Announcement

Collapse
No announcement yet.

Marshall Lead 100 Mosfet 3210 producing loud intermittent/random noises

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

  • Marshall Lead 100 Mosfet 3210 producing loud intermittent/random noises

    Help needed with...
    Marshall Lead 100 Mosfet 3210 ss amp, pcb: jm88
    The quality control sticker inside is dated 06/06/86 & appears to be unmolested.

    This is the schematic I'm using since the only other one I could find is too difficult to read...
    http://drtube.com/schematics/marshall/3210-iss7.gif

    This amp powers on, but I'm getting loud intermittent cracking & popping noises thru the speakers [2x12" 8Ω Celestion V30's wired into 16Ω].
    When patched thru another amp [same speakers] it produces a nice clean sound so that should rule out the preamp correct?

    I'm a novice to guitar amplifier repair but I have a grasp on electrical components & the tools/ability to solder,
    so please don't become impatient if I don't completely understand everything right off.

    At first glance after opening it up, I saw melted insulation on the wires connected to the K135 mosfet.
    Click image for larger version

Name:	IMG_9614 20%.jpg
Views:	1
Size:	185.5 KB
ID:	871437 Click image for larger version

Name:	IMG_9615 20%.jpg
Views:	1
Size:	217.7 KB
ID:	871438

    Although it was still producing sound the pin on K135 with the yellow wire was actually able to move [up & down] in the mosfet.
    Obviously this part got really hot & needs replacing so I'll replace both mosfets to keep them closer together in spec.

    I removed the board & inspected the solder joints which all looked ok.
    However, there were a few dark spots on the circuit board.
    Click image for larger version

Name:	IMG_0121 50%.jpg
Views:	1
Size:	518.5 KB
ID:	871439

    Click image for larger version

Name:	IMG_0119 50%.jpg
Views:	1
Size:	512.7 KB
ID:	871440

    Click image for larger version

Name:	IMG_0102 20%.jpg
Views:	1
Size:	235.6 KB
ID:	871441

    I'm not very experienced at following voltage across a schematic but I'd like to make sure the components near these dark spots aren't damaged from overheating.
    Perhaps with your help I can test these components & be sure to correct any problems the first time & not blow up new parts by rushing thru the repair.


    I sourced some parts but have yet to dive into replacing anything yet,
    until I can get some opinions & advice from those with the experience I am lacking.

    I don't want to make any rookie mistakes, so if anybody would like to offer their expertise in which components should be tested & why, It would be very much appreciated.
    Hopefully I didn't leave out any pertinent info.

    Thank you guys for your time & any help you can offer...

  • #2
    Hi Mattinthebox, welcome to forum,

    Have you verified that the pin on the yellow wire has become detached from the mosfet. I can't quite work it out from the picture as to whether those FETs are in sockets or just mounted to a piece of chassis. If they are in sockets then it would be good idea to investigate the pins and the fit. Maybe the socket is the problem, or part of the problem and not the FET. Might be wise to dismantle ad check the fit.

    If it was a bad connection to the FET that would tie in with a LOUD pop.

    The toast marks on the board are not too bad. You should measure those components and if they check out ok, then resolder. Best approach is to desolder the part, clean the board with solvent like IPA (alchohol) and toothbrush to ensure good joint. If hard to get a good joint, then you can wirewoll the resistor legs to remove any oxidisation.

    You should also google dim light bulb tester and consider making one up. It will help prevent damage to the amp in case anything goes wrong with the disassembly et

    Comment


    • #3
      mikeydee77
      Have you verified that the pin on the yellow wire has become detached from the mosfet
      It isn't detached, the wires are soldered directly to the pins [gate/drain], no sockets.
      The black item which appears to be a socket is nothing more than a spacer to isolate the mosfet from the heat sink.
      The actual pin or leg [drain] that connects to the internal guts of the mosfet is whats moving. I believe the heat loosened the material that holds the pins & had allowed the pin to slip in & out of it's factory position.

      For replacements I got a set of BUZ900/BUZ905 mosfets.
      According to the data sheet [as long as it's the same pin out for the K135] the drain pin on the original K135 mosfet is whats moving.



      The toast marks on the board are not too bad. You should measure those components and if they check out ok, then resolder
      Ok, so should I only check the components where it's browned? TR4, TR5 & R44, R45? Or are there other critical parts I should test?



      You should also google dim light bulb tester and consider making one up
      Gotcha on that. I've read up on that & whipped one together.
      I've read, "the higher the bulb wattage the lower the bulb resistance."
      It was recommended to use a bulb of 250w but I only have a 60w on hand, everyone has been switching to fluorescent & or led bulbs so would a 60w incandescent bulb be sufficient & safe for testing?


      On a side note, can anyone tell me if the grey silicone insulators under the mosfets were original when these amps were manufactured?
      I'm trying to determine if they have been replaced previously.
      The grey pads look a bit tore up, they have slight perforations from the rough surface of the K135/J50 mosfets so I replaced them with mica & white thermal paste.

      Comment


      • #4
        I would "tickle" TR6 and TR9 with some freeze mist. That's a beta matched pair. If the hfe of either has drifted it will pop like a whore chewing gum. I actually have an untouched working chassis and matching Reverb tank for one of these (last time I looked) if you need one. I gut these and build 20 watt tube amps in the head cabs. They look nice.

        Comment


        • #5
          The 60W bulb is a better choice than the 250W. There is some misinformation on the web about using higher wattage bulbs, which will not offer much protection.
          Those original Fet's are very rare and probably worth more than you think, even with a loose leg like that. I would guess there are people out there that would glue it in place and use it, if the connection is not intermittent.
          "Everything is better with a tube. I have a customer with an all-tube pacemaker. His heartbeat is steady, reassuring and dependable, not like a modern heartbeat. And if it goes wrong he can fix it himself. You can't do that with SMD." - Mick Bailey

          Comment


          • #6
            If it pops copper is already separated

            If **absolutely** necessary he might grind 1.5 to 2mm of plastic and solder a wire to the stump; if a TO3 package forget it because its encapsulated in glass.

            EDIT: its TO3 and glass so forget it
            Last edited by J M Fahey; 03-28-2017, 01:00 AM.
            Juan Manuel Fahey

            Comment


            • #7
              I would "tickle" TR6 and TR9 with some freeze mist.
              Sorry I don't follow... "tickle" how?

              That's a beta matched pair. If the hfe of either has drifted it will pop like a whore chewing gum.
              If they pop will there be visible damage or will they need to be tested with a meter? These appear to look ok & the unit does still produce sound.

              I actually have an untouched working chassis and matching Reverb tank for one of these (last time I looked) if you need one.
              Thanks! If I need one then I'll definitely take you up on your offer.
              Would you mind shooting me a pm with what you'd like to get [$] for it?

              I would guess there are people out there that would glue it in place and use it, if the connection is not intermittent.
              Is there any way to test these mosfets with a amprobe multimeter?
              I've also got an old 3 lead micronta 22-025 dynamic transistor checker but it says for NPN/PNP transistors so I don't think that meter will do for these.

              Besides verifying if the original hitachis are still good, I wasn't paying attention when I removed the hitachis & think I may have installed 1 of the buz9xx's upside down on the heat sink.

              So are all n-channel & p-channel pin outs the same?
              Are the source & drain pins oriented the same way for both of the 2SK135/J50 & BUZ900/BUZ905 mosfets?

              Here's the data sheets for the BUZ900/BUZ905 mosfets. I couldn't find any readable data sheets for the 2SK135/J50 mosfets.

              BUZ900:http://www.mouser.com/ds/2/414/BUZ900-242310.pdf

              BUZ905:http://www.mouser.com/ds/2/414/BUZ905-242291.pdf

              I'm trying to *not* sound like a dumbass, [Red Foreman], but I lack experience & don't want to grenade the new transistors.

              Thanks for the helpful replies so far...

              Comment


              • #8
                Originally posted by mattinthebox View Post
                Sorry I don't follow... "tickle" how?


                If they pop will there be visible damage or will they need to be tested with a meter? These appear to look ok & the unit does still produce sound.


                Thanks! If I need one then I'll definitely take you up on your offer.
                Would you mind shooting me a pm with what you'd like to get [$] for

                Thanks for the helpful replies so far...
                Tickling just means use a can of freeze mist with a reducer tube to thermally shock those components one at a time. If the problem is there it will go away or get worse. It's a common (sometimes too common) troubleshooting technique for isolating intermittentcies.

                Comment


                • #9
                  Well I tested a few transistors with my transistor meter.

                  TR6 & TR9 tested good but TR4 & TR5 are bad.

                  TR5 was completely dead & TR4 is failing, the light on the tester faintly glowed showing very little if any current amplification.

                  The schematic shows these [TR4 & TR5] near the reverb unit.
                  Are they necessary components of the reverb circuit because if I remember correctly the reverb worked?

                  Is there anything else around these failed parts that I should be testing?

                  Comment


                  • #10
                    Dont lose focus or wander all over the place.
                    1) you confirmed preamp working well so stop messing with it,reverb, etc.

                    2) you have an apparently working power amp, which now and then pops loudly.
                    You also found a loose power transistor leg, which is cosistent with loud pops.

                    So concentrate on solving *that* first and foremost:
                    a) build a lamp bulb limiter, a 60W bulb is fine, if the old type is not available (you might have one still working at home), the new "small quartzlight bulb inside a larger glass bottle" work fine, what you want is a smallpiece of red hot wire between mains and your amp.
                    Check actual power consumption: "75W equivalent" will probably mean some 52W or so, fine for this test.
                    b) replace thosen transistors with the new ones, preferrably use mica and grease, if not *new* Silpads becausethey aresingle use
                    c) turn amp on, no speaker attached, check for no (or very little, less than 100mV) DC at the speaker out.
                    If fine, connect a speaker without turning amp off and play a couple chords at low power, say bedroom approved 1 or 2W.

                    If fine, you might plug amp straight into mains.
                    Juan Manuel Fahey

                    Comment


                    • #11
                      Ok, I attempted to reassemble & power up this amp but things did not go to plan.
                      Here are my findings.

                      Reinstalled TR6 & TR9, these tested good so they were returned to their respective locations.
                      Replaced TR4 & TR5 with new parts.
                      Installed BUZ900/BUZ905.
                      Reinstalled circuit board, jacks, pots, etc...
                      -Reverb unit NOT connected-
                      Connected light bulb limiter.

                      1st attempt
                      Turned unit on.
                      Noticed red power switch was *not* illuminated even though amp had power applied.
                      Measured dc voltage at speaker output -5.83v
                      Connected a 15" 8ohm speaker & there was a motorboating effect, or oscillation coming thru the speaker.
                      Adjusting every pot on the amp had no effect. Attempt to control volume increase or decrease had no effect, just a constant tone.
                      BUZ900 quickly heated up, limiter lightbulb became very bright then went off.
                      At the moment the limiter bulb went off the red light on the amps power switch lit up again.
                      I also noticed that TR4 & TR5 got very hot to the touch.
                      Bridge rectifier was also hot to the touch...

                      ---
                      2nd attempt limiter bulb did not light up.
                      Amp power switch lit up red.
                      Speaker out was 200mv & counting down.
                      Both mosfets were cool to the touch.
                      TR4 was hot to the touch again, TR5 was not.
                      R44, R45, R46 & R47 were quite hot.
                      Bridge rectifier was not hot.

                      For chits & giggles I connected a speaker once again to see if there was any output.
                      The speaker made a buzz like a 60cycle hum, not the same motorboating I heard the first time.
                      Adjusting pots had same effect as before, no effect.

                      So I'm right back to where I started, actually further back since I have no sound now.
                      It looks like I've burnt up at least 1 mosfet if not both & something else must have failed.

                      Is it possible that not connecting the reverb tank could be part of the cause?
                      I don't know what I did wrong. I only removed, tested & replaced the 4 transistors & 2 mosfets, nothing else.
                      Any suggestions?

                      Comment


                      • #12
                        What bulb size did you use?
                        Juan Manuel Fahey

                        Comment


                        • #13
                          The bulb I had originally thought was 60w was actually a 40w.
                          I forgot to mention that when I attempted to use this bulb, it initially came on with a series of flashes like a steady pulsing.
                          That's when I checked the bulb & discovered it was 40w.
                          I looked through my bulbs & found an old 100w, which is what I ended up using.

                          Comment


                          • #14
                            I did some testing to find the failure butTR4, TR5, TR6, TR9 were getting very hot, I removed these, tested them & left them out of the circuit.

                            TR4 MPSA56 PNP tested good
                            TR5 MPSA06 NPN tested good
                            TR6 MPSA06 NPN tested good
                            TR9 MPSA06 NPN tested good

                            I tested one of the mosfets, BUZ900, which tested good.
                            I have yet to pull the BUZ905 to test yet.

                            I tested for rail voltages & have +/- 49v, 37v & 16v
                            I'm still getting -49v at the red wire & 0v on the black wire on the speaker output.

                            I don't know how hot the components should get while sitting idle with no signal going thru it but several items were quite hot.

                            Component --- Value --- Temp --- Voltage In/Out
                            R47 _________ 270Ω 2w __ 100 ___ -49.6/37.4
                            R46 _________ 270Ω 2w __ 110 ____ 49.6/36.5
                            R45 _________ 470Ω 2w __ 130 ____ 36.5/16.0
                            R44 _________ 470Ω 2w __ 160 ___ -37.4/16.8

                            C38 _________ 100F 25v __ 185 ___ -16.4/0
                            C39 _________ 100F 25v __ 130 ____ 15.9/0
                            C40 _________ 47F 63v ___ 110 ____ 36.5/0
                            C41 _________ 47F 63v ___ 90 ____ -37.3/0

                            Are these temps normal? Isn't 185 at C38 on the verge of failure? Shouldn't it be much cooler?

                            Everything testing good to the best of my knowledge but I know I'm missing something...
                            I could really use some direction which way to go.

                            Comment


                            • #15
                              Well I've returned TR6 & TR9 back into the circuit.
                              I double checked everything once again.

                              Apparently TR6 & TR9 break down the voltage that was at the speaker out.
                              The speaker output now had -.6v on the red wire & -.3mv on the black wire.

                              I connected a 15" 4Ω speaker just to see if I had sound & eureka! I have sound once again.
                              I plugged in a guitar & lightly strummed some notes, I've got notes coming through the speaker but it's still crackly.
                              Hopefully reattaching the screws & pots to the chassis will ground any noise, I'll report my findings after I get to that point.

                              I don't know if this was a dumbass move but the first time around I had the circuit board isolated from the chassis (to easily flip front to back for pulling components).
                              My observation is that I broke the ground between the circuit board & the chassis which caused all the buggy results I had gotten.

                              I still have TR4 & TR5 out of the board. I noticed the lead coming off the board for the reverb tank was frayed & about to break, it literally detached when I touched it.
                              I'll repair that & check the reverb tank before I put TR4 & TR5 back in.

                              Yes progress! ...at least I'm going in the right direction again...

                              Comment

                              Working...
                              X