Zap! Pop! Sizzle! R1 is toast [resolved]

Bartguy · 1451

0 Members and 1 Guest are viewing this topic.

Offline Bartguy

  • Jr. Member
  • **
    • Posts: 32
on: July 16, 2020, 03:57:45 PM
Hi, something shocking just happened. I had previously built my Kaiju up through mounting the C4S boards using .1uf Audyn copper caps between 4U & 32U and 23U & 46U. I did my Driver testing voltage checks and all my numbers came in fine on both boards (IA-511VDC, OA-299VDC, 32U-175VDC 46U-175VDC).

Then in another thread, I found out that I'd need to secure those caps much more firmly to the board, and might have to order a solder pot to feel confident about soldering the stranded Audyn leads and decided to substitute some .22uf 1400vdc Mundorf supreme caps instead because their leads would be easier to solder securely. I confirmed the higher cap rating would work and proceeded to replace the Audyns. So I just resoldered 4 connections, which look fine.

When I was done, I decided I should redo the Driver voltage tests just to confirm I was still good to go. Plugged it in, all seemed fine. The A side board measured the same as the previous test. On the D side board, IA also measured fine, but when I touched my multimeter probe to OA, I heard a load zap and pop and the R1 resistor burnt out (see photos.) Did I do something with my multimeter probe that did it? I checked the wiring on the D side board and it looks correct, and the cap soldering between 23U & 46U looks good.

Any idea what happened? Can I just replace R1 on that board or should I suspect that I damaged some other components?

Any advice on how to proceed would be welcome. Thanks.

« Last Edit: July 28, 2020, 04:49:42 AM by Paul Birkeland »



Offline grufti

  • Sr. Member
  • ****
    • Posts: 320
Reply #1 on: July 16, 2020, 04:37:03 PM
You have melted quite a few parts. The LED's are toast and I would consider at least one of the transistors suspect. It's easier to replace those than to try and figure out, if you need to.



Offline pfontana

  • Jr. Member
  • **
    • Posts: 19
Reply #2 on: July 16, 2020, 05:26:30 PM
Hi,
Your story is very, very familiar.  I also built my boards and while measuring voltages, my probe strayed slightly off the measuring point.  End result I also blew up R1.  As you realized too, there's no coming back from it!

I'd build another board completely AND replace all the Zeners.  Ask Bottlehead for the scientific explanation if you need/want the details.  I'm not an electrical guy.  I just know it's what it took to get my amp testing and working properly.
And.... use alligator clips whenever possible for testing.
Good luck,
Philip



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19319
Reply #3 on: July 17, 2020, 06:13:29 AM
Philip is correct, you really want a new board, board parts, and zener stack. 

At the bare minimum, you would need to replace R1, the LEDs that exploded, the MJE-5731A, and the PN2907 on the A side.  That might not fix the problem though, and desoldering and removing transistors from a PC board without damaging the board is far harder than building a new one. 

I suspect what happened is that your meter probe slipped and you shorted two terminals together.  It's much easier to test the OA voltage by probing the metal tab of the MJE5731A.

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline Bartguy

  • Jr. Member
  • **
    • Posts: 32
Reply #4 on: July 17, 2020, 10:01:58 AM
Thanks everyone. I'll order an new board and parts. Sigh. I was so close to finishing.



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19319
Reply #5 on: July 17, 2020, 10:13:38 AM
Also if the 5670 side of that new coupling cap somehow touches the chassis or connects to a circuit ground, it would blow up that board in that exact manner. 

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline Bartguy

  • Jr. Member
  • **
    • Posts: 32
Reply #6 on: July 17, 2020, 06:02:43 PM
That's important to know, Paul. If you're talking about the barrel of the cap, I thought it was insulated by the plastic coating. But it's on top of two plastic zip tie mounts and not touching the ground lug on the 5670 side. The zip tie is the only thing separating it from the 5W 22 ohm resistor nearby, however. It is possible that one of the clipped leads on the bottom of the C4S board could have been touching the side of the film cap. I'm positive that the pop happened when I touched the multimeter probe to the top of the board, though - not before even though the amp was powered up.
In the photo I posted in my other thread with the Audyn coupling cap and the big Jantzen caps that you said were glued, how are they insulated from the chassis? Is the glue enough to insulate those caps? I focused on making sure the Mundorf caps were really securely fastened to the top plate because of the warning that they could shake the terminal strip loose.

I've upgraded caps in almost all of my DIY gear (Bottlehead Stereomour, Bee Pre, Foreplay, and 5 Audio Note and Elekits) and never had an issue with film caps shorting anything. But I've focused more on keeping them clear of anything else rather than fastening them really tight.



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19319
Reply #7 on: July 17, 2020, 06:08:00 PM
The plastic cases of the caps don't need to be insulated from the chassis. 

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline grufti

  • Sr. Member
  • ****
    • Posts: 320
Reply #8 on: July 17, 2020, 06:20:53 PM
I see a long thick lead sticking out between the black cap and the number 48 in your photo. That could some day cause you grief. The cases of the capacitors will not. They provide insulation to the level of the capacitors rating and beyond for safety.



Offline Bartguy

  • Jr. Member
  • **
    • Posts: 32
Reply #9 on: July 27, 2020, 05:23:05 PM
Thanks, Paul and Grufti. I rebuilt the C4s board and zenner stack and all is good. Kaiju is paired up with Bee Pre in the main system and sounds really nice. Going to listen for a while to get a baseline before doing the DC filament upgrade.

Thanks again for all your help.