Testing the TK Tar Baby

[sparks]

  I  studied logic a long time ago so give me a little leeway here if I err.  Upon employing logic to predict solutions to a problem many times logic prooved to be a valuable tool for problem solving.  Just about as many times it was useless because reality does not always conform to logical thought processes.  I can make a prediction using logic that an apple I am buying in a store will taste like an apple .  It is a prediction based on logic.  I bought something that looks like smells like and is marketed as an apple so by employing logical thought processing I deduce that it will taste like an apple.  If I get home and it tastes like an orange logic has failed me.  Stuff like that happens all the time. I will say however that I believe intuitave thought processing relies heavily on subconscious application of logic to databases that on a conscious level we are not aware of.  The brain can rapidly remember similar problems and situations and even respond to minute chemical traces or body posturing etc.  This is why any successful businessman will not make an important deal unless whoever he is dealing with is in person. 

[TinselKoala]

Ahh...limits.

At last. I don't consider a research project completely interesting until I find out where some limits are.

I believe this limit is attained when one of the cheap Radio Shack RF chokes (rated at 2 amps) overheats from high current, shorts through its insulation and suddenly drops from 100 uH to 0 uH. This allows all that power circulating in the loop to bang through to the mosfets.

I still haven't found all the pieces. It sounded like a pair of firecrackers.

[Magluvin]

Hey T

Maybe the fuse is a bit high. Im working on a project at the moment and will be building me a loop device soon.
The IRF1010e are close in power to larger devices, bout 200w.

From fiddling with car amps, I can see its better to use a clamp to mount the transistors to the heat sinks, forcing the whole tab side evenly against the sink. When pushing the transistors hard, the top of the tab bolted to the sink, the tab itself can distort/bend the tab, and the part of the tab right where the chip inside is mounted might not be against the sink, even with heat paste, enough to get rid of the heat fast enough.

Not saying it would have made a difference here. But I noticed the screw impressions on your tabs. ;]

Was thinking of trying the larger disk caps from RS. 10nf 500v x 6.

Mags

[TinselKoala]

This wasn't fused.... so yes, a bit high!
No, I don't think better heat transfer would have helped in this case. But these are cheap, three of these for the cost of one PG50.

I agree that clamping is better than screwing, all things considered. But these heatsinks, thermal paste, and screws are perfectly adequate under normal circumstances. What goes on in this lab is anything but normal circumstance, though. If I'm not blowing mosfets, I'm not finding where the limits are.

[Magluvin]

So what are the limits?  ;]

Mags