a center-tapped transformer. The schematic would help.

But the same methodology should work for testing the diodes, provided they are wired as a full wave rectifier. Can you see the bottom of the circuit board to tell how the diodes are wired together?

Where are the 3 transformer wires connecting? The 2 red ones look like they connect to the diodes, but I can't really tell about the black wire.

- Brian

 

[bmsmith]

Re: Darell\'s dead power supply - LIVES!

Ok after some more research (google) I think the center tapped transformer allows Tripp Lite to combine the outputs to double the current capacity. Each winding provides half the current needed.

Given that assumption about how it's being used, then my theory still holds. The 4 diodes are providing full-wave recification to DC, and the large capacitor on the left serves to smooth out the DC voltage ripple that is a side effect of AC to DC rectification.

Here is my best ASCII I can do on this board for a schematic:

|----1----D1----2----D2----3----D3----4----D4----|
|--------------------------------------------------|

D1, D2, D3, D4 are the diodes. D1 connects to D2 which then connects to D3 which then connects to D4 which then loops back to connect back to D1. It's like a circle.
The points 1, 2, 3, and 4 are connected as follows:
1 = AC in (from transformer)
2 = DC out (to regulator circuit)
3 = AC in (from transformer)
4 = DC out (to regulator circuit)

You've already verified that you see 17 Volts between points 1 and 3. Now see if there is 17 Volts DC on points 2 and 4. The easiest way to do that is to measure the voltage across the capactitor, which should be connected to points 2 and 4.

- Brian

 

[Darell]

Re: Darell\'s dead power supply - LIVES!

OK, just for accuracy, I have TWO of those large capacitors. The top one just hides the one below it.

Next, the diodes are paired. By that I mean that one red wire is traced to the input of the first two diodes, and the other red wire is traced to the second pair. 1&2 are ganged as are 3&4.