Open circuit voltage vs voltage under load

Having had a little time to myself this morning I did a little experiment to see how the output voltages on my mig are affected when the machine is under load, i.e actually welding.
This was just to satisfy my curiosity but I thought the results might be interesting to have on here too

The machine specs are Portamig 361, 40-360 amp, 17-46V

I first tested the open circuit voltages (OCV) using a voltmeter set at 200V DC with the positive lead connected to the contact tip using a crocodile clip and the negative probe clamped in the earth return clamp.
I then tested the voltages under load using the same method making sure the negative lead of the multimeter was contacting the plate I was welding on.
Pressing the trigger gave me the following readings;
(C,F = coarse and fine settings, figures on the left open circuit, on the right under load)

c1,f1 - 16.7, 11.0
c1,f2 - 17.4, 11.9
c1,f3 - 18.2, 13.6
c1,f4 - 19.1, 14.7
c1,f5 - 20.1, 16.0
c1,f6 - 21.2, 16.8
c1,f7 - 22.3, 17.5
c1,f8 - 23.6, 18.2

c2,f1 - 25.7, 20.8
c2,f2 - 27.3, 20.6*
c2,f3 - 29.3, 23.0
c2,f4 - 31.5, 24.6
c2,f5 - 34.0, 26.0
c2,f6 - 37.1, 28.0
c2,f7 - 40.5, 31.6
c2,f8 - 44.7, 33.0

As you can see, the voltage drops considerably when the machine is under load, anywhere from around 4V to over 11.
It's interesting that coarse 2 fine 2 is actually lower than c2, f1 and it wasn't a fluke either, I rechecked it to make sure

If anyone else decides to do the same experiment, please be careful, obviously you can't read a multimeter with a welding mask on so you have to set up a shield of some sort to protect yourself from the arc

Excellent experiment!!!
I am curious about the same thing too.

I just want to ask you and somebody very familiar to this matter about OCV if the clips should be connected in this way:
- one clip somewhere around the driving rolls,
- another clip on the machine, exactly at ground connection.

I think, in your way, you suffered some voltage drops comparing to the machine with voltmeter inside. I am sure those machines measure voltage at one point around the wheels and at another point near the exit for ground connection.

The voltage drop measurements was the main point of my experiment, I knew they would drop under load but didn't realise by how much.
The voltage isn't applied to the wire at the drive rolls, it's applied to the torch connection at the machine and then onto the wire at the contact tip. I might try it the way you suggest and see what difference it makes .

Good point, zst. I seem to remember the manual saying the voltage readout on the fascia could differ from that at the torch by 'x' amount. I recollect being quite surprised at how much. I think it said up to 2 volts but that's probably worst case scenario with long, thin leads or something.

Do we really care what the voltage is inside the box? It seems to me that the important factor, and the one directly affecting the weld, is what the voltage is at the contact tip, where the work is being done.

As an extreme example, suppose the voltage inside the box is 100 VDC, but at the tip it's 33 volts. Does the 100 volts really make a difference (except in terms of design; e.g., that would mean a very long current path or very high resistance, both of which would indicate a poor design)?

It is depending on lengths of the cables. If the lengths of cable for ground and torch are around 3.5 m each, the voltage would be more, measured on the machine, around 0.5 V-1 V (according to some books) comparing to measurement points on the gun tip and ground clamp.

Only experiment could confirm that.

By the way, the dropa of voltage 4-11V, from OCV to loading voltage are not full data. We should know the applied amperage and then estimate drop of voltage per 100A.

I came accross somewhere on the net (maybe at www.weldreality.com), that good drop from OCV to working voltage should be 3-4 V/100A.
If your machine does not have ampermeter, you could tell us value of wire feed speed. I think there are some conversion chart betweeen A and m/min for each diameter wire (in Lincoln MIG guide, free to download from their site.)

The wire feed speed you may measure with clock. Press the trigger on the gun, and hold it 5 or 10 seconds (time is measure with clock). You could measure the length of the wire and if you divide length with time you may get wire feed speed in m/min for each turn on your machine.

Only insofar as if there is a non-negligable difference between the voltage at the electrode tip and the voltage displayed by the machine then it may be necessary to adjust the volts upwards to compensate. Since I don't know what the drop on mine is I ignore the fact that it's there and adjust by eye. In fact, having just looked at the manual it even says to do this! Interesting experiment, Woz. Though I shan't be getting too hung up on the results. My welding probably doesn't require such precision!

Cripes, zst! Now we're getting complicated!

zst,

Yes, the length of the cables, the diameter of the wire used in the transformer and the wiring of the machine, the combined resistance of the diodes, the resistance of the ground connection, etc. etc. etc. all affect the voltage drop, which is what I meant by "high resistance". The information you refer to all makes perfect sense, but is mostly about how well designed a welder is, along with a factor involving use (ground clamp connection, for example). The lower the voltage drop the better because that means less heating from IR drop, including in the transformer.

I would guess, based on the current range of the welder and the voltages measured, that 100 amps would be around:

c1,f3 - 18.2, 13.6
c1,f4 - 19.1, 14.7

You need to measure under a small load or the OCV volts will be high. Try it with a small load like 2 car tailights in series. On a MIG with noise filtering (better ones do), you can get a 40% difference between the actual and measured OCV voltages.

Is this a 3 phase unit? That will be closer in OCV/load voltages.

Thank you, Shen, resident genius. I was just coming on to ask why I was getting a voltage of 17 on the multimeter when the welder said 5.8! Question answered without even having to ask it!

EDIT: Ok. Question not quite answered, Einstein! Could you please explain, for us lesser mortals, where the bulb goes in order to test the OCV? Athankyou.

You may also need a true RMS meter that can do RMS on DC to get a accurate result.

BTW, an old analog meter will be more accurate as it mechanic ally averages the voltage.

Mine when set to 17V will measure about 24V with no load. As it is a 330A machine, there is a small drop in voltage (there are 8 different transformer taps available to set the droop.) The main difference in OCV and under load is the filter caps will hold the peak voltage.

There should be some drop under load. I see the "droop" mentioned on weldreality as being critical for consistent robotic welds. Note: the drop is usually via the reactor so it is not wasted power.

Just to put the cat amongst the pigeons a little more, spray transfer occurs at voltages above 27 I believe, guess what setting I had to have mine on to achieve spray? Coarse 2, fine 6! OCV 37.1 with no load whatsoever and 28.0 under welding load.
So when we use say, the miller calculator which gives us a voltage, wire diameter and a wire speed range to work with I've always assumed it was the ocv in which case I'd assume mine would drop into spray on a much lower setting.
Would the two bulb trick give the same approximate readings as my under load test or would the resultant voltages be higher?

archie.henderson said:

Thank you, Shen, resident genius. I was just coming on to ask why I was getting a voltage of 17 on the multimeter when the welder said 5.8! Question answered without even having to ask it!

EDIT: Ok. Question not quite answered, Einstein! Could you please explain, for us lesser mortals, where the bulb goes in order to test the OCV? Athankyou.

Click to expand...

Archie, you'd connect the bulbs in series then connect them between the contact tip and the multimeter or the return clamp and the multimeter.
I'm gonna try it tomorrow and see what readings I get

Aaarg, Woz. Don't! So far all this thinking has led me to think my readout is probably telling me nothin useful at all! I'd go back to before and carry on oblivious but, as they say, 'curiosity got the cat'. Well this cat's well and truly hooked!!!

Wozzaaah said:

Archie, you'd connect the bulbs in series then connect them between the contact tip and the multimeter or the return clamp and the multimeter.
I'm gonna try it tomorrow and see what readings I get

Click to expand...

That's what I thought but it made no difference. I think I'll wait for you to report back!

I'd say your readout would be pretty useful, it will be telling you the readings under load which is what you want to know.
When you say it made no difference, to what, your readout or multimeter readings?

Either. Before striking an arc it displays the OCV (or so I assume). Whilst welding it displays the actual voltage. Having just checked, I can say that these two values differ by about 0.4 of a volt max. This seems to make perfect sense. What doesn't us that when measuring the volts under no load and with a multimeter the value on the meter is much higher than that on the machine. This must be where Shenion's point of requiring a small load comes in. Either way, I am fairly satisfied I get it now!!! Too much thinking for a Saturday!

Ah, so if I understand correctly, the built-in meter on your machine must already provide the required load to read correctly? Now I wish I'd had a digital readout fitted on my Porty when I ordered it anyone know if one could be retro-fitted and if so, how?
I've got one on my Butters and always thought of it as being pretty useless as you can't actually read it while you weld but now I'm not so sure.

Yeah I guess so, Woz. It'll also show the amps but that really is useless since it only works whilst you're welding when you can't actually see it anyway! As far as the volts go it's pretty essential on mine side the adjustment is continuously variable. Without it it'd be damned near impossible to reliably set the thing to the same setting twice.