How to calibrate a MIG set.

[Jim Davey]

As I had to fit some meters to a new Tech Arc that we sold today, I thought I'd take a few pics and show anyone interested what's involved. (This is my setup, others vary and there are many valid ways to do the same job).

Here's the machine in question, interconnection cable is straightened out and is in circuit so the results are as accurate as possible. The yellow box on the left is my load bank, this consists of a bank of resistors that I can switch in or out with the big black levers on the front.

Above is my Dinze to welding cable adaptor, this has a toggle switch connected to the trigger pins allowing me to stop and start the welder as needed. The white box contains a 600 amp 60mV shunt for super accurate reading of current. It has a couple of banana plug sockets on it for quick connection to my multimeter.

I use two multimeters, one measures volts DC and I connect it by pushing the probes through the insulation in my cables directly into the copper. Positive to the euro socket adaptor. Negative to the cable going to the earth return socket, this machine has two, either will do.

 

[Jim Davey]

Once the machine is connected up I switch the set on, dial up about 100 amps on the load and let it warm up a bit while I start filling out the paperwork, usually before any work of this nature I'd kick of by testing the machine for electrical safety but as I did that yesterday when I fitted the mains plug I thought I'd skip to the calibration part of the test.
Here's the gear...

I have a pair of multimeters, a current clamp for double checking, a thermometer and a tacho for checking wire feed speed.

The clamp meter has a nifty removable magnetic display which makes taking the readings a much easier affair.

The new meters were about 15 amps and 3 volts out, on this particular meter kit there are small PCB mounted trimmer pots on the back of the kit that you adjust with a screwdriver until the meters on the set read as close as possible to the readings you get from your own calibrated test meters.

Keep going until it's close, then try high low and medium amps/volts and see if the meters are better or worse in any particular area, you might need to adjust up or down to get a meter kit to be within tolerance through the machines entire output range.
Once the meters were adjusted to the point I was happy to start testing I start by measuring max OCV (open circuit voltage).
All the relevant test info is on the rating plate on the back of the set.
(Ignore the fact this machine has the wrong sticker, for some reason it's badged as a single phase 240 volt machine!)

We can see from the plate its a 40-450 amp machine with a welding voltage of 16-36.5 volts.
OCV is 17 to 48.
Now, this is the important bit as far as the test is concerned, for a Standard Grade calibration on a set with digital meters the tolerance is 2.5 percent of maximum no load voltage.
And 2.5 percent of maximum rated amperage.

Ie.

Volt 2.5% 48 = 1.20 volts

Amps 2.5% 450 = 11.25

What this means is the meters on the machine can be up to 1.20 volts out and 11.25 amps out, quite a big tolerance on amps, not a lot on the volts.

OCV measured we can proceed with the rest of the test. The procedure is start low and work your way up through the amps volts from minimum through to maximum and three evenly spaced setting in between. Here's an example on my test sheet.

You do the same on the voltage settings. The voltage steps are not chosen randomly but based on the amperage you should expect a given voltage for any MIG, TIG or MMA machine. I can never remember the A/V correlation so I use a simple crib sheet. Obviously min voltage on this machine will be 16 through to the max of 36.5.
You can see the amp/volt correlation for mig, tig and stick sets plus some common tolerance plus and minus calcs for reference.

 

[Jim Davey]

It's then a case of inputting settings on the MIG set and adding or removing load and trying to get the meter on the machine to read the amperage or voltage you are looking to test it on. This can take a bit of shuffling around with the switching but with a combination of coarse and fine steps on the machine and the load you can normally get it very close if not spot on.

Once you've got the amps or volts where you want them on the set, record the readings off your meters and trigger the machine off, then repeat allowing the meters to stabilise for a few seconds, I usually do this three times but two is enough. This ensures the readings are repeatable and consistent.

Meter on the right is reading volts, the one on me left shows millivolts but this is from the shunt box and equates exactly to amps if you move the decimal point by one space to the right, so 13.48 mV is actually 134.8 amps. This is more accurate than the clamp (which is very good, but not as good as a shunt as you can see from the picture it's reading a little high.

 

[Jim Davey]

That's enough of welding sets for one day. More on measuring wire speed tomorrow if I can be a**ed

 

[dan.taylor.1]

How much does a shunt cost to get gauges on a mig that hasn't got them?

 

[Jim Davey]

A shunt is cheap, about £15 to start for a budget one. It might need fettling to be accurate though.

 

[Maker]

I love the foam lined case for all the multimeters, is there a name for that stuff, it looks like it's made from loads of squares?

 

[Jim Davey]

Pick and pluck I believe, it came in the case which is a good copy of a Peli case, made for RS Components but I got mine from Solent Plastics in Hampshire, I want a few more too while they've got them as they are tough as old boots but quite reasonably priced.
The squares are half cut and you just prise them apart a little and you can make quite neat holes for things.
Only issue is it's easy to catch an edge as you place things in the foam and tear out squares unintentionally but it's good to gauge layouts of stuff until you get around to having some tougher foam CNC cut (never gonna happen lol).

 

[Maker]

it's good to gauge layouts of stuff until you get around to having some tougher foam CNC cut (never gonna happen lol).

You could cut it out layer by layer with a knife from those interlocking foam floor mat things, that'll be tougher.

 

[Jim Davey]

Yeah, this will do for a few years. I'll probably have to change in a few years anyway as the new meters are bound to be a different size.

 

[Wedg1e]

I use a similar principle with a set of three shunts to calibrate Mag Particle Inspection units... but 500A would be the lowest setting on a lot of them, many go up to 6000A. 1500 to 3000A (AC, HWDC or FWDC) is more usual. We often melt the (welding) cable used - and it's usually doubled or tripled to take the current!

We use Peli cases at work; bitter experience tells us they are one of the few designs that can cope with the sort of handling that oil rig helicopter loaders dish out.

If you want a neat finish to foam case linings, try Plastazote. I use it in cases that carry heavy test blocks: unlike pick'n'pluck density foam it retains its shape with chunks of steel in it. You can cut it with a scalpel for a lovely edge and it's available with self-adhesive backing so you can create multi-layered linings for parts that have differing thicknesses (difficult to achieve with most foams).

 

[fenman]

Nice post Thanks

 

[optima21]

How much does a shunt cost to get gauges on a mig that hasn't got them?

I got a 200A 100V display and shunt off ebay for about £10. when I tried it by putting it in series with some steel I was welding (one terminal bolted to the metal and one terminal had the earth clamp on it). it was very difficult to see when welding, and the current would vary massivly due to the amount of stick out of the wire. as a result it never got fitted to the case of my mig. here it is shown with my bench supply that has a similar one 10A 20V display fitted

 

[AndyStobbs]

Do you have to use a shunt or can you use a clamp meter?

 

[Jim Davey]

You can use whatever you want, I just find the results from a shunt more consistent. Clamp jaws spread or of the clamp moves slightly the readout can change by 5 amps or so, I just use it as a backup or when I'm servicing, but I don't rely on it for calibration as such.

 

[AndyStobbs]

Ah fair enough. Looking into one of those foreign voltage/current meters on ebay for my innacurate TIG set, but they need a hold feature or a very slow refresh rate to be any use on a retro fit.

 

[DanielW]

I love the foam lined case for all the multimeters, is there a name for that stuff, it looks like it's made from loads of squares?

Ye old Alumimium
SLR camera cases come with it....this was the start of electronic bods using it to suit their tools and meters

Think the foam was called Neoprene?

 

[DanielW]

The squares are half cut and you just prise them apart a little and you can make quite neat holes for things.Only issue is it's easy to catch an edge as you place things in the foam and tear out squares unintentionally but it's good to gauge layouts of stuff until you get around to having some tougher foam CNC cut (never gonna happen lol).

Eh! you don't need a CNC just to cut holes out in foam a wooden jig like bow saw and 30 amp fuse wire pinned across the bow and both ends connected to a high amp variable power supply.

 

[DanielW]

http://www.ebay.co.uk/sch/i.html?_f...TRS0&_nkw=aluminium+cases&ghostText=&_sacat=0

 

[Jim Davey]

Yeah, if I could be bothered I would do that