Testing components (Diodes) in-circuit

Discussion in 'Technical Repair' started by Standisher, Mar 20, 2017.

  1. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    Not sure if is the right board for this thread so happy for moderators to move if necessary.

    Background - feel free to skip

    I acquired a faulty Function Generator (Instek SFG-1013) for a very low price and am hoping, despite my quite basic skills, to repair it. The unit powers up and generates a display of frequencies. The TTL output seems to be functioning also and the set frequency corresponded with the frequency measurement on my oscilloscope. The main output however delivers NO signal at all. On opening the unit I found 2 resistors badly scorched and one of the output transistors was also running very, very hot.

    I have been in correspondence with the manufacturer. They agree with my belief that the 2 main transistors in the main output circuit were blown (which the say causes the resistors to run hot). They also think several other general purpose npn transistors are also likely to be damaged and are currently in the process of securing a total of 9 components which they will supply free of charge (very good of them) and asked if I will take a photo of one of the npn transisters to make sure they send the correct items (5 of them the same)

    New concern - the point of this thread

    I had the unit open today to take the requested photo and thought I would check a few of the diodes in a small cluster in the output path. Using my DMM in the diode mode. Alternating the pos & neg probes on anode & cathode I got a reading in both directions (on a total of 7 diodes - see pic). Typically the diodes read 0.612V (with pos lead on anode and neg lead on cathode) and 2.30V with probes reversed. I had expected to get NO reading when testing with pos probe on the cathode. All diodes have a very similar reading but diodes elsewhere, outside the output path gave me the expected reading (i.e NIL reading when reversing probes). Could I be missing something basic when testing diodes in-circuit like this or is it likely that all 7 diodes could have shorted simultaneously? I haven't even examined the 2 zener's (not sure how to test them) and if all these components (total of 16 to date have gone south, I wonder if zeners and opamps are likely to have blown too :eek:

    I've attached a photo of the out portion of the pcb . Sorry for the long post. diodes4.jpg
     
  2. ian field

    AAC Fanatic!

    Oct 27, 2012
    5,406
    961
    It used to be easy with through hole parts - just snip one of the diode leads and you can test it unencumbered by other circuit elements.

    You should ignore any reverse leakage because its probably not the diode - and some diode types normally have some leakage.

    You could probably reduce the task by using a DMM diode check function to read the Vf of each diode, I'd certainly remove any that read shorted and have a closer look - but some diodes have a lower Vf, and other circuit elements can sabotage the readings.
     
    Standisher likes this.
  3. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    Thanks for the reply @ian field . When you say:

    I was in diode check mode on the DMM and I thought the 0.612V I quoted WAS the Vf...was it not? I was more concerned that I was getting 2.3V reading when I reversed the probes on the diodes.
     
  4. LesJones

    Active Member

    Jan 8, 2017
    587
    120
    The reverse voltage reading is probably due to the components that the diode is connected to. You would need to trace out the schematic of the commponents connected to the diode.

    Les.
     
    Standisher likes this.
  5. ian field

    AAC Fanatic!

    Oct 27, 2012
    5,406
    961
    Out of circuit; it should read "out of range" for most diode types in reverse.

    You can't trust any kind of reverse leakage reading with the diode in circuit.
     
    Standisher likes this.
  6. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    OK guys, I think the best thing I can do is desolder one of the diodes with strange readings and see what that produces out of circuit. Trouble is with these miniature, surface mount components, I'll have to be careful not to damage it in case it was OK in the first place :)
     
  7. ian field

    AAC Fanatic!

    Oct 27, 2012
    5,406
    961
    Those round SMD diodes are a PITA. I'd order the service manual and find out what they are - and then just replace any suspects.

    As I'm a cheapskate - I'm just as likely to form the leads on regular diodes to match the pads - but its still worth finding out what they are.

    Predictably - its easier to solder the new part in than get the old one out. My method is to tin the ends with fresh solder and flux, then keep moving the iron from end to end till it heats up and comes adrift. You can get a special ultra low melting point solder that dilutes what's there and makes unsoldering SMD parts much easier - maybe someone can remind me the brand name
     
  8. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    Thanks Ian. I think the low melting stuff you're thinking of is Chip-quik (or something like that)? Would love to be able to get a service manual but can't find one out there despite giving it the heavy google fingers. Even the manufacturer doesn't seem very knowledgeable about what components are in there...hence asking me to take a photo even though I had given them the part # stamped on the transistor (1AM) !

    If anyone knows a good source for service manuals it would be great.
     
  9. ian field

    AAC Fanatic!

    Oct 27, 2012
    5,406
    961
    When I did TVs for a living - most manuals places charged too much for practically unreadable lousy copies - any time they were readable; they turned out to be for a different version of the chassis.

    There are online archives of test equipment manuals and schematics that just take a bit of finding. There was the BAMA boat anchor archive - but I haven't looked in for a while and can't confirm they're still there.
     
  10. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    I'm re-posting to this thread I started a long time ago to avoid starting another with an associated question. Yes, I'm still trying to find the fault in the amplifier part of a G W Instek (SFG-1013). I've posted below, an annotated photo of part of the circuit on the negative side of the NPN/PNP power transistors as I'm puzzled by the voltage readings and am hoping someone can explain how the significant change in voltage from -15V to -0.634V after the 7.5K resistor could come about. I have tried removing the NPN transistor (Q908) to see if that was somehow responsible but there is no change in the readings. I really struggle with these boards which have 'vias' so am finding it hard to find what could have this effect (incidentally, the 7.5K resistor is fine).
     
  11. dl324

    Distinguished Member

    Mar 30, 2015
    4,674
    999
    You need to post a schematic. If you don't have one, trace enough of the circuit around the components in question so we can do analysis.
     
  12. BR-549

    Distinguished Member

    Sep 22, 2013
    2,869
    672
    Standisher.......have you ever heard of an oscilloscope octopus? The octopus circuit is easy to build.....and will display the p/n junctions on the scope. Without any un-soldering.

    Google........oscilloscope octopus. Saves a lot of time.
     
  13. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    No, I don't have (and can't get) a schematic @dl324 (the manufacturer won't supply one and there is nothing online) . As I say, the difficulty for me is that traces from some of the components lead to a 'via' and I ham having great difficulty establishing where those vias lead. I was not expecting a full analysis from anyone (I know that is difficult/impossible with my inability to trace the circuit out in view of these vias which send me down rabbit holes). I was just hoping for some kind of clue on the circumstances in which such a voltage differential could occur. I'm suspecting that you may be saying that there is too wide a variety of things that may cause this for anyone to help?
     
  14. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    I have built a very helpful curve tracer @BR-549 , that's what you meant by 'oscilloscope octopus' and have, with the help of that, moved on from worrying about the diodes as they check out fine. I re-opened the thread to talk about voltages but realise the original title of the thread is misleading ... sorry about that. :oops:
     
    Last edited: Aug 14, 2017
  15. ian field

    AAC Fanatic!

    Oct 27, 2012
    5,406
    961
    Some types of ESR meter will accurately read very low resistances and most have defined test signal that won't forward bias any (most) semiconductor junctions. Its seriously tedious tracing tracks this way, so last resort - but it can be done.

    You'd be surprised what lurks in schematics archives - but not easy to find, mostly stuff turns up when I'm looking for something else.
     
  16. dl324

    Distinguished Member

    Mar 30, 2015
    4,674
    999
    Can you determine if the board has more than two layers? Do all of the vias have traces on both sides of the board?

    If the board has inner layers, tracing the circuit could be very tedious, but tracing the circuit in question on a dual layer board shouldn't be too difficult.

    Doing the whole board might be a different story. I understand this situation because I have a couple power supply repair projects where I need to trace the whole circuit to understand how it works and I always find something more enjoyable to do.
    Can't do any analysis without a schematic.
     
    Standisher likes this.
  17. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
    Update - Success! Today, I replaced all 5 surface mount transistors (MMBT3904LT1G), 2 1N4148 and the two power transistors (2N2219A and 2N2905A) and fitted their heatsinks (which were not in the unit when I got it). I originally bought the unit for £25 and have spent £16 on parts so not at all bad for a unit that costs circa £140 to buy new. Everything is now working perfectly. Thanks for your time and contributions to my questions.
     
    Last edited: Aug 15, 2017
    RichardO likes this.
  18. RichardO

    Well-Known Member

    May 4, 2013
    1,567
    540
    It looks like you got more than £140 in education. :):):)
     
    Standisher likes this.
  19. Standisher

    Thread Starter Member

    Jan 16, 2015
    89
    8
Loading...