[martinr]

For anyone, like me, unsure of exactly what the advanced pulse function on high-end inverters does, the following might help

A picture is worth a thousand words, so I hooked up my PicoScope oscilloscope to my welder (using a x10 voltage probe and a 600 Amp current clamp).


About the Inverter Welder under Test.

My welder is an R-Tech 210P TIG/MMA inverter bought in 2012; R-Tech is a UK firm. On the outside it looks the same as the Everlast 210EXT of that vintage as shown in https://www.youtube.com/watch?v=MaFnUm88rYc Everlast states that it is different inside; nevertheless, the hidden menu is the same, so I was able to set advanced pulse. My welder has all the other functions you’d expect on a top quality machine eg square wave, sine wave and soft square, as well as VRD, arc force (dig) and hot start in MMA. (It is an excellent welding machine and after nearly 11 years, I'm still delighted with it.)


AC Balance and the AC Current Waveform

To make sense of what advanced pulse is, I needed to understand what AC balance and the AC current waveform really look like. My welder’s AC balance shows 0 when the balance is set at 50%; the displays shows a maximum +30 (75% EP cleaning, 25% EN penetration) with the knob fully one way, and a maximum -30 (75% EN penetration, 25% EP cleaning) with the knob fully rotated the other way.

The voltage signals have been removed from all the waveforms, leaving just the current waveform. Voltage adds little to the understanding and just clutters the screen.

AC Balance comparison.jpg (30.7 KiB) Viewed 1273 times

(You might notice that the zero line is not halfway between the current peaks - there is a bias towards the electrode negative current - this is of passing interest only and has no relevance here. The point to note is that AC balance is controlled by altering the duty cycle of the AC waveform.)


Advanced Pulse Waveform

The advanced pulse waveform shows that part of the cycle consists of AC current (with AC balance applied), and the rest of the cycle is pure DCEN current. One takeaway I have from this is that at a pulse rate of 10Hz there is no danger of the electrode hitting an oxide-covered area, but at the lowest pulse rate setting, and a relatively fast travel speed, there could be a problem with DCEN being applied to a portion of the surface not yet hit with AC.

Advanced Pulse Parameters.jpg (21.96 KiB) Viewed 1273 times

The settings for advanced pulse were chosen only to illustrate the topic and are not necessarily practical settings.


I hope these visualisations help make sense of the topic

[BugHunter]

Great post, very interesting.

I think it was on the Miller site I found similar images of scope traces showing the effects of various settings. There are probably similarities of all brands, but likely every one is unique when you look deeper into the details.

[martinr]

Many thanks. It wouldn’t surprise me if you had seen it on the Miller site: only the other day I found an excellent explanation on arc force (dig) in stick welding on their site. In fact, it’s an excellent article on the SMAW process in general.

https://equiptoweld.millerwelds.com/int ... /Stick.pdf

[Oscar]

Very neat, I may have to get one of those picoscopes and probes!

[martinr]

Thanks, Oscar. I’ve got one extra bit to add, just to tie up a possible loose end, later today.

If you’re serious about PicoScope, I believe AESWave are the main US suppliers to your automotive industry and hobbyists. (They are a super outfit: I bought one of their connector kits; it was worth the shipping and import fees to the UK, and their after-sales support is outstanding).

Paul Danner in the US has made 2 excellent videos on PicoScope Basics. Part 1 https://youtu.be/81AGbZcgCZs

Part 2 at https://youtu.be/ntJJYhkmR4I taught me a great deal even though I’d been using a ‘scope for years.

Sorry I’ve gone way off topic, but ScannerDanner is the equivalent of Jody but on the automotive side, and I expect that many users of this forum, being practical engineers, also have a keen interest in taking care of their own vehicles and, if they don’t already know of ScannerDanner, they might possibly be interested in his troubleshooting videos. (Many are free on YouTube and there’s also a premium channel for a small subscription.) Certainly, anyone with a ‘scope, or considering buying one, would not be wasting their time on those 2 videos, and might even save money if they decide it’s not for them.

[martinr]

There's one loose end that should be tied up: what's the significance of "Advanced"? Or, if that's what advanced pulse is, how does it differ from non-advanced, or standard, pulse? One picture says it all:

standard pulse.jpg (31.45 KiB) Viewed 1112 times

So, advanced pulse isn't all that different (on my machine, at least) to standard pulse: the electrode is also negative at all times during the background part of the pulse, it's simply alternating between, in the example above, -20A and -60A. (Note: it’s not alternating between +ve and -ve: it’s alternating between 2 electrode negative current values) Given the duty cycle of that waveform, my guess is that the average negative current during that background part of the cycle is about 55A. [Footnote: the PicoScope software is extremely powerful: it measured the DC average of the (negative) current in the background portion of the cycle at 53.7A ie the alternating background current is equivalent to 53.7A DCEN. So 55A was a fair guess.]

[Oscar]

Do you have your axis reversed somehow? Is the horizontal line the true 0V reference? In that list picture the background AC-wave form is completely above the horizontal line, where in reality it should be alternating positive to negative, just like the main AC welding waveform.

[martinr]

Do you have your axis reversed somehow? Is the horizontal line the true 0V reference? In that list picture the background AC-wave form is completely above the horizontal line, where in reality it should be alternating positive to negative, just like the main AC welding waveform.

Great question, Oscar; I’m pleased that people are interested enough to think about it and to query it.

The dotted horizontal line is zero Amps - I zeroed the Amp clamp at the start. Everything above that line represents the electrode being negative - the normal state for TIG - and everything below represents the electrode being positive (cleaning action to bust through Al oxide). Bear in mind, I had the machine’s AC balance set to max penetration, which translates to 73% of the time electrode negative and 27% of the time electrode positive (forgetting pulse time). So if you look at the 4 large square waves that cross the zero Amps line, you could well believe, looking at the duty cycle, that for 73% of the time the electrode will be negative, and the rest of the time (below the axis) it’s positive. (You also have that bias effect mentioned above such that when the electrode is positive, it hits a peak of about 30A, whereas when it’s negative it hits a much higher peak of about 110A.). If you look at the first screenshot in the very first post above and look at the lower waveform of the pair, you’ll see the underlying AC waveform with 73% of the time electrode negative and 27% cleaning-action time.

So you get no cleaning action during the background pulse time, in advanced pulse or standard pulse because the current flow at all times makes the electrode negative. The only time you’re getting cleaning action is during the time those 4 elements of the square wave are below the axis.

Would it help if I re-ran the test with more representative settings? AC balance of 73% (EN) and 27%(cleaning action) is perhaps a little bit extreme. And I had my pulse settings at 60, 10, 40 ie background current 60%, pulse rate 10Hz, and pulse time 40% of total time. I picked those numbers (40 and 60) so I could easily see which number related to what. I’d be very happy to produce a new set of waveforms with different AC balance and pulse settings eg 65% AC balance and Jody’s typical 33%, 33% pulse settings, if you want? Would that help clarify things do you think?

[Oscar]

It clarifies one thing, since I am used to seeing the axis above the horizontal as positive, but having that part the negative side is trivial to accept and understand. No Big deal there.

The big deal that I do see, is the background pulsing on the standard AC pulse. It cannot be pulsing up-and-down solely on the negative side because you cannot have unequal amplitudes unless your machine is specifically capable of doing so. Correct me if I' wrong, but your machine does not have independent amplitude control? It's not the same thing as balance, which then means the rest of the pictures are also off. In a machine that does not have independent amplitude control, the peaks above and below the 0V reference line should be (aside from those slight over-shoots) relatively identical in amplitude. You cannot have up-down shifting with just using balance because balance only manifests itself in the horizontal axis (time domain). To have different peaks above & below the 0V reference line, a machine must have independent amplitude control, which if I'm not mistaken, that machine does not have. So something is off in the o-scope sofware it seems.

[martinr]

You’re quite correct: the only control I have on the AC balance is that of the duty cycle of the AC wave. There’s no independent control to shift the horizontal axis up or down, as it were. I can’t control the magnitude of the positive and negative peaks, only the overall current setting. That bias, whereby the peak of negative current is a lot higher than the positive peak, is out of my control.

[Oscar]

You’re quite correct: the only control I have on the AC balance is that of the duty cycle of the AC wave. There’s no independent control to shift the horizontal axis up or down, as it were. I can’t control the magnitude of the positive and negative peaks, only the overall current setting. That bias, whereby the peak of negative current is a lot higher than the positive peak, is out of my control.

Then in that case your analysis of the base-current AC (in standard AC pulse) is incorrect on the part that says it's solely on the "negative side". By definition Alternating current must alternate above and below the 0V reference, not just random alternation. It must switch direction, but it's own very definition of alternating current. If it "alternated" just on the negative side, that would just be pulsing, not alternating. Perhaps the offset of the current clamp is something that you can adjust? Because honestly that is all you would need to have a true picture of what is really going on.

[BugHunter]

I can't help but thinking there's an issue with measuring the current but attempting to use a 0 V reference. It's almost like your current is being shown as a Delta, not as absolute. It's too late for me to really put enough thought into this though.

[martinr]

I have another current clamp - a Fluke - which I’ll use instead and will repeat the capture. I originally wrote “oscillating” to describe the square wave during the background part of the cycle, but that didn’t feel right, so I changed it to “alternating”, but that carries with it the suggestion there should be a positive part and a negative part, whereas the current shows only as negative.

I’ll re-run with the Fluke and let’s see what we get.