[luvdunhill]

So, what is this thread about. Let's show how powerful the DIY community is here on Head-fi. This would be a first, on Head-fi but need not be the last time. Why, swt61? Why not.

The details. At the recent Houston meet, we were listening to swt61's Beta22 at full volume, as the source was extremely low output. Then someone plugged in the most coveted headphones that exist, namely the Fostex T50v1 and the amp immediately caught on fire. Here's what happened (all pics are high resolution, zoom in for the gory details):



Removed the device, and here's the damage:



... and another angle:



now, the following devices were determined to be bad and removed. Lots of flux was added to aid in the removal. I'm not sure about CR6, but I'm inclined to replace it, and would appreciate any opinions on whether or not this device should be okay or not (and if there are any other devices I should check or replace). Yes, the carnage was massive:



Even the rear side of the board, and the case itself took damage:



Finally, everything was removed, holes cleaned up and ready for the new parts:



I will be posting a part list soon and ordering the parts from Mouser and the extra devices from amb where appropriate. I'd like to rally everyone together and brainstorm on the best way to repair the PCB tracks and determine what else (if anything) might have been damaged. swt61 is a great guy and here's hoping that the community can get this amp back in his hands as soon as possible!

 

[Pars]

Why not just get a new board and transfer everything that is good over? Much better than any trace/pad repair I know of (though I am certainly no expert by any means).

 

[spritzer]

That's some damage the T-50 managed to do. Full volume or not this shouldn't happen as the T-50 isn't that tough a load at least compared to the speakers that mine is pushing these days.

A new board would be a lot of work so why not just use wires on the underside to jumper the damage. I'm a man of simple solutions though...

 

[n_maher]

I have quite a collection of beta22 spare parts, including output devices so please send a list of what you need once you have it. I'll start by volunteering a matched set of the output fets. PM me an address and they'll be on the way tomorrow.

 

[naamanf]

Was the T-50 plugged into the 1/4" SE Jack? I am guessing it was and the volume was turned up. When being inserted the 1/4" causes a momentary short from ground to the right channel. Poof up in smoke it goes. Don't ask me how I know this but I can say that I have verified it

 

[spritzer]

The T-50 is SE wired so a short in the jack would be a probable explanation.

 

[luvdunhill]

Quote:

Originally Posted by naamanf /img/forum/go_quote.gif Was the T-50 plugged into the 1/4" SE Jack? I am guessing it was and the volume was turned up. When being inserted the 1/4" causes a momentary short from ground to the right channel. Poof up in smoke it goes. Don't ask me how I know this but I can say that I have verified it

naamanf: yup, that's exactly what happened!

Thanks Nate for volunteering to help! I'll send you a PM once I get a list of what I need.

I'm not sure what to do. I could easily use the center leg of the MOSFET and connect it to the remaining copper trace by just running the leg on top of the board...maybe add a bit of heat shrink. Then, I could put the 2 resistors on the bottom of the board and then connect R32 up through the board to the top side and then connect it to the MOSFET directly.

I need to figure out what type of resistors R34-R35 were, as it would be nice to match them to the other 3 boards.

 

[spritzer]

The rest of the resistors are RN55's so those two could be the same. The other boards should tell you all you need to know. Creative bending of leads would make the amp good as new, I would also install a DIN socket or something...

 

[naamanf]

Quote:

Originally Posted by luvdunhill /img/forum/go_quote.gif naamanf: yup, that's exactly what happened!

I think maybe there should be a little warning somewhere about this. I warned all the people I built B22s for to never plug/unplug the SE jacks without switching the outputs off first. Doing it with no source/volume should be sufficient but better safe than sorry.

For the board that looked just like that one I just used wires to jumper from the MOSFET to parts it attached too. Better than trying to reuse and already stress trace.

 

[luvdunhill]

The resistors I'm not sure of are blue and have traditional color coding markings. Look at the last picture above, and look at the background and you'll see them from one of the other boards.

I'm surprised it took out as much as it did. naamanf, what all did you theoretically loose when you theoretically saw this same thing happen?

 

[naamanf]

Just the resistor (R32) which cooked (because when the MOSFET fails it just shorts itself out internally) and both the + side MOSFETs. Since they are matched of course they all need to be replaced.

Edit: R32 should be a 75ohm resistor.

By saying I verified this, I verified it can happen when being plugged in and when being unplugger

When being unplugged the MOSFETs just fail with out the resistor cooking because there is not place for the current to go. Of course if you then plugged something in....

 

[jantze]

Quote:

Originally Posted by naamanf /img/forum/go_quote.gif I think maybe there should be a little warning somewhere about this. I warned all the people I built B22s for to never plug/unplug the SE jacks without switching the outputs off first. Doing it with no source/volume should be sufficient but better safe than sorry.

Would ε12 fix this? I have 4 x β22, 2 x σ22 and 2 x ε12 in my balanced β22 build. My plan is to use Neutrik Combo Series XLR chassis connectors, so that I can use one balanced headphone or two unbalanced headphones.

 

[Pars]

Quote:

Originally Posted by luvdunhill /img/forum/go_quote.gif The resistors I'm not sure of are blue and have traditional color coding markings. Look at the last picture above, and look at the background and you'll see them from one of the other boards.

Those appear to be the metal oxide resistors and look just like ones on the sigma22:

R34, R3522W 5% metal-oxide resistor 0.47Ω

To the poster above: I'm not sure if an e12 would protect you from this or not. It is designed to disconnect the output in the event of DC offset exceeding 70mV. When you plug a SE headphone in and short the dignal to ground, it MIGHT generate enough of an offset to trigger the e12. Whether it switched in time to save the FET or not??

 

[naamanf]

Quote:

Originally Posted by jantze /img/forum/go_quote.gif Would ε12 fix this? I have 4 x β22, 2 x σ22 and 2 x ε12 in my balanced β22 build. My plan is to use Neutrik Combo Series XLR chassis connectors, so that I can use one balanced headphone or two unbalanced headphones.

I am going to say it's either not going to work or the ε12 relay is going to be shot. But I could be wrong.

 

[shellylh]

Unfortunately, I have no knowledge in this area so all I can add it moral support. I would be happy to contribute to a parts fund (via check or paypal) if one was started though.