SMT Diode id & home automation system repair

[(*steve*)]

Yeah, there looks like a ground-plane very close around the power connections. Is there any chance you've bridges one of the power connections to it?

 

[mypcbis$crewed]

Oooh,. sorry, been a while since I read through this thread. (I guess I should've looked up that part number.)
No problem Mitch

A TVS diode, treat it almost like a zener. That guy should read as an open until you apply >= 15V to it. Then it will start to conduct. What kind of voltage are we talking about on those pins?
Edit: And hey, those joints don't look bad at all for a cheap iron! Nicely done.

That's whats weird the guys at Vishay were able to tell me that the part that burnt out was a smbj15A but I was a little hesitant to believe that because there was no polarity markings and the diode when still attached and somewhat functioning (even though it was puffed) had resistance in one direction and an unequal resistance in the other direction. Keep in mind the other three diodes on the board with identical markings had equal readings forward or reverse! I could only conclude that this part was a bidirectional diode.The closest thing I could get my hands on was a sa15ca

And it's not exactly two diodes in series, reverse of eachother... it's two zener diodes in series, reverse of eachother. While one is forward biased, and passes current normally, the other goes into avalanche mode at a certain voltage and passes current.

Thanks for the explanation, I was assuming that the circuitry inside the diode were both equal and that is why I couldn't understand how that would work. In this case with the diode being across the 12-13v power supply entrance to the board how would this diode protect the circuitry?

Resqueline, the most distant ground plane I could find measures 0.3 ohms between negative lead. Between a positive spot on the board where the old diode was and the red lead is the same, 0.3 ohms.

Steve, I checked again with the sun shining on the part to really see it well, and I don't see any solder traces bridging over to the ground plane. I checked resistance as well and the positive lead is still shielded from the negative side, no continuity.

At this point I am thinking about just desoldering the part to verify that the unit still works.

ok, very weird..... I was doing some more testing on the board while powered with a 9 volt battery....the soldered power connection receives power and so do the various rails on the board. When I checked for power to the little positive spot (see pic below - small silver spot between the blackened lifted pads)where the diode was with the red lead from my multimeter and the black vom lead on the negative power rail, the unit squeaked to life!

Resqueline, you had asked me about the voltages on the 78m05, they are input 7.96 and output 4.96 using 9v battery.

 

[Resqueline]

The damaged pad area might be a bridge carrying power to the circuits. Try to use a powerful magnifying glass to see where there might be a break.
The power might come up through a hole from the underside, and might be supposed go back down through another hole after passing by the pad.
Solder in a small link wire (anywhere) to connect these two tracks if you find a potential break.
Resistance measurements pos/neg seems just as would be expected.
Ok, the 78M05 is obviously unharmed.

 

[55pilot]

I have not read this whole saga, so I may be completely off here....

I am assuming that it has already been determined that the blown D5 is a transorb. For D5 to blow up like that, there had to have been a lot of current flow. Where did that current travel and what else did it destroy? Could it be that when you touched the spot with the fat probe, you make a connection that was previously open? You need to clean up the area and carefully inspect it under a 5x or higher magnifying glass to see what got damaged.

---55p

 

[Mitchekj]

Thanks for the explanation, I was assuming that the circuitry inside the diode were both equal and that is why I couldn't understand how that would work. In this case with the diode being across the 12-13v power supply entrance to the board how would this diode protect the circuitry?

This protects the input by "absorbing" (shunting) the excess power when anything over 15V is applied... since the unit is designed to be run at 12V, anything higher will damage it. Spikes and such... a tranzorb is a good high-end TVS, very quick. Does it's job well, usually.

So in summary the diode is there, doing nothing, when all is well. When a spike or overvoltage comes along, he shunts the excess and limits the input voltage to 15V, keeping everything else safe.

 

[(*steve*)]

I would definitely go with Resqueline's advice. The board is *very likely* to have one (and possibly both) power leads going via the ex-transorb before connecting to anything else.

You mentioned that the old component came off with some o the tracks. In retrospect you should have been asked to check that nothing had been damaged.

I recommend that you clean up the board around the old transorb location then repair the track(s).

When you touched the board with your meter probe it probably flattened a bit of copper that was pulled up, but the circuit, even if working now will suffer from intermittent failure unless you repair it.

 

[mypcbis$crewed]

This protects the input by "absorbing" (shunting) the excess power when anything over 15V is applied... since the unit is designed to be run at 12V, anything higher will damage it. Spikes and such... a tranzorb is a good high-end TVS, very quick. Does it's job well, usually.

So in summary the diode is there, doing nothing, when all is well. When a spike or overvoltage comes along, he shunts the excess and limits the input voltage to 15V, keeping everything else safe.

Awesome, understood TY for the good info.

 

[mypcbis$crewed]

Spot on Resqueline! The positive side does indeed have a trace running under the board and terminates in a hole that goes up through the board to that silver spot that was left behind. The spot coincides with the backside of the burnt out diode as well so the trace must be somewhat loose. What is the best method for cleaning up the burnt out areas for inspection/fixing? My current magnifier is one that you get with the screws for glasses so its not the greatest Thanks all!

When I thought on its some more....the diode thing still has me confused...why do the other diodes on the board (the smbj15ca's) have a forward and reverse reading on the vom diode check while the sa15ca does not???

 

[(*steve*)]

It's possible that the SA15CA is a bidirectional device and the SMBJ15CA is unidirectional. Hmmm, the "C" indicates bidirectional in both cases.

It is possible that if power was applies in the reverse direction that *all* of the reverse biased parts could have gone short circuit, but it's not something I'd bet on.

Is it possible that the surface mount components don't have that "C" on them?

 

[mypcbis$crewed]

It's possible that the SA15CA is a bidirectional device and the SMBJ15CA is unidirectional. Hmmm, the "C" indicates bidirectional in both cases.

It is possible that if power was applies in the reverse direction that *all* of the reverse biased parts could have gone short circuit, but it's not something I'd bet on.

Is it possible that the surface mount components don't have that "C" on them?

The surface mount components have hardly any lettering on them, however, they have the same forward and reverse readings. This is not a fluke because all 4 diodes had the same readings plus or minus 1 ohm. IIRC the reading was around 7400 ohm (or whatever figure the VOM uses in diode check mode). That is why I was certain that they were bidirectional. The Vishay engineer that I gave the part markings to thought it was unidirectional as well, but they had a hard time id'ing the part because it was from a company they took over General Semiconductors. This is kinda why I was betting on it being bidirectional.

At this point I will wait until I receive instruction on how to properly clean up the trace and repair it. I am guessing using a scotchbrite pad or something, but I dont want to make it worse

BTW, where in God's creation is somewhere between China and the Antarctic?

 

[Mitchekj]

Hehe, Australia, no?

 

[mypcbis$crewed]

Good point, I was thinking somewhere remote like malaysia or the phillipines.

 

[(*steve*)]

Good point, I was thinking somewhere remote like malaysia or the phillipines.

More remote that Malaysia or the Philipines.

In fact, you can only say we're not "remote" if you're from Honolulu.

Back on topic -- I'm really surprised that the other equivalent components read such a low resistance. If they are what we think they are, then they should appear like the new one (I guess that's stating the bleeding obvious). Where are the others? are they all across the same supply rail, or different ones? Or are they attached across something else (like inputs or even outputs)?

As for cleaning the board, Resqueline may have better suggestions, but I'd go for isopropyl alcohol and an old toothbrush. You're unlikely to destroy anything with that .

Once you have cleaned off the muck, repair the tracks by soldering a wire bridge over it. You could expose all the copper and solder to the track, or simply bridge the break as best you can. For relatively low voltage and current devices you don't have to take any special precautions.

 

[mypcbis$crewed]

D4, D3, D5 the three diodes closest to the pin/wiring harness connector show the following on VOM set to 20k resistance check:

D4
6.09k forward 6.09k reverse

D3
7.15k forward 7.15k reverse

D5 - burnt diode
5.20k forward 2.24k reverse (of course d5 is no longer there and now reads straight continuity backwards and forwards.)


All the diodes have the same marking of BM 45 (and apparently they are bi directional).

The left side closest to the edge of the board on D4 and D3 have continuity with the negative input from power supply. The right hand side, D4 goes to the green lead and D3 goes to the yellow lead of the power connection. These leads go back to the main control board and are as far as I can remember 2-way data communication pathways for the controller.
The input pin of the voltage regulator below c4 (left leg) has continuity with the right side of D2 (cathode side) while the left leg of D2 has continuity with the + dot where D5 used to be. But for whatever reason the anode of D2 does not have continuity with the positive battery lead?? That makes no sense to me unless that is where the break in the circuitry is? D2 is protecting the voltage regulator so the anode must be connected to the + terminal.....how else can it allow electrons to flow through it?

 

[(*steve*)]

D5 is straight continuity, or an open circuit?

Also, remember that you are measuring these in-circuit. You are most likely measuring the resistance of the rest of the circuit. To accurately measure these things, you'd have to de-solder at least one end. I'd suspect it's not worth it.

 

[mypcbis$crewed]

True. To clarify, the d5 diode registered continuity off the board now (toast)

 

[Resqueline]

D3 & D4 is on data i/o lines while D5 is on power line, so D5 would measure different from them when in-circuit.

Cleaning carbon & muck is best done with rubbing alcohol & a toothbrush like steve says. If you need to get down to bare copper to solder on then I often scrape with a jewellers screwdriver, or use a fiberglass eraser if it only needs brightening. Scotchbrite can be used only if the tracks sits firmly on the board, otherwise it will easily catch on their edges and rip them (more) off.

 

[mypcbis$crewed]

SUCCESS!!

You guys are awesome. I was able to clean up the pad area a little bit, but the trace lifted, which was no problem. I inserted a 22 gauge wire from the opposite side (there was a positive trace connection with a drilled lead hole) and soldered there and on the lifted trace. Tested for continuity and reinstalled into system. The system works! I put my VOM in series with the external horn and triggered the system (the vom is fused on the 2A scale) and it drew 137mA. So all go at this point. Again BIG thanks to Resqueline and everyone else that helped me for months!

The only mystery that is still unresolved is what happened and will it happen again? Resqueline has already provided some theories and the only possibility that I could think of that I think Resqueline mentioned was if one of the battery leads grounded out. I don't think this happened because I believe that both leads were connected at the time. But if memory serves me wrong: If the positive lead was off and happened to ground out - What sort of damage would we see?
Thanks again all

 

[Resqueline]

Congratulations! Very happy to hear it turned out well!
If the battery lead grounded out it could (possibly) lead to thermal overload of the regulator (even if it's supposed to be protected) whith the regulator shorting out and delivering full voltage to the circuits, burning diodes & tracks.
A wiring short at the remote panel could have the same effect. It's very hard to say what happened and if it could happen again. You'll just have to inspect all the wiring and keep your fingers crossed.

 

[mypcbis$crewed]

Anything is possible, and those are excellent possibilities. I just hope that it works for a good long time! Thanks again ever so much for all your help!