Hotheaded? Check Motorhead's thermals.

There's nothing much going on in the Motorhead, dissipation-wise. The main victim of dissipation will be the LV1117 linear regulator that drops the supply down from 12V to 5V to power the relay. Beyond that, the A4988 stepper driver might become slightly warm but it's a PWM controller, so should be bothered by what we are asking of it, not least as it's running from 12V here.

The linear regulator on the Motorhead gets hot, not surprisingly. It's not easy to predict the thermal performance of an SMT device with any precision when it's almost entirely reliant on PCB area for cooling.

When running the unit yesterday, I was able to confirm that it's certainly running above ambient and body temperature. However, that doesn't tell me much, not least as the pain threshold for most of us is around 50C at most and I'd expect this thing to be running closer to 100C, against a max junction temp of 150C.

Let's check, using the thermal camera - but first some calculations:

• The relay is a "3V" version, with 64.3R nominal resistance. 
• The RP2040 GPIO are 3.3V outputs and there's a 5V-to-3.3V LDO on board the Waveshare RP2040-Zero 
• The 12V-to-5V LDO I have used to supply the RP2040 Zero is the 5V version of the LD1117 in the SOT_223 package.

This should suffer ~ 110C/W. What is the dissipation?

The worst case duty cycle is 100% ie when the relay is permanently enabled. At this point, the current in the relay could should be ~50mA. 

Yes, that is (just) within max ratings for the GPIO outputs:

...and the dissipation in the LV1117 regulator will be approximately Ptot = (12V-5V) * 0.05A = ~350mW. With an Rth.j-a of 110C/W, that might result in a rise of ~38C. This would be above the pain threshold but shouldn't be a problem for the regulator.

What does the thermal camera tell us? Looks like 78C ie a 53C rise. I guess the assumed PCB tracking isn't present. However, a case temperature of ~80C is fine.

The A4988 stepper driver is just about visible under the crosshair (you can just about make out the fins). It's not even breaking a sweat.

All looks well - but it's good to check.

Butchering the MIG torch - ready for action!

Firstly, I need to install the 3.5mm phono socket switch thingy. There aren't many places it will fit but this is just about manageable once I've chopped the switch body with the cutters.

Next, put the cover back on the MG welder.

And here we are. I'm just waiting for some 3.5mm mono plugs to arrive later in the day. Then I can connect the 12V supply up to the Motorhead unit and - who knows - possibly even have a go at using it.

The socket looks innocuous enough. I assume it will still function after being doctored / butchered to fit in the torch body. Time will tell......

Butchering the Arc Captain MIG200 to provide power to the Motorhead

I need to bury a 12V PSU somewhere in the MIG welder and connect it up to the spare pins on the front connector. Currently, only 2 are used - for powering the spool gun. That's handy.

Some double sided foam tape holds the 12V PSU in place. This was about the only flat surface I could find to fit it.

The connector can be removed easily enough, so I can solder the additional pair of wires on:

If you look closely, you can see that the lower pin is the +12V and the upper pin is the 0V:

With toroidal chokes fitted to both the input and output cables, hopefully that SMPS will behave when the machine is running in anger. Here it is with the lid ready to go back on.

Now I just need to connect up the 3.5mm switch wiring in the torch handle and the 12V supply to the Motorhead. Then I can try this thing out for real. 

Motorhead - system and mechanical design

I need to think about how I will house the Motorhead controls and integrate them with the generic MIG torch that was supplied with the Arc Captain welder. I also need to supply some power to the controller, ideally from the welder end of the cable.

Mangling the MIG torch:
Here's what we've got:

Comes apart easily enough:

The gas comes down the main "pipe", with the electrical power coming down the outer braid. Then there's a pair of wires for the trigger switch:

The other end has a generic Euro connector, which takes care of the gas, power and trigger connections in a quick connect connector system.

It also comes apart easily enough. 

The trigger wires are routed alongside the power braid, not inside the pipe itself.

The MIG wire comes along this Bowden cable:

The trigger "switch" is little more than a washer that shorts across a couple of simple spring loaded contacts.

So, I guess I need to insert the normally open Motorhead relay contacts in series with the torch trigger switch. For this, I will use a 3.5mm mono jack socket, the sort with the in-built switch that will leave the trigger switch operational with nothing plugged in - but will put the external switch in series when the plug is inserted. That way, the torch will work fine without the Motorhead but will assume Motorhead powers when plugged in. If you see what I mean.

I've ordered some Chinesium 3.5mm sockets. I will fit one of these into the MIG torch housing.

PCBA housing and interconnections:

Meanwhile, the Motorhead PCBA needs some sort of housing and some interconnections to the welder.

Here's the Oriental Motor (Japanese - but made in China of course)

The MIG cable is marked with a part number but I couldn't find any information on the internet. No problem, these torches are cheap as chips, so if I need to get another one to butcher it's not a major outlay.

Anyway. Over to Fusion 360. Here's the imported PCBA positioned magically in mid air, roughly where I imagine it needs to be in the final assembly:

And with a simple box-like housing, it's looking a little more like a workable concept. The dimensions are taken from the NEMA37 motor, so that the housing will sit neatly on top, flush with the sides. I won't win any industrial design prizes for stuff like this but for now it's a start:

Right - let's make one of these things. It's not a finished design by any means but I want to get on and try the thing out before getting too far ahead of myself.

Looks OK:

And not surprisingly, the PCBA fits inside nicely:

With some double sided foam tape, it becomes one with the existing motorised head:

Now to sort out matters "at the other end of the cable"...