PDA

View Full Version : Testing a circuit board



cuemaker
04-16-2010, 01:40 PM
Pictured below is a circuit board to my welder..

I am told that the 2 black boxes (relays) are more than likely whats wrong with the board.

1. How do you test? What do I turn my multimeter to? what reading should I be looking for..

Thanks..

http://i201.photobucket.com/albums/aa129/xringx/DSCN2344.jpg
http://i201.photobucket.com/albums/aa129/xringx/DSCN2345.jpg

MotorradMike
04-16-2010, 02:18 PM
If you have it on good authority that the relays are subject to failure on this model then replace them. I'd start with the big one. It's contacts are rated for fairly high current and they may be burned.

It's best to learn how to troubleshoot electronics on something you don't care about IMHO.

If you aren't good at soldering you might want to get somebody else to remove the relay.

ckalley
04-16-2010, 03:11 PM
Those big white things at the top center of the board are a type of resistor that are known to fail fairly often (don't ask me how I know....).

Since it sounds like the relay(s) might be the problem, look at the top of the relay for a wiring diagram. That should show you show you what pins are connected to what inside. The coil may be represented as a rectangle or something that looks like a coil spring. The conatcts may be shown either their functions C (common) NC (normally open) NO (normally closed) or as symbols - 2 parallel lines with nothing between them (NO) or 2 parrallel lines with a diagonal line between them (NC).

If you can identify the coil, set your ohm meter on OHMS and measure across those 2 pins. You should see some kind of resistance probably some where between 10 and 1000 ohms. If you get and open (no change in indication) or a dead short (0 ohms), replace the relay.

You can also get an idea if the relay contacts are burnt or welded together with the ohm meter on the same setting. measure between one of the common terminals and the NC terminal - you should see 0 ohms. Then measure between the common terminal and the NO terminal - you should see an open circuit (no change in reading). If you don't get these readings, replace the relay.

Craig

cuemaker
04-16-2010, 04:35 PM
If you have it on good authority that the relays are subject to failure on this model then replace them. I'd start with the big one. It's contacts are rated for fairly high current and they may be burned.

It's best to learn how to troubleshoot electronics on something you don't care about IMHO.

If you aren't good at soldering you might want to get somebody else to remove the relay.

I already have a new board on the way... This is for education and profit!

lynnl
04-16-2010, 04:35 PM
What is the little orange thingy at PCT2?

I have a board from a depth finder on which a similar looking item has broken away at one leg due to corrosion.

They kind of remind me humpty dumpty, except round rather than egg shaped, with two legs soldered to the board.

Fasttrack
04-16-2010, 04:58 PM
What is the little orange thingy at PCT2?




That looks to be an epoxy dipped tantalum capacitor. Got a big box full of 'em here - .01 mfd to 10 mfd.

edit: (Also known as "conformally coated" tantalum capacitors)

Bruce Griffing
04-16-2010, 05:05 PM
Remove the relay to test it. Look up the part number to find the contact arrangement. Hook up a couple of drill batteries or car batteries or whatever to get a 24 volt dc source. connect that to the relay coils. You should be able to test the contacts. The normally open contacts should close (show zero resistance) and the normally open contacts should open (go from zero to infinite). To do this you will need some wire and a multimeter, as well as your voltage source(s).

MotorradMike
04-16-2010, 07:09 PM
I already have a new board on the way... This is for education and profit!

Well then. When you get the new board, compare the good large relay measurements to the bad board prior to installation.
Looking at the bottom side you can see it has 5 pins. 3 have larger pads and are to the left(in your pic). 2 have smaller pads to the right. These 2 pads are coil and are probably OK. Expect 660 Ohms.
Then compare measurements on the 3 larger pads bad board to new. As ckalley has said, these readings will be 'very low' or 'very high'. Any discrepancy indicts the old relay.

Couple other things.
Yes the big white things are prone to failure because they are power resistors. These two however, have never broken a sweat. They turn circuit boards brown, melt their own solder joints, and discolour and crack the ceramic they're made of without failure at times. You may find values on the sides you can verify but maybe not. Unlikely to be the problem.

If PTC2 was a tantalum it would have a 'C' ref. des. not PTC and would also have a + sign polarity indicator. My guess(and it is a guess) would be Positive Temperature Coefficient Thermistor.

For more info on the big relay look here (http://search.digikey.com/scripts/DkSearch/dksus.dll?vendor=0&keywords=pb115-nd).

Have fun, let us know what happens.

J Tiers
04-16-2010, 08:34 PM
That looks to be an epoxy dipped tantalum capacitor. Got a big box full of 'em here - .01 mfd to 10 mfd.

edit: (Also known as "conformally coated" tantalum capacitors)


PTC = Positive Temperature Coefficient

typically a device used as a sort of circuit breaker..... resistance increases with temp (and current)

As for the relays, you test by applying coil voltage and seeing if they close with low resistance...... unless you open and inspect them which may be hard as they are enclosed.

Looks to be quite a simple little board, but teh conformal coating may make repairs a pain of large proportions.... that stuff both sticks, and falls apart into dust.... doesn't like to come off. We use it, I hate it. And it looks like they slathered the whole thing with it... even the heatsink

MotorradMike
04-16-2010, 09:24 PM
teh conformal coating may make repairs a pain of large proportions.... that stuff both sticks, and falls apart into dust.... doesn't like to come off. We use it, I hate it. And it looks like they slathered the whole thing with it... even the heatsink

Yes, I hate it too. There are lots of different kinds. Slathering the heatsink was a bad plan for sure, regardless of the kind. I'm guessing they didn't need the heatsink too much.
However, remember, we conformal coat for good reason. Just because you and I don't like to fix things that are encapsulated, doesn't mean it's a good plan not to. Conformal coat guards against all sorts of evil, just be thankful it isn't potted in epoxy!

boslab
04-16-2010, 09:59 PM
replace with a ssr solid state relay, should last indefinately but you will have to stick on a seperate sink
mark

J Tiers
04-16-2010, 10:15 PM
Yes, I hate it too. There are lots of different kinds. Slathering the heatsink was a bad plan for sure, regardless of the kind. I'm guessing they didn't need the heatsink too much.
However, remember, we conformal coat for good reason. Just because you and I don't like to fix things that are encapsulated, doesn't mean it's a good plan not to. Conformal coat guards against all sorts of evil, just be thankful it isn't potted in epoxy!

Some kinds are worse... That looks like teh Dow Corning stuff we use, which is an RTV type silicone, but thin liquid. i forget the number, we have lots of pint cans of it.

That stuff is reasonably easy to remove, but it crumbles into dust, and the last layer is hard to remove. We end up using a Dremel with a soft wire brush.

I don't do that work, other than diagnosing some field failures, but the dust is crazy.... bouncy whitish dust.... doesn't brush off the bench well, static cling like the devil, ugh.

We don't use larger cans because it goes bad once opened, and a pint covers a LOT of boards.... One board we have been making for an HVAC company ever since designing it.... lots of relays, and goes outside, under a cover, but not otherwise protected, so we have to conformal it. Pollution degree 3 plus humidity, rain, snow, etc....

The stuff does work..... but.....

doctor demo
04-16-2010, 11:08 PM
we have to conformal it.

The stuff does work.....
I thought that was to keep the magic smoke in:D

Steve

J Tiers
04-17-2010, 12:31 AM
I thought that was to keep the magic smoke in:D

Steve

Actually, it can only keep it "local", but not "in"..... sorry.

And, when applied to a heatsink, it is likely to actually be a sort of solvent for magic smoke, allowing it to slide out more easily.

Paul Alciatore
04-17-2010, 12:47 AM
No obvious signs of smoke. If I didn't have the schematic I would just try replacing parts one at a time until it worked. The most likely ones first. If your info on the relays going bad is good, then they are a good first choice. The transistor on the heat sync would be another candidate. Then perhaps the electrolytic capacitors, the large power resistors, the tantalum capacitor, and the IC. Your choice of the order to try can also be influenced by their prices.

If you are not too sure about the relay thing, start with the electrolytics and put the relays further down the list. Relays are usually fairly reliable.

macona
04-17-2010, 05:23 PM
Miller dips all their boards in a thin RTV. Keeps the boards sealed and keeps metal filing from shorting out traces. Other companies like SanRex (Who makes the Thermal Arc inverters) do not coat their boards and I have had several cases where aluminum dust got in the control and made things go all wonky.

The big resistor are just dynamic break resistors.

The big relay is the only one you need to replace. Digikey part number PB115-ND $2.95, Buy an extra.

J Tiers
04-17-2010, 06:37 PM
The big one is like the Potter and Brumfield T-90 series.

The originals are a very tough relay, but that appears to be a "chinese copy" that does not carry the same ratings. For instance, where the T-90 is rated 30A 28VDC, that one is rated 14VDC....

So it is perfectly credible that it might fail early. Electro-mechanical stuff fails before electronics, usually.