Extending (relocating) MOSFET from amp board

I got a couple old Dayton 250 watt sub plate amps that have given lots of issues over the years w/ the amp section.

Finally tired of repairing the amp, I decided to purchase a couple cheap class d amp boards online and will re-use the xover, preamp, power supply in the plate amps and install these little class d amps in place.

Here is the exact ones I purchased: https://articulo.mercadolibre.com.mx/MLM-1368826029-modulo-amplificador-clase-d-irs2092-500w-_JM

Nothing special, these are the dime-a-dozen IRS2092 boards available all over from ebay, ali express and other sites.

The problem: These boards have two MOSFETSs mounted (to the PCB) under a tiny heatsink w/ fan. My plate amps have the normal external heatsink with interior space for mounting the active devices.

What I am proposing to do:

1) Remove the heatsink and clip the MOSFETS from the board
2) Add 2-3 inch long extension leads to the MOSFETS and re-attach (the new leads) to the amp board.
3) Mount the MOSFETS directly to the heatsink as the original amp was. I have insulators and mounting pads to ensure they are electrically isolated.

The question: would there be any issue with extending the leads like this? Do the MOSFETS have to be close to the board, or can "extending the leads" be safely done? I hear stories of capacitance issues and other similar issues with MOSFETS.

Thanks!
 
Normally this is a recipe for disaster, especially with high-power switching design. They tend to be much more susceptible to parasitics than class AB, so extending the leads of the FETs will be worse. It might work, but I don't think anybody would recommend it.

A more obvious solution from my POV would be to simply remove the fan and heatsink and replace them with either a mounting bracket (aluminium angle) to transfer heat to your plate/heatsink, or replace the heatsink entirely with a solid block of aluminium to accomplish the same thing?
 
I agree. In a Class D amp there is a carrier of several KHz whith harmonics ad infinitum and although MOSFET are vety fast, any piece of extra wire will add capacitance and inductance. So the added parameters not only future load the driver, this reduces MOSFET's speed addind too a more important radiated field. So IMO keep it as is.
 
What if I tried this:

1) Remove/clip MOSFETS as explained - currently mounted to the TOP SIDE of the PCB
2) Carefully re-solder them to the BOTTOM side (pins extend through the PCB already)
3) Fold the MOSFETS flat so that I screw the board down to the heatsink - sandwiching the MOSFETS to the heatsink

The MOSFETS are mounted in the middle of the board, a very akward place that does not allow easy extension of the heatsink.

Check out this YouTube video - fast forward to around 5:00 in that video, you will see the guy has taken the heatsink off and showing the MOSFETS.
 
That should work better. The reason that it's not done often is that unless you have mounting holes in the PCB you can only properly clamp the transistor to the heat sink by putting pressure on the PCB (and bending it), which isn't ideal. Also, air circulation is a problem so the transistor is going to run hotter. It's probably going to work though.
 
Nisbeth - thanks for the idea.

A thick chunk of aluminum angle (L) from the door and window shop down the road should work I think.

Board would mount 90 degrees from the inside of the main heatsink.

Technically, one MOSFET would be further away from the main heat sink than the other - and heat from the furthest one would contribute to the one closest to the main heatsink.. but.. this IS class D and, if the aluminum angle bracket is thick enough, this should not be an issue - correct?

The alternative would be a large piece of aluminum C-channel and mount the entire board up-side-down. One side of the C to the MOSFETS, the other side to the main heatsink. But this seems a bit un-needed, no?

Again, this is Class-D.. I do not imagine the MOSFETS are going to run as hot as typical class a/b. Plus, I am going to be running this board on the lower end of it's design range (+/- 58 volts -vs- the 70-90 some run it at). I only need/expect about 175-200 watts out of it.

Pushing some cheap-o Dayton "classic series" 15" subs
 
More I think, the angle (L) idea makes the most sense... it also provides a way to secure the entire board to the back of the plate amp.

The factory mounting solution for this amp board is a hole that runs throught the entire PCB and clamps the MOSFETS between the heatsink and top-side of the PCB.

I got some insulating pads AND... since the mounting screw/bolt runs BETWEEN the two MOSFETS, no need for the nipple.

I plan on updating the old thread found here: https://www.diyaudio.com/community/threads/dayton-spa250-plate-amplifier-repair-help.258020/page-4

Going to take some pics for all along with some notes on the wire-up for others who have these flakey amps. They are notorious for failing, and this is a "modular" solution that will make future repairs as easy as purchasing another $15-$20 module.
 
You may be able to drill two holes, in a neutral position, into the PCB. So you can screw down the MosFet's which are soldered to the underside. The PCB must be secured by spacers, of course. This is a very common way of mounting such modules. You bend the MosFet over, fasten them on the heat sink, secure the PCB and then solder them from the upper side. This will not be perfect, but keep them in position. Now remove the whole thing, Moset's and PCB from the heat sink and solder them again, this time final, from the under side. Mount them on the heat sink again, don't forget isolation and heat transfer compound.
May seem a bit complicated, but works fine.