Markbass Hot Coil

[drewl]

Okay, have a Momark 800, that uses two IRFP264 mosfets with one of the output filter coils overheating.


It melted C86 which I believe is .68uf and also the .33uf connected to the 4.7ohm 5w section of the filter, that was bad.

I replaced all the damaged/bad caps, tested the others in the output section and the larger coil is still overheating.
It's not as bad, and the high frequency oscillator level is lower than it was before on the output

On the one end of the coil that's overheating the high freq signal is about 220Khz at about 200v peak to peak
The other side hi freq is the same level as the output signal, it goes through the 2nd coil, the two 1ohm 5w resistors then to the output relay.

There's about 1v DC on the hot coil, is it possible a leaky FET could be causing it to overheat?

I tried another coil close to the originals' value of about 15uH, it also gets hot


I was referencing this informative post by Greg Robinson from the Markbass thread below
Thanks


"Sorry for the necropost, but I just repaired A CMD 102P (same amplifier, different speakers in the combo) with the same fault. Figured here was as good as anywhere for posterity.

At least in my case, C86 - 680nF 400V failed first, causing R53 and R54 (4R7 5W) to overheat, leading to C87 (330nF 250V) and C88 (100nF 400V) melting, as they are right next to each other.

All these components, along with L2 and L10, form the class D output filter. L2 and C86 form the first pole, L10 and C88 form the second, and C87 is in series with the parallel pair of R53/54 to form a Boucherot cell/Zobel network. Without C86 to work with L2 to do the bulk of the filtering, R53/54 overheat, with escaped magic smoke and extreme EMI emission being the result."

[Jon Snell]

"On the one end of the coil that's overheating the high freq signal is about 220Khz at about 200v peak to peak
The other side hi freq is the same level as the output signal"

There is your answer, the coil has a shorted turn, causing eddy current that heat up the coil and reduce the inductance causing poor/no filtering.
Rewind/replace the coils.

[drewl]

That's the answer?
Been a while since I studied coils.....


I tried another coil that is about the same 18uH, but it does look burnt inside like the original
Any idea about what value it should be?

I'll search for another
Here's the board for comparison

[drewl]

No matter what coil I try they all get hot.

Could the problem be the smaller coil heading to the relay?
It measures about the same at 16uH

[drewl]

This is the output filtering

[Jon Snell]

The output chokes need to be around 33uH at around 6 Amps for a 400kHZ output stage so for 200kHZ 100uH is near.
16uH is good for around 800kHZ.

[drewl]

Thanks for your help and research

So what you're saying is both coils are bad, reading too low causing the Eddy currents generating heat.

Well I'll try some coils as close to 100uH as I have, for some reason I don't think it will help.
But I'll try.
Thanks again

[Jon Snell]

Make sure there is no DC offset on the output. DC offset will destroy the coils as will square waves from an overloaded amplifier.

[drewl]

No DC offset, gets hot while idling.
Ugh, this is driving me nuts.

Tried a larger coil the reads about 120uH, still gets hot.

A smaller one that measures about 78uH works best, takes longer to heat up, but still will overheat.

The search for the perfect coil continues.

[Jon Snell]

"Heats up", with or without a load?

[drewl]

Both.
While idling, no signal it still gets too hot.

[Greg Robinson]

Have you tested for a DC short to ground on the output leg of the choke? Or after the relay?

[drewl]

Yes sir, I have, output path measured good and amp was operating fine except for the coil overheating

Just to wrap this up, I was reading some engineering forums on class D amps and this appears to be a common problem

I wired two 70uH coils in series and that seems to have done the trick
Puts out plenty of clean power without the coils heating up.

Don't know what the value of the original was supposed to be, or if it was damaged from severe overheating

Thanks for the help Jon.

[Mars Amp Repair]

Hello folks,
I must say I just figured about these DClass amps that are such a glued-glopped together mess that anyone is attempting to repair them! Other than obvious issues, I haven't ventured much further into them as I figured items like transformers & coils would not be available.
Perhaps I'm going to have to change my state of mind around these little buggers.
As always I really appreciate this forum. It's the best. g

[drewl]

Although these are a pain to work on, due to size and the goop, most of the problems are typical as found in any amp.
bad preamp stages or opamps, blown output mosfets etc.

I'd like to goop some of these designers.....