IRF840 mosfet question

[Bieworm]

I might have done someting very stupid...
I built the TK imperial clone and followed the schematic mainly, except the reverb circuit, in which I copied the blackface reverb circuit. It's not that lush, but that is due to the lower voltage supply compared to BF circuits.
But back to the possible stupidity...
It sports that IRF840 N power mosfet in the tremolo circuit. I built this mosfet into 3 turrets, standing straight vertically. But due to the low chassis and the turretboard + mounting studs consuming 1/3rd of the room in height, I had to cut off the metal part of the mosfet. Tis should have been used to attach it to the chassis , insulated with one of those white rubbery sheets, no? After playing the amp for hours I haven't noticed anything bad, but is this bount to fail? Or is this a wait and see what happens kinda thing?
I have spare IRF840 mosfets, so I'd like to know if this is necessary to act now or wait...

FWIW right now it's performing really great!

[tubenit]

I've not used an IRF840 mosfet.  I've used a few IRF820 mosfets as cathode followers on turret boards and never had an issue. I've never attached the metal part of the mosfet to a chassis.  Look at the photo on this link:  https://el34world.com/Forum/index.php?topic=21873.0

I don't know if the IRF840 or using a mosfet as a tremolo would make a difference or not?  Hopefully someone who know more can give you a better answer.

With respect, Tubenit

[Bieworm]

My guess is that when the mosfet pops I don't have trem anymore. No other hazard.  But I would like to know if there's no other fails on the way then.
Guess I'll have to keep playing the snot out of the amp before gigging with it.
2 obvious things that give me confidence in succeeding this are
- covid: no gigs in the near future
- love the amp,  so playing it a lot is kinda bount to happen

[PRR]

I have never heard of cutting the power tab off a TO220. If I did it I'd expect to crack the seal, let damp air into the chip, and it fail in seconds to weeks. Maybe not. Good luck.

[shooter]

+1

Never seem it either, I have ground the pointy ears round, 1/8" maybe
my concern, if it's too tall, how close is it now, slight bump - arc - bump.......

[glass54]

Hi Bieworm
The IRF840 is a higher current spec of the IRF820. (IRF820: 500V/2.5A/50Watts/Rdson 3R, IRF840: 500V/8A/125Watts/Rdson 0.8R all at 25deg Celsius). As you are probably using a "Cathode" resistor in the range 82k to 100k, the current requirements (thus power) are low.
In this application, the IRF840 is a slight overkill BUT one uses what one has in their kitty.
Beware that the Drain (Pin 2 and TAB!!) connects to +HT, so I would not connect the TAB to chassis without a suitable insulator on mounting screw and screw and thermal insulation pad (usual TO22 mounting kit). I just tested one to confirm Drain connection to TAB as Datasheets don't show this.
I personally wouldn't cut the TAB off as it does help with thermal dissipation in free air 
Also see attached data if you are bored 
https://www.vishay.com/docs/91057/91057.pdf
https://www.vishay.com/docs/91070/sihf840.pdf
Kind Regards
Mirek

[tubeswell]

Without a heat sink those transistors won't be able get anywhere near the data sheet maximum power rating, and without the Drain tab there's no easy way to heat sink it. And if you've cut the drain tab off, and somehow the thing still works for hours on end, its because you didn't damage it electrically, and it isn't dissipating anywhere near enough heat to kill it. Measure the drain-to-source voltage and multiply it by the current (that you can measure through the load resistor) and you have the power dissipation. It probably isn't even 0.5W.

[Bieworm]

Thx guys. :)