Announcement

Collapse
No announcement yet.

Testing IGBT with a meter... especially fgd4536 and 30f131

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Testing IGBT with a meter... especially fgd4536 and 30f131

    Read thru lots of posts about fake items so there is great potential for coming across these. Of course, the simple test of the failed units showing a short isn't the point of this post.

    I wanted to get some people's experience with "stock they have on hand" to see if the "simple diode test" tells you anything more about the potential for having a good device (ie. maybe not fake) before soldering it to the board.

    In diode mode, with the black lead on the emitter, and red on the center pin, the meter should read nothing. The test then is to put the red on the gate (opposite side from the emitter) for a short period to charge the gate, then put the red back on the center pin. Doing this with one set of fdg4536, I get a reading of .9v. One of these devices failed during my initial replacement effort. I had some more of these plus another batch. I tested the other batch and them seemed to show 2.6 volts. I used them and the second attempt resulted in two of those blowing their top plus others shorted. Not necessarily blaming them as my soldering maybe suspect but failure all the same.

    I also some 30f131's ("far east" of course) and tried the same test.
    However, I can't get them to "conduct" as touching the gate never seems to switch the emitter collector junction. Are the 30f131 that much different in design than the fdg4536's???

    So my question to others here is... if you have some fdg4536 that you think / know are good, can you do a little test on one of them and state what the results of the diode test is and similarly if you have some "good" 30f131's around can you also try the same thing with one of them and post some of the results.

    I know that the test doesn't give certainty on the "fakeness" of the device but perhaps the simple diode test might reveal something for us "one timers".

    Thanks for any help / results.
    Last edited by budwich; 09-21-2018, 08:55 AM.

    #2
    Re: Testing IGBT with a meter... especially fgd4536 and 30f131

    I have a cheap "put it together yourself " transistor tester. It is an E.S.R. meter also.
    On it a bad igbt will read as a capacitor. I have also had a few that read only as a resistor. Here is a few pics of good and bad parts.
    Attached Files

    Comment


      #3
      Re: Testing IGBT with a meter... especially fgd4536 and 30f131

      Originally posted by budwich View Post
      Read thru lots of posts about fake items so there is great potential for coming across these. Of course, the simple test of the failed units showing a short isn't the point of this post.

      I wanted to get some people's experience with "stock they have on hand" to see if the "simple diode test" tells you anything more about the potential for having a good device (ie. maybe not fake) before soldering it to the board.

      In diode mode, with the black lead on the emitter, and red on the center pin, the meter should read nothing. The test then is to put the red on the gate (opposite side from the emitter) for a short period to charge the gate, then put the red back on the center pin. Doing this with one set of fdg4536, I get a reading of .9v. One of these devices failed during my initial replacement effort. I had some more of these plus another batch. I tested the other batch and them seemed to show 2.6 volts. I used them and the second attempt resulted in two of those blowing their top plus others shorted. Not necessarily blaming them as my soldering maybe suspect but failure all the same.

      I also some 30f131's ("far east" of course) and tried the same test.
      However, I can't get them to "conduct" as touching the gate never seems to switch the emitter collector junction. Are the 30f131 that much different in design than the fdg4536's???

      So my question to others here is... if you have some fdg4536 that you think / know are good, can you do a little test on one of them and state what the results of the diode test is and similarly if you have some "good" 30f131's around can you also try the same thing with one of them and post some of the results.

      I know that the test doesn't give certainty on the "fakeness" of the device but perhaps the simple diode test might reveal something for us "one timers".

      Thanks for any help / results.
      I suggest that you 'knock up' a proper IGBT test circuit as described here:-
      https://alpengeist-tvrepair.blogspot...ansistors.html
      This site has helped me greatly with fixing Panny Plasma sets. I have a batch of 50 FGD4536's from UTSource and they are fine... I can give you figures from the test circuit. All bulk 30F131s appear to be fake and no good long term. Only trust those supplied singly in Panasonic bags.
      The blog also describes some of the 'fun' I had due to fake parts!
      Regards,
      Steve
      Last edited by foxyUK; 09-26-2018, 03:48 PM. Reason: more info added

      Comment


        #4
        Re: Testing IGBT with a meter... especially fgd4536 and 30f131

        Thanks a bunch. That's looks readily doable for even a "one timer diy".

        One question... the blog didn't show testing of the FGD4536 but did they act similar to the 30f131's? The reason for the question is the fdg4536 that I used initially, instead of 30f131, readily were "checkable" with the diode test in the DDM using the method that you described in the blog while the 30f131 showed nothing. Of course, the fdg's didn't survive and hence my post.

        Anyways, I will see what I have around to do the circuit. Meters I got, variable resistor might need some digging... but I got an old "bread board" so it should work out overall.

        Comment


          #5
          Re: Testing IGBT with a meter... especially fgd4536 and 30f131

          I ran a couple of 30f131 using my "crude circuit tester". Of course, I can't tell how "tough" they are (ie. loading) but did see that the gate voltage below about 3.5 volts opened the device (no longer conducts). Above that value, it looks to conduct fine. I haven't tried the fgd4536 but those were "gateable" using the diode function on the meter so I am not cure the test circuit would be much different.

          Perhaps at least the 30f131 that I got are OK... and the fgd's were not good (ie. they blew in the SC board). I do appreciate the comments on testing the igbts.

          Comment


            #6
            Re: Testing IGBT with a meter... especially fgd4536 and 30f131

            weight is often a factor with fakes ...

            Comment


              #7
              Re: Testing IGBT with a meter... especially fgd4536 and 30f131

              interesting... I don't have a known sample to compare against... but I do have my "marjawana scale" to get an accurate reading if I do get a "benchmark"... :-)

              Comment


                #8
                Re: Testing IGBT with a meter... especially fgd4536 and 30f131

                Originally posted by budwich View Post
                interesting... I don't have a known sample to compare against... but I do have my "marjawana scale" to get an accurate reading if I do get a "benchmark"... :-)
                simple balance scales fine .. if original to start with all original should balance out . p.s dont tell this to the fakers or they will overcome it becoming an issue with them

                Comment


                  #9
                  Re: Testing IGBT with a meter... especially fgd4536 and 30f131

                  use the scales you use for gold scrap if you scrap gold that is

                  Comment


                    #10
                    Re: Testing IGBT with a meter... especially fgd4536 and 30f131

                    I looked at the link mentioned earlier in this thread with details of the IGBT tester, which in turn had a link to a site https://www.transistormosfet.com/ for sourcing components.

                    Has anyone tried parts from this site and any idea where they are based? I've put a selection of part numbers in from a blown TNPA5105AB and they all show as in stock. Everything is $3 a piece too.

                    Comment


                      #11
                      Re: Testing IGBT with a meter... especially fgd4536 and 30f131

                      as a note i have had real ones that wont switch on with my meter but worked perfectly in circuit . so in my view diode test is not reliable even for go no go testing .

                      Comment

                      Working...
                      X