Welding a thermocouple to a TO-220 tab

Hello gents,

I'm trying to weld a thermocouple wire to a TO-220 tab. Anyone done it successfully?
I tried charging a 560 uF capacitor up to 400V, with one side electrically connected to the TO-220 tab, and the other, to the thermocouple wire, then touching the wire to the tab to get a big arc. I get a pit, but it doesn't stick. I tried the same thing, sticking the thermocouple wire to a piece of 12 gauge stranded wire, and it stuck nicely. I wonder if the problem with the TO-220 tab is the thermal mass.

My wire is Type K... I just ordered some type T, which is solderable. I don't know if I can solder it to the tab.

Anyone use a small welder successfully? I wonder if this small resistance welder would work:
https://www.amazon.com/Pulse-Welder-Battery-Charger-0-1-0-2/dp/B01MS27R08

Cheers.

You want to tack-weld. I believe the pieces should not have a big difference in thickness, otherwise the thin piece vaporizes and you get a pit instead of contact.

One thing you might try is to 'splatter' a bit of copper wire close to your pieces. Send house current through a few thin wires in hopes that the rapid (almost explosive) melting splashes into the space where your pieces touch. This might require many attempts.

- - - Updated - - -

The tack-welders I've seen have the electrodes pressing the pieces together at the same moment the electrodes send current. Can you arrange to do the same thing with your pieces?

"The tack-welders I've seen have the electrodes pressing the pieces together at the same moment the electrodes send current. Can you arrange to do the same thing with your pieces?"
Hmm. Thing is I don't have a high current switch to apply the high voltage from the bulk cap. I close the circuit by touching the thermocouple to the tab.

Are you sure you want to do this? Because you'd be adding
a third junction with maybe unpredictable results (if
"thermocouple wire" is the two wires with the weld-ball -
if it's only the copper or only the constantan, what's the
point?).

What's wrong with a tiny dab of high temp epoxy? Arctic
Silver if you want a low thermal impedance connection?
Or heat sink compound, with another low mass form of
mechanical "keep"?

jcuadra said:

"The tack-welders I've seen have the electrodes pressing the pieces together at the same moment the electrodes send current. Can you arrange to do the same thing with your pieces?"
Hmm. Thing is I don't have a high current switch to apply the high voltage from the bulk cap. I close the circuit by touching the thermocouple to the tab.

Click to expand...

Take a copper wire 3-4mm dia, press it rapidly to the tab with the thermocoule weld (the small ball like junction) and you need to press fast and hard. Also use high capacitance and lower voltage (with the same joules). I guess the tab is made of copper and is difficult to spot weld in this manner.

Tabs are copper but plated with "stuff" that can make
a spot-weld fail to adhere (although solder will).

Another option could be to find small tube stock just
barely greater than thermocouple ball diameter, solder
it down (high temp solder, if you're really worried about
temperature excursion?) and swage the tube down at
from the ends inward, once the thermocouple end is
inserted. Or you could form thin copper sheet around
the thermocouple and try to solder / spot weld that.

Problem with trying to spot weld to copper is, its
conductivity prevents developing much heat in the
workpiece to get a true melt. All heat has to come
from "the other piece" or a lousy lossy interface.

What about rolling a thin copper sheet around the
thermocouple, with a "flag" the size of the TO220
body and a matched mounting hold, and sandwich
this between the case and the heatsink / mount
plane? Squish it tight about the ball and you ought
to have more than good enough thermal contact,
ball well surrounded by source and conduction >
convection losses.