Phoenix Gold Phorum

I have a few tantrum amps in my collection that are kind of my "white whales". I thought I had finally discovered what was going wrong, but I rebuilt my 500.2 for the second time this weekend, and it still isn't right.

A little back story. My first tantrum amp I bought to repair was a 600.4. I rebuilt the PS and one blown channel and ever since it has randomly blown outputs. I have gone back to this one many times over the years. I have now pirated so many parts off it that I may never fix it, but, I digress. I spent a little time on a Tantrum 500.2 a few weeks ago, I would replace outputs, it would randomly short a single output fet. I ended up putting it aside and working on the sapphire amp. In the mean time I did some research and it seems the IRF540N which I used in both these tantrums is very prone to spontaneous destruction. It looks like an almost perfect match on paper, but it is not. A lot of research on DIYaudio.com confirmed this, almost any amp that has had these transistors swapped in has had sporadic failures. The fix is to of course use the IRF540 with no "N". When I originally bought parts, the N suffix was the only available replacement. I think it was supposed to replace the previous version. IRF then brought the old version back just because the N suffix was not a good match for all applications (supposedly this holds for the IRF640N as well).

Anyhow, I bought a mess of IRF540's to have another go at the 500.2. I rebuilt all the channels and now it works, but it is still fuckered up. At low volume/minimum gain it seems to play reasonable well, but if you increase the output at all it immediately goes very distorted and then quickly degenerates into oscillation and no longer plays audio unless you reset the power to the amp. It also starts to draw a lot of current when it goes into oscillation.

I think that it is probably a problem in the feedback loop and I will look into it further, but do any of you have any advice on where to start?

Thanks,
Jason

Well, I found the issue. I feel sheepish even admitting this, but even though I had meticulously cleaned the heat sink and insulators, something had managed to get between one of the output transistors and the insulator. It was tiny, just a spec. I could see it with a 10x loupe after the fact, but not before. It eventually made a tiny arc on the sink. Even then the only reason I found it was because I unclamped that side of the amp to start checking for shorted outputs again.

Moral of the story, no matter how clean you think you are, double check everything on assembly. I think the spec got there the last time I had the amp out of the sink. I left the heat sink sitting out with insulators on it. I did not clean and reapply compound the last time because it was all fresh just a week or so before.

I'm on to a old Tantrum 1200.1. Pretty sure it is shorted outputs and not the Tripath module. Well, the original tripath module was bad, but I have a 103A that I think is good that I am going to try. Supposed to be backward compatible. We will see!

Later,
Jason

And that is why am hesitant to take any boards by themselves for repair. I just don't know how careful or tedious the owner will be when assembling them, not to mention I had a board come in for repair and turned out they used two TO220 insulators for the large Toshiba outputs with tiny areas of the device's sink exposed.

Good luck on the 1200.1 tantrum.

Well, the Tantrum 1200.1 didn't work out so well. I forgot, I had already replaced the outputs and with the tripath module installed, one half of the H-bridge mosfets are enhanced. Pull the module and they test perfect, so there is something bad in the modules. I have the cover off one of the modules and it looks like they have two half bridge drivers that are just epoxy blobs... That is beyond my capabilities to repair without more data...

So, anyone have any input on the tripath modules on ebay right now? There seem to be a plethora of them, most from China... Did they finally reverse engineer these things? I'm thinking of grabbing one for $20ish. Some of them say seller refurbished. It might be worth getting one and cracking it open to see how it was repaired...

Thanks,
Jason

If you could get the datasheet for the tripath that would help out tremendously as they usually always have a diagram of the internal components.

I have all the data sheets for all the tripath modules and I think some of the white paper as well. There is no description of the internal components or circuitry of the module.