Previous board repair appeared to swap TD62064AP for TD62064APG. Should I swap?

[barfdogg]

So I have a "Densha De Go!" arcade cabinet and I have come to find that a critical indicator light (0.3 amp 12V Incandescent Bulb) is not illuminating when commanded. Testing revealed that the negative wire was sat, but I get nothing on the 12V+ wire. (I had this wrong, the NEGATIVE wire is the dead one) Tracing it back to the board sets I narrowed it down to a specific PCB. (I am unable to find any repair manuals or schematics as everything is in Japanese, so I'm sort of winging it here.) Anyway, I believe I am on the right board... I started inspecting for failures and came across a TD62064APG that had clearly been swapped out as I see some solder blobs and naked pins. My first urge was to repair the solder job, but I have a few things I am unsure about...

First, am I correct in assuming that you would use this IC to drive a low voltage lamp, and am in the right area?

Second, it appears that the "G version of this IC is quite similar to the original version, and looking at the datasheets I'm not sure what the difference even is... My first instinct is to find the original IC and replace this with a new one to ensure that this one isn't fried, fix the missing solder, and have the original part number in there, just to be extra safe.

Input?

(Thanks BTW)

[Benta]

The "G" version si exactly the same as the non-G, except it's lead-free. This might explain the bad soldering job.

[barfdogg]

Ah, that makes sense. Thank you.

As for replacing it, I see that they are EOL. I have always replaced components with the exact same model in the past. I can find them on eBay but am worried about authenticity. I also see there are similar options currently in production from other manufacturers like ULN2064B. Should I just get one of those from a legitimate supplier or what?

Thanks again.

[Ian.M]

You need to prove IC81 is faulty, not just jump to it because its been (badly) resoldered.  Trace the faulty lamp circuit all the way back to the specific pin of the specific driver IC.  If its not IC81, you are barking up the wrong tree.   If it is, use its datasheet to locate the corresponding input pin and see if its getting a drive signal. 

[wizard69]

As others have said I'd almost certainly start by fixing the botched soldering job.    It looks like the last hack didn't even bother to solder some pins (maybe they are not used) but the whole job leaves me thinking that the old solder job itself may be a problem.   

As for actual chip replacement I also believe that you should verify that it has actually failed.   That is verify the state of both inputs and outputs.   With chip replacement you always run the chance of board damage, so random replacement of IC's should not be avoided.    This isn't at all like the repair of decades old equipment with dodgy capacitors.

[james_s]

I would definitely start by checking the solder joints and look closely for damaged traces and torn out through-hole plating, that is very common with amateurs lacking the proper tools and skill trying to replace parts on arcade boards.

[mikerj]

The TD62064APG could be used to drive a low power incandescent lamp but note that it is a low side switch.  Since you diagnosed that the lamp is missing a +12v supply, this fault would not be caused by this part since it would be switching the ground side of the lamp. 

[barfdogg]

Thank you for the suggestions.

I have checked continuity from from the bulb and traced it all the way back to the O2 Pin 7 on the TD62064APG at IC81. Even with the garbage solder, it seems to have continuity.

Either way, I decided to remove and re-solder the IC, so even if still defective it will at least look nicer.

I then traced the I3 Pin 11 on the TD62064APG back to a PWM pin at IC67 which is a M68HC11. There was continuity between these two as well. (Mind you I checked this AFTER I re-soldered the TD62064APG pins which I did not think to check beforehand)

At this point I think I will have to simply assemble the PCBs and start the machine up to see if any changes occurred, as there are three PCBs sandwiched together and I have no clue how I am going to check for a PWM signal while it is in operation to see WHICH IC has failed, but I guess I will cross that bridge when I get there.

It seems at this point I have confirmed that there is continuity between everything, so if the reason it does not function was not the poor solder, then it would have to mean that one of these two ICs are toasted. I will report back the results when I get it back together.

[james_s]

Once you had the IC out would have been a great time to install a quality machined pin IC socket. I normally don't bother with IC sockets, but in the case of a board that has already been damaged or a part like that which drives external items and potentially has a high failure rate I make an exception.

[barfdogg]

Ok, so I tested all the traces for continuity to/from the TD62064APG and am confident in my soldering job. Reassembly and testing resulted in the same outcome (of no change on the output.)

The TD62064APG could be used to drive a low power incandescent lamp but note that it is a low side switch.  Since you diagnosed that the lamp is missing a +12v supply, this fault would not be caused by this part since it would be switching the ground side of the lamp. 

Your comment made me second guess my initial diagnosis, and you were correct- I apparently did not notice the polarity in my initial testing. There is a blue wire going to all of the indicator lamps and this is the 12V supply. The 12V supply to this location is functional as with the others. When I send a command to illuminate the lamp, there is no appreciable change in voltage from when not active.

Some interesting behavior I noticed, is that when switching on other lamps (driven by TD62064AP ICs as well) I will see 11.4V appear when ON, and when off I will get .0001V with no variance (which is what I would expect) but the negative line going to the TD62064APG in question is always floating around. My meter will jump up and down randomly between +.04 and -.04. I know it is a small number but everything else seems locked at 0 when off. Is this difference in behavior indicative of anything?


Also, since I cant figure out how to have this thing assembled and ON to test the PWM line to the TD62064APG I am going to wing it and install a new one. That will at least narrow it down a lot more. I ordered a NOS TD62064AP and hope to get it by next weekend.

Thanks!

[Ian.M]

Also, since I cant figure out how to have this thing assembled and ON to test the PWM line to the TD62064APG ...

Solder a thin insulated wire to the pad of the offending IC83 I3 Pin 11, and snake the other end out to where you can access it to test it running.  Tape it down near the end to an insulating surface so it cant accidentally short to chassis, a power rail or another signal.

[barfdogg]

Good idea. I will do that.

[james_s]

0.04V is irrelevant, that's a ridiculously small amount of voltage, likely just the result of the ground potential changing slightly from other loads coming on and off. Can you measure continuity from the terminal of the lamp all the way to a pin on that IC?

[thinkfat]

Did you entertain the idea that the lamp is broken? If the driver chip pin is always low, the lamp should be lit. This is a low-side switch. If it's open, you should see the lamp supply voltage at the pin. If it's closed, you should see 0V or thereabouts. You get 0V and no light, so either the lamp or the connection to the lamp is broken.

[barfdogg]

Yep, I attached a test lead to from the leg of the IC and get continuity at the very end of the wire. (I removed the lamp base for testing.)

I will be checking for the PWM signal that should be triggering the TD62064APG when I get another chance to work on it. If I can see the signal that will pretty much doom this IC.

I know the voltage is small, just found it odd this one was moving slightly when on or off due to the the lamp wiring NEXT to it would stay exactly steady when on or off. I assume this is due to it being "open"

[barfdogg]

When I first started repairing the machine a dead lamp was my first assumption as a few others were out as well.

When I went to replace it, I found that someone had removed the factory lamp base and replaced it with a non standard one. There was already a new bulb installed, but I replaced it anyway with no change. I then tested the lamps in other sockets and they were functional.

Once I started diving deeper I found repair work had been done to the IC that drives this lamp, so It appears someone had previously assumed the lamp base was faulty, then either fixed or attempted to fix the issue by replacing the IC. "if" they fixed it, then the replacement has failed as well, or if it did not fix it, then the culprit is likely the M68HC11 not sending the PWM signal the driver is waiting for. I think we are close to an answer.

[james_s]

Ground the wire that leads to the lamp and confirm that it comes on. I've seen a lot of bad lamp sockets, bad connections, broken wires, etc. It's very easy to verify this.

[barfdogg]

Just finished replacing IC81. It was in fact the culprit. While it was receiving a signal, it was not changing the output state. I swapped it out and put it all back together and good as new.

Thank you everyone that commented with input on the issue.