Home made TIG cooler

Here is a TIG cooler I made out of bits and pieces, I cut up an old greenhouse water heater for the aluminium sheeting to make the case and the fluid container as I no longer used the heater. Made some hose fittings out of aluminium and brass rod on the lathe, pulled apart an old discarded air conditioner for the fan and the radiator.

New parts were an electric motor driven low volume, high pressure pump, a pressure gauge, 2 switches, one for the fan and one for the pump, a fuse holder, a cheap aquarium digital temp gauge, assorted S/S ring clamps and a ball tap to adjust the flow pressure. All up it cost about $100 to build.

The tap is adjusted to let through about 3.5lts/minute at 26PSI and the holding tank has about 17lts of fluid to which I added coolant to stop corrosion. The only complaint is that the pump is a bit noisy but at least I won’t forget to switch it off. Only modification after I built and painted it was to add a fan shroud, which then forced about 3 times the amount of air through the radiator than before it had it.

Works well

Apart from a pedal, a cooler this the next best thing to sliced bread. Nice job

Is 26psi high enough pressure? I have been looking into making a cooler, and it seems 50psi and something like 4.5ltrs a minute flow are generally the figures recommended? Also whats the reason for on-off switches, as this means it would be possible to weld without the cooler working, which will very quickly burn the torch out if on high amps?

Engineer said:

Is 26psi high enough pressure? I have been looking into making a cooler, and it seems 50psi and something like 4.5ltrs a minute flow are generally the figures recommended? Also whats the reason for on-off switches, as this means it would be possible to weld without the cooler working, which will very quickly burn the torch out if on high amps?

Click to expand...

Hi Engineer, As far as I found out 60PSI-70PSI is the max before most of the internal hoses become borderline blowout, when I was looking at doing this I came across a few units that other people had made and a lot of them used aquarium pumps, most aquarium pumps are high volume and low pressure, max out at 3-4mts head height so that means they can only produce about 5-7PSI so hooking one up to a tig torch would only have given a very low pressure and a resultant very low volume of fluid flow, as you need the pressure to force the fluid through the narrow tubes inside a TIG gun, but in any case they claimed they had used these setup for 5-10 years like that without problems, occasionally just replacing the pump but their torch was fine. With the pumps they used they would have been lucky to even get 5 PSI out of them

My pump will put out more pressure but it doesn't seem to need anymore than the 26PSI as it seems to work fine and the water doesn't warm up much at all, 26 PSI should go up to a head height of about 50ft.

I should crank up the amps all the way and see if that changes anything but I have had it up to 170 amps without any dramas at all. I did about 3 hours the other day on about 125amps and the temp didn't rise at all.

Will do some more TIG tomorrow so will test it out, will let you now if anything does change.

As for the on/off switches, seeing as there is water and electricity involved I decided that it would be quicker to switch it off from the front instead off trying to pull the plug out from the back should it spring a leak inside and make the case "live".

I do have RCD's in the garages circuit box but it doesn't hurt to add extra protection.

Not common over here, but in the states it appears alot of people just connect main water to their set, and poke the return in the drain. So I would say 26psi is more than enough, as mains water is usually only 1-1.5bar (14.8-22.2psi)

As long as you have checked to amount of flow per minute, and its ok then I guess the low pressure doesnt matter.

Mains water flow is generally likely to be around 20-25 lpm, and connecting this directly to a TIG torch needing 5lpm may well result in problems.

I would think anyone using mains water simply turns on the tap until the amount of water being returned looks ok, and leaves it at that.

Unfortunately in common with coolers fitted with on-off switches, its possible to forget to switch the water on, which will unless you have a fuse in the torch lines, quickly melt the hoses, if you are running high amps.

Engineer said:

Unfortunately in common with coolers fitted with on-off switches, its possible to forget to switch the water on, which will unless you have a fuse in the torch lines, quickly melt the hoses, if you are running high amps.

Click to expand...

It would be simply enough to "interlock" it using a pressure or flow switch to enable the Tig trigger.

Neat job. CK worldwide just state a min flow of 1 litre/min, Weldcraft say

Water cooled torches require cooling water flow of 1 quart (.94 liter) to 11/2 quarts (1.4 liters) per minute at a maximum of 50 PSI pressure.

Click to expand...

Or to connect the torch trigger lead to the water cooler, then the cooler to the welder, that way the welder can't start unless the coolers running...

A simple double pole relay would do...

Omniata said:

Or to connect the torch trigger lead to the water cooler, then the cooler to the welder, that way the welder can't start unless the coolers running...

A simple double pole relay would do...

Click to expand...

That method would still enable the trigger if the pump failed..

Robotstar5 said:

That method would still enable the trigger if the pump failed..

Click to expand...

How about setting the relay up so that it's only connected when the pumps running?

Leave the pump on a manual switch and link the trigger through...

That way it won't arc until the trigger button is pressed, but can't unless the pump is running...

ie: the pump is driving the relay...

Omniata said:

How about setting the relay up so that it's only connected when the pumps running?

Leave the pump on a manual switch and link the trigger through...

That way it won't arc until the trigger button is pressed, but can't unless the pump is running...

ie: the pump is driving the relay...

Click to expand...

If you mean use the power feed to the pump to operate relay, it will still work with a pump failure, pressure/flow switch is the best way to go

Remember the pump needs to continue after you have stopped welding to remove latent heat from he torch head

At the risk of seeming to make this sound too simple, is there any big issue with connecting the cooler to the welding sets power lead, so when the welder is plugged in the cooler is running? Switches of any kind seem to be source of potential problems, and not needed here as when the welder power is switched off so is the cooler.

As to cooler flow, you need to remember the hoses in a TIG torch are pretty small, which means most pro coolers have pumps which can flow around 5LPM open, with a pressure of around 50psi. A quite neat cooler can be made using a stainless steel catering pan, and the Bernard coolers use something very similar.

Engineer said:

At the risk of seeming to make this sound too simple, .

Click to expand...

None what so ever.

I built a cooler a while back. I first tied a fish pump, and as others have said, too low a pressure. I then tried a central heating pump. No good at all.

The last one I tried was a beer pump. That one worked

I wonder why was the central heating pump no good? Luckily I was able to get hold of a pump from a Kempi cooler that had been scrapped, but I dread to think what one of those would cost new!

Looked at central heating pumps, before I found the Kempi pump, and it seemed as though they would work, but an adjustable bypass circuit might be needed, as the flow rate seemed as though it would be much too high?

Got a rad and cooling fan from a beer chiller, and this was like the Kempi pump FOC, so only costs so far are for a stainless steel gastronorm (catering pan!), some quick release hose couplers, and a cheap pressure gauge from Ebay.

CK WP20 torch wasnt cheap though, and still need to buy a leather sleeve for the hoses, and probably a fusible link.

They need a head to create the pressure

Is that not possible if you use a vertical tank, with the pump mounted at the bottom? Central heating pump seems on paper a good choice, as they are easily available, and well made. Seen all sorts of weird suggestions related to pumps, and a fair amount of them dont seem to figure how important reliability is in this application.

Engineer said:

Is that not possible if you use a vertical tank, with the pump mounted at the bottom? Central heating pump seems on paper a good choice, as they are easily available, and well made. Seen all sorts of weird suggestions related to pumps, and a fair amount of them dont seem to figure how important reliability is in this application.

Click to expand...

Correct on all counts.

Engineer said:

Is 26psi high enough pressure? I have been looking into making a cooler, and it seems 50psi and something like 4.5ltrs a minute flow are generally the figures recommended? Also whats the reason for on-off switches, as this means it would be possible to weld without the cooler working, which will very quickly burn the torch out if on high amps?

Click to expand...

Hi Engineer, cranked up the TIG today to about 240 amps and did a pass on both sides of a 12mm steel T joint 200mm length each side, the fluid temp rose by 2 degrees above ambient, and the head was just warm up to the where the cup fits in, did a further 4 hours welding at about 130 -140 Amps and the fluid temp got to 25 degrees which was 2 degrees above ambient by that stage, so it all works as it should at my pressure and volume output. Hope this helps.