[martinr]

I previously posted a topic on this subject, but the images seem to have disappeared, and I previously used a Hantek (Chinese) current clamp, which I lost all faith in. So I've decided to revisit and get fresh data.

My welder is an R-Tech 210. R-Tech is a UK manufacturer/supplier. Although R-Tech deny any links with Everlast, there is a striking similarity between my welder and the Evertlast 210EXT, even down to the hidden menu.

Anyway, because a picture is worth a thousand words, I'll try and let the screenshots of the oscilloscope screenshots do the talking.

ALL THE SCREENSHOTS BELOW WERE TAKEN WITH 70% ELECTRODE NEGATIVE (OR 30% CLEANING ACTION) SET:

To set the scene, the following 3 shots are of AC settings, no pulse.

This is what the sine wave looks like:

70 EN Sine wave.jpg (323.27 KiB) Viewed 6339 times

This is the square wave:

70 EN Square wave.jpg (302.14 KiB) Viewed 6339 times

And this is the trapezoidal wave:

70 EN Trapezoidal wave.jpg (326.64 KiB) Viewed 6339 times

On my machine, the pulse settings start with background current as a percentage of the peak current (1% to 100%), the pulse frequency (0.5 to 10 Hz), and the percentage of the duty cycle during which the AC current is at maximum (20% to 80%). [ I'm sure not all machines are the same in how these parameters are defined, so please don't read across from my data. Until I set up my 'scope today, I thought the duty cycle part of the setting was the opposite way round to how it is. There is nothing in the manual about these settings, which is poor, especially as the display on the machine does not make it obvious.]



Looking now at the Advanced Pulse with sine wave, with 60% background, 4 Hz pulsing and 80% of duty cycle set. Here we see the defining feature of Advanced Pulse: a period of continuous DCEN, punching into the workpiece. In this case, that period is 20% of the duty cycle.

Advanced Pulse Sine WAVE 60 3 80.jpg (435.52 KiB) Viewed 6339 times

For comparison, this is Standard Pulse with sine wave and also 60% background, 4 Hz pulsing and 80% of duty cycle set:

Standard Pulse Sine WAVE 60 3 80.jpg (315.38 KiB) Viewed 6339 times

Finally, just to make it more obvious that there is 70% electrode negative on the AC part of the waveform, here is Advanced pulse with square wave set (as opposed to sine wave above, where the percentage EN is not so obvious):

Advanced Pulse Square WAVE 60 3 80.jpg (390.99 KiB) Viewed 6339 times

I hope these pictures help to make obvious the difference between standard pulse and advanced pulse. The first 3 screenshots - sine, square and trapezoidal waveforms - have certainly helped me to appreciate that, with square wave, there is no 'dead time' i.e. the current is always either maximum positive or negative, whereas with the sine and trapezoidal waveforms, noticeably less energy per second is going into the workpiece because of the ramping up and down of the current. (I can now appreciate why terms such as "softer" are used to describe eg triangular waveforms, or, indeed, anything other than square wave.) So, if welding a thick piece of aluminium, and maxing my machine out, I certainly wouldn't want anything other than square wave; that might be obvious to some, but it certainly helps to see why on a screenshot.

[Jack Ryan]

Thanks for the images.

The Everlast description (also same as R-Tech) describes that but there are no images. Your images complement the Everlast description.

https://www.everlastgenerators.com/blog ... g-part-1-2

Jack

[martinr]

Thanks for the images.

The Everlast description (also same as R-Tech) describes that but there are no images. Your images complement the Everlast description.

https://www.everlastgenerators.com/blog ... g-part-1-2

Jack

Thanks, Jack.

I also found and read Part 2 of that Everlast description:

https://www.everlastgenerators.com/blo ... e%20pulse.

I like to see pictures for something to make sense.

Thanks for your feedback.

[Jack Ryan]

I also found and read Part 2 of that Everlast description:

Yes, sorry - part 2 discusses the advanced pulse function.

I meant to ask what clamp meter you are using now.

Thanks
Jack

[martinr]

I also found and read Part 2 of that Everlast description:

Yes, sorry - part 2 discusses the advanced pulse function.

I meant to ask what clamp meter you are using now.

Thanks
Jack

Hi Jack,

I'm using a 200A/2000A Pico TA167. My 'scope is a PicoScope. I also have a Fluke i30s, which is good up to 20A RMS.

It's the usual story: you tend to get what you pay for, and, when it comes to electrical measurement, an untrustworthy bit of kit is worse than no kit at all, all the more so if you haven't yet realised it's untrustworthy.

[martinr]

I thought it might be useful to look a little closer at the shape of the sine wave, the trapezoidal wave and the square wave, shown in that order below, just to be more aware of how much more energy is going into the workpiece with square wave per unit time:

sine wave up close.jpg (289.95 KiB) Viewed 6319 times

Trapezoidal up close.jpg (294.85 KiB) Viewed 6319 times

square wave up close.jpg (257.7 KiB) Viewed 6319 times

[Jack Ryan]

I'm using a 200A/2000A Pico TA167. My 'scope is a PicoScope. I also have a Fluke i30s, which is good up to 20A RMS.

I have a Fluke i1010 which will handle the current but has neither the bandwidth nor the basic accuracy. It has sufficient to see what is happening though. What I never seem to have is time.

It's the usual story: you tend to get what you pay for, and, when it comes to electrical measurement, an untrustworthy bit of kit is worse than no kit at all, all the more so if you haven't yet realised it's untrustworthy.

Reminds me of some advisers.

Regards
Jack

[martinr]

I'm using a 200A/2000A Pico TA167. My 'scope is a PicoScope. I also have a Fluke i30s, which is good up to 20A RMS.

I have a Fluke i1010 which will handle the current but has neither the bandwidth nor the basic accuracy. It has sufficient to see what is happening though. What I never seem to have is time.

It's the usual story: you tend to get what you pay for, and, when it comes to electrical measurement, an untrustworthy bit of kit is worse than no kit at all, all the more so if you haven't yet realised it's untrustworthy.

Reminds me of some advisers.

Regards
Jack

“ What I never seem to have is time.”

If there’s something you’d like me to look at, I’d be happy to do so: it’s all a good learning opportunity for me.

Thanks, Jack.

[Jack Ryan]

I thought it might be useful to look a little closer at the shape of the sine wave, the trapezoidal wave and the square wave, shown in that order below, just to be more aware of how much more energy is going into the workpiece with square wave per unit time:

I think we'd have to assume that the waveforms don't really look like that - they are just likenesses

The sinusoidal waveform doesn't have a 50% balance so obviously it isn't a sinusoid. There is evidence of quantization levels but also of some overshoot. The square wave shows significant overshoot possibly due to capacitance in the measurement/test circuit.

Nevertheless, the images give a very good insight into the operation of the pulse function and of the power supply itself.

Thanks again Martin.

Jack

[Jack Ryan]

If there’s something you’d like me to look at, I’d be happy to do so: it’s all a good learning opportunity for me.

Thanks Martin, I'll keep that in mind.

Jack

[Arno]

One thing that's always interesting to see would be an overlay of the amps with the weld circuit voltage on the same graph with these kinds of A/C or mixed curves to see if perhaps the volts are shaped like the described functions or not at all.

As inverter machines (try to) control the various components of the circuit separately, up to their rated capacity, it may be interesting to see if these current waveforms correspond to actual voltage waveforms on the machine too or if the voltage is largely just a simple square wave that's mainly determined by the arc length and the inverter control logic is just adjusting/ramping the amps in a more controlled way up to the max kW rating the welder is rated for.

TIG being a 'constant current' process is hardly the case with inverters and advanced settings anymore

Bye, Arno.

[martinr]

One thing that's always interesting to see would be an overlay of the amps with the weld circuit voltage on the same graph with these kinds of A/C or mixed curves to see if perhaps the volts are shaped like the described functions or not at all.

As inverter machines (try to) control the various components of the circuit separately, up to their rated capacity, it may be interesting to see if these current waveforms correspond to actual voltage waveforms on the machine too or if the voltage is largely just a simple square wave that's mainly determined by the arc length and the inverter control logic is just adjusting/ramping the amps in a more controlled way up to the max kW rating the welder is rated for.

TIG being a 'constant current' process is hardly the case with inverters and advanced settings anymore

Bye, Arno.

Hi Arno,

Good idea; I will do that.

I have done it in the past - one channel showed the voltage at the electrode, and the other showed the current through the electrode - but, as I remember it, the results were extremely puzzling/confusing. And, for that reason, this time I only looked at the more important parameter: current.

But let’s give it another go. I think it might generate quite a bit of discussion. Let’s see. Give me a couple of days.

Martin

[Arno]

Yeah.. Because the power source is being actively controlled and generating the actual 'signal' in an inverter welder and we're not looking at a traditional 'dumb' A/C source (like a bare bones transformer just 'translating' the incoming A/C sine to a different/lower voltage but doing nothing else), the normal neat linked behaviour between voltage and current pretty much goes out the window..

Here there's likely a very loose relationship between them with mainly the limits 'known' like at 0V there pretty much has to be 0 Amp close after that, the voltage will have a certain upper limit the device can sustain and the total wattage of amps and voltage at any time can't exceed the device's capacity.

But because the inverter is actively controlling the actual current and the voltage as it sees fit, the result can indeed get a little bizarre

Bye, Arno.

[martinr]

Yeah.. Because the power source is being actively controlled and generating the actual 'signal' in an inverter welder and we're not looking at a traditional 'dumb' A/C source (like a bare bones transformer just 'translating' the incoming A/C sine to a different/lower voltage but doing nothing else), the normal neat linked behaviour between voltage and current pretty much goes out the window..

Here there's likely a very loose relationship between them with mainly the limits 'known' like at 0V there pretty much has to be 0 Amp close after that, the voltage will have a certain upper limit the device can sustain and the total wattage of amps and voltage at any time can't exceed the device's capacity.

But because the inverter is actively controlling the actual current and the voltage as it sees fit, the result can indeed get a little bizarre

Bye, Arno.

Arno

I’m not having much success with the voltage probe. I’m using a x10 (10:1) Fluke probe, but even on the maximum 2kV range, the trace goes above the maximum as soon as the HF comes on. I’ve tried AC coupling rather than DC, and bandwidth limiting to 20kHz, but it makes no difference.

However, when I connect my Fluke 117 DVM, DC range, the voltage jumps from -21V to -57V when the HF start is on, and then falls to -4.3V during welding. I was measuring at the DINSE connectors at the welding machine, and I believe I must be inducing a very high voltage into the x10 probe, giving me a false measurement: it certainly is not >2kV, and not when the DVM is reading 4V!

I have tried moving the probe away from the machine and using some coaxial cable with the shield grounded to earth to try to eliminate any induced voltages, but that made no difference.

So, unless you have any ideas, I think I will give up on trying to measure voltage along with current.

[martinr]

Further to my last post, I also tried using (a) a capacitative automotive HT lead pickup, and (b) and inductive pickup. The HT pickup similarly over-ranged on the maximum setting (just like the x10 probe), whereas the inductive pickup did show an AC voltage at the same frequency as that of the current clamp waveform.

I don’t have a differential voltage probe, and I don’t think I can justify buying one, but I suspect that would be the next step if I wanted to follow through with this.

[Jack Ryan]

I’m not having much success with the voltage probe. I’m using a x10 (10:1) Fluke probe, but even on the maximum 2kV range, the trace goes above the maximum as soon as the HF comes on. I’ve tried AC coupling rather than DC, and bandwidth limiting to 20kHz, but it makes no difference.

I don't know what settings you are using but can you try some (eg square wave) that don't need continuous HF?

HF only at start might help but lift start would be quieter.

Jack

[Arno]

Yeah.. I was kinda expecting the HF to be a headache there as far as voltage measurement goes as that basically indeed induces massive voltage spikes.

Thanks for trying anyway!

Bye, Arno.

[martinr]

Yeah.. I was kinda expecting the HF to be a headache there as far as voltage measurement goes as that basically indeed induces massive voltage spikes.

Thanks for trying anyway!

Bye, Arno.

I’ve not given up yet: I’m looking at differential probes, because this isn’t the first time I’ve wished I had one.

Bye for now, Arno.

Martin

[martinr]

Hi Arno,

I must thank you: I finally decided to get a Pico TA057 differential probe. Now, it all makes sense: there is no puzzlement when overlaying voltage and current.

On all the tests below, I had the AC Balance set at zero, i.e. 50% electrode negative and 50% electrode positive. However, when I measure the peak currents on the 'scope, it is clear that there is quite a discrepancy. A typical result is shown on the trapezoidal wave: 91A positive and 114A negative. This is a minor point, but just in case anyone notices the difference. ( I could, of course, calibrate and make a correction, if I wanted to at a later date.)

Anyway, here are the waveforms. To clean up the signals a bit, I employed low-pass filtering (10kHz on voltage and 80kHz on current)

VOLTAGE IN BLUE, CURRENT IN RED

You might find it surprising that the welding voltage is only around +/-10V
Tell me if there's anything else you'd like to look at. Thanks again for your encouragement and advice.

Martin

sine wave.jpg (402.46 KiB) Viewed 6229 times

Square wave.jpg (408.02 KiB) Viewed 6229 times

Trapezoidal.jpg (358.49 KiB) Viewed 6229 times

And Advanced Pulse

Adv Pulse.jpg (479.38 KiB) Viewed 6229 times

[Jack Ryan]

Hi Martin,

There may be an actual imbalance between the EN and EP currents due to partial rectification.

What is the nature of the common mode noise?

Thanks for your efforts - very interesting!

Jack

[Arno]

Very nice to see the V/A graphs

Eg. you can clearly see the spike in voltage right after the crossover point where the arc need to re-establish (higher resistance) even though HF is helping to keep the ionisation. Once the arc path is there the voltage drops to the base welding level as the path now has low resistance and the inverter logic then kicks in to start controlling the actual weld amps.

I expect you have the torch in a clamp and at a fixed distance to a block of alu or similar? Probably nice and short at the moment which should keep the voltage low.

If you do a trace with a longer or shorter arc you'd likely see the welding voltage get higher or lower with the arc lenght and at a certain (too) long distance the inverter may start drooping on the amps as it hits it's max kW rating on the total power it can provide.

And yeah.. As you can stick-weld with just a car battery and an electrode (not advised, but possible..), 10V doesn't seem too odd

Bye, Arno.

[martinr]

Hi Martin,

There may be an actual imbalance between the EN and EP currents due to partial rectification.

What is the nature of the common mode noise?

Thanks for your efforts - very interesting!

Jack

Hi Jack,

I don’t know what the nature of the common-mode noise is? How would I go about finding out for you? Unless I measure, using the differential probe, between the welding return lead and mains earth. Will that reliably show the common-mode noise?

Martin

[martinr]

Very nice to see the V/A graphs

Eg. you can clearly see the spike in voltage right after the crossover point where the arc need to re-establish (higher resistance) even though HF is helping to keep the ionisation. Once the arc path is there the voltage drops to the base welding level as the path now has low resistance and the inverter logic then kicks in to start controlling the actual weld amps.

I expect you have the torch in a clamp and at a fixed distance to a block of alu or similar? Probably nice and short at the moment which should keep the voltage low.

If you do a trace with a longer or shorter arc you'd likely see the welding voltage get higher or lower with the arc lenght and at a certain (too) long distance the inverter may start drooping on the amps as it hits it's max kW rating on the total power it can provide.

And yeah.. As you can stick-weld with just a car battery and an electrode (not advised, but possible..), 10V doesn't seem too odd

Bye, Arno.

Hi Arno

And that voltage spike as shown is a good bit lower than it really is due to the low-pass filtering I set, both to clean up the signal and to reduce the height of the spike relative to the rest of the waveform.

I held the torch by hand and moved it along about one inch, keeping as short an arc as I normally would do. (I set the ‘scope to single trigger.)

That’s a great idea for a DC setup: voltage variation vs current with arc length. I must give that some thought and then give it a try. And with the maths channel, a plot of kW power would be simple to display. Or, I’m sure I can display X-Y, so I could display Volts on the Y axis and Amps on the X axis (and kW as well, perhaps). Leave it with me.

Ha ha! Going back many, many years I have indeed welded from a car battery. The electrode holder had a weak spring in it which allowed a small contractor to make and break several times a second, as I remember it. It worked - sort of, provided you didn’t get a spark exploding the end cell and blowing the battery case apart.

[martinr]

Very nice to see the V/A graphs

Eg. you can clearly see the spike in voltage right after the crossover point where the arc need to re-establish (higher resistance) even though HF is helping to keep the ionisation. Once the arc path is there the voltage drops to the base welding level as the path now has low resistance and the inverter logic then kicks in to start controlling the actual weld amps.

I expect you have the torch in a clamp and at a fixed distance to a block of alu or similar? Probably nice and short at the moment which should keep the voltage low.

If you do a trace with a longer or shorter arc you'd likely see the welding voltage get higher or lower with the arc lenght and at a certain (too) long distance the inverter may start drooping on the amps as it hits it's max kW rating on the total power it can provide.

And yeah.. As you can stick-weld with just a car battery and an electrode (not advised, but possible..), 10V doesn't seem too odd

Bye, Arno.

Hi Arno

I had a look at how arc voltage might vary with arc length, with a view to gaining insight into the Voltage-Current welding characteristic of the inverter. I set up TIG DCEN. After trying to plot X-Y, Volts vs Amps on the 'scope, I gave up and simply looked at both traces as I increased the arc length from less than the electrode diameter to a massive 2 inch arc length. The results show that the voltage (and current, as you'd expect) stayed remarkably constant as the screenshot shows.

VOLTAGE IS DARK BLUE, CURRENT IS DARK RED

TIG DC Volts and Amps.jpg (107.73 KiB) Viewed 6146 times

I think I might try a MMA setup instead of TIG: voltage might vary more with arc length, but with TIG there is practically no variation of voltage with arc length.

I think

[Arno]

Nice!

I'd have to dig out my old physics books to find out what the effective resistance would be of an ionised path through argon. Could well be low enough that the actual effect on any 'normal' welding distance variation is in effect pretty much 0.

Quite interesting though as many docs and videos mention the arc effectively getting 'hotter' (and of course more diffuse) when long-arcing, while it seems the effect is more about the diffusing and putting the heat in a much bigger cross-section on the work piece than when keeping a short arc making it seem like it's hotter. It seems it's just getting less focused.

MMA/stick should be pretty similar as it's (supposed to be ) constant-current as well, but a good comparison to plain TIG.

MIG and such should be a lot more dynamic/spiky voltage and current wise, but that's of course also part of the process esp. when using short-circuit. May well be that the spray-tranfer types here make for a much more smooth/consistent power delivery similar to TIG once it gets going.

Bye, Arno.