Converting TO-263 to TO-220?

Question

I’m trying to repair an amp where the power MOSFETS (HUF76639) are no longer available in a TO-220 package. They’re currently available in TO-263 package. I’m thinking of adapting the TO-263 package as a solution. The setup is pressing the TO-263s against a vertical heatsink with a bar, insulated by Sil-pad. See photo belows. The G and S leads would be bent (slightly) to clear the heatsink. A 21 AWG wires would be soldered from the leads to the board. Wires would be used to alleviate mechanical stress (between the board and TO-263) on the soldered connections. The questions/concerns I’m running into are

1. Will the TO-263 case withstand the mounting pressure?
2. The drain lead is ~1.5 mm long, barely enough lead to solder it seems. Is this too short?
3. Has anyone tried out the above setup or similar and had success?
4. Any obvious limitations that would make this venture too risky?

The power MOSFETs on this amp usually stay below 50C. See links for TO-263 and TO-220 case specs. I couldn’t find a datasheet on the TO-263 case mounting pressure. I’m guessing is doesn’t exist since the package is designed to solder onto a board.

[

Device - HUF76639S3

TO-220 Mount Pressure

TO-263 Package

TO-220 Package

I used this Molex connector as a sort of cast form soldering the leg to the middle pin, and to align the other two legs for soldering. https://www.mouser.com/ProductDetail/Molex/560124-0101-Cut-Strip?qs=1YYxJJMLJH1jZhYDkeoY4Q%3D%3D

So far so good after 1 week. Heat cycling is my main concern.

Answer 1

The G and S leads would be bent (slightly) to clear the heatsink.

Just a caution, do this carefully, as bending the leads against the body will stress them quite a lot. Get a pair of thin-jaw needle-nose pliers and grasp at the base of the pin, then bend against that. That'll be a safe way to straighten the pins out (or bend them wherever else you need).

And then tacking on extension leads of solid wire or whatever to reach the board, is a fine strategy here.

1. Will the TO-263 case withstand the mounting pressure?

I don't see why not; it's made of the same solid encapsulant as anything else.

2. The drain lead is ~1.5 mm long, barely enough lead to solder it seems. Is this too short?

I'd probably avoid using the pin nub, just because it's short and small, prone to breaking, and may have encapsulation on it making soldering difficult.

You can just as well lap a wire on the top side of the tab; enough sticks out to solder onto, and the wire length likely doesn't make a difference here (I might be concerned if this was an SMPS). There is some risk of solder flowing around the backside of the tab; avoid getting flux here, and if it ends up lumpy (mind that the native tin plating (when applicable) may itself flow on heating), you can always file or lap the surface flat/smooth again. (If using a file, use a fine flat type; if lapping, use a medium stone that's known to be reasonably flat, or the old wet-or-dry sandpaper on plate glass or the like.) Not that it'll make much of a problem with the Sil-Pads there, really just as long as it's not blobby, it'll be fine.

3. Has anyone tried out the above setup or similar and had success?

I can't say I've seen many D(2)PAKs used in maximal power dissipation setups (i.e., they're rated more or less identical as TO-220s, nevermind they're almost always surface-mounted making this spec useless), but aside from the hand work to fit them here, I don't see why not.

And it doesn't sound like this is a very powerful application anyway.

4. Any obvious limitations that would make this venture too risky?

Not really, other than the jankiness of the solution, which I suppose you're already aware of. I would prefer finding a substitute; but I appreciate that may be easier said than done, or beside the point ("rolling" different types), and in any case isn't the object of this question, so that's fine.