[icedvolvo]

Hi,

I have an AC/DC machine which has a fixed AC frequency of 60Hz but also has pulse capability of 300Hz and AC balance control.

While I understand the principle of pulse and AC balance whats the point of having a pulse frequency greater than the AC frequency as per the attached pic??

acdc.jpg (30.35 KiB) Viewed 1164 times

[LtBadd]

What machine do you have?

[Poland308]

Your using the same measure to reference two different things. Pulse generally refers to high and low amps( relationship is in frequency of times between highs and lows). Frequency of the AC is in reference to the actual sine wave of the electricity.

[icedvolvo]

Your using the same measure to reference two different things. Pulse generally refers to high and low amps( relationship is in frequency of times between highs and lows). Frequency of the AC is in reference to the actual sine wave of the electricity.

Nope I'm not ... as you can see from the diagram I drew above .. the frequency of the sine (or square wave in my case) is non adjustable and set to 60Hz on my machine (Unimig KUMJR200AC/DC) and the pulse function runs between the high and low amps ...

Lincoln (or Miller??) machines which allow u to set the AC frequency but limit their pulse frequency to 2/3 of the AC frequency .... in other words the machine will nor let you set the pulse frequency higher than the AC frequency ... but my machine which is set to AC 60Hz will allow pulsing up to 300Hz ...

so what I am asking is:

in AC mode is there any point of pulsing at a frequency higher than your AC frequency?

And as a followup, I always used my previous (much more expensive machine) at a pulse of about 1-2Hz and I would time my weld pools/filler addition to coincide with the pulse to get that nice dropped penny weld ....... so what is the function of having a pulse frequency of 300Hz anyway ???

[Louie1961]

I think Poland 308 is correct. AC frequency is measuring volts (potential) as in how many times the voltage crosses zero potential. Pulse effects current (amps) not potential (voltage). In order to mix the two you would have to graph power, which is volts x amps. Your graph can't be correct. The point of pulse in all cases is to reduce overall power input to a weld (power=heat), while not necessarily effecting voltage

[LtBadd]

From this PDF file

"The addition of a fully adjustable pulse function of frequency, base current and pulse width gives you the added
capability to better control heat input into the work, control penetration & control distortion."

Website for the machine

[Oscar]

The OP's diagram is correct, IMO. When the pulse frequency is greater than the AC frequency, the welder wants to (and will of course) pulse between the high-and-low-currents, but since the square wave output is still on only one side (either EP or EN, don't matter for the sake of this explanation), the pulsing will happen only on that side because the current will not have switched to the other polarity yet (the pulsing is "out racing" the AC frequency). In other words, the pulser-dude is doing his thing switching from high-to-low currents, and the AC Frequency dude is just sitting there because it's not time to switch to the other polarity yet (even though the pulser dude is modulating the current output between the main welding current and the base current). Once the AC dude switches over to the other polarity, then the pulser dude has no choice but to do his pulsing on that side of the "current diagram". He can't slow down and say "this isn't right, I'm still on the same polarity!", because the pulser will do what is set on the panel. He's just gonna pulse away, up and down between the set main welding current and the base current.

[Louie1961]

I get what you are trying to say Oscar, but you can't graph amps on the same scale as volts. Its apples and oranges. you could however graph watts by multiplying amps x volts at any given moment and then plot that. But that is not what his graph is doing. By definition pulse doesn't change the voltage potential, only the amps.

[cj737]

To Oscar’s point: I have a different understanding of pulse behavior with AC.

Take a square wave, if set to 60Hz, and pulse is at 300, then my understanding is that the amperage for the arc on that side of the wave is cycled 5x the waveform on both sides, thereby modulating the heat input regardless of being north or south of Zero. For pulse, you set Peak, Background and Time. You will get 5x the HI amperage input, 5x the Background, for each pulse cycle, per frequency cycle.

Perhaps I am wrong?

I have personally not found an application for pulse rates over 2pps within my limited use, though I do understand it’s benefit in certain applications.

[Oscar]

I get what you are trying to say Oscar, but you can't graph amps on the same scale as volts. Its apples and oranges. you could however graph watts by multiplying amps x volts at any given moment and then plot that. But that is not what his graph is doing. By definition pulse doesn't change the voltage potential, only the amps.

I know what volts and amps are, and I'm not saying you can graph them on the same scale. The OP is depicting current vs time. The implication is that because it is electrical current that is flipping polarities, then that implies that the potential difference must be changing, because it is the sudden flipping of the voltage that causes the change in direction of the current flow (IE: from the EP side of the square waveform, to the EN side of the square waveform). But even then we both digress, as the OP was never arguing anything about voltage.

To Oscar’s point: I have a different understanding of pulse behavior with AC.

Take a square wave, if set to 60Hz, and pulse is at 300, then my understanding is that the amperage for the arc on that side of the wave is cycled 5x the waveform on both sides,

What exactly do you mean by that underlined part? I'm not sure I follow what you're saying.

[Louie1961]

Here's the thing with electrical current. Amps is equivalent to volume, like gallons per minute in a fire hose. Volts is equivalent to pressure, like PSI. The only thing is in electricity it is electrons not water molecules, obviously, but the analogy still holds pretty well. I can have a garden hose with 100 lbs of pressure and it will only flow so much water on to a fire (gallons per minute). I can have a 2 1/2 inch hose line at the same 100 lbs of pressure and the amount of water going onto the fire will be like 10 times that of the garden hose, despite being the same pressure.

Mixing volts and amps on the same chart is like mixing PSI and GPM....doesn't work. Frequency only relates to volts and the number of times the volts crosses zero. Amps relates to the overall volume of electrons moving. Pulse modulates the amps. Overlaying pulse on top of AC cycles really doesn't work. it is confusing because we talk about pulse frequency and AC frequency, but in this case the word frequency is not addressing the same things.

[Oscar]

Here's the thing with electrical current. Amps is equivalent to volume, like gallons per minute in a fire hose. Volts is equivalent to pressure, like PSI. The only thing is in electricity it is electrons not water molecules, obviously, but the analogy still holds pretty well. I can have a garden hose with 100 lbs of pressure and it will only flow so much water on to a fire (gallons per minute). I can have a 2 1/2 inch hose line at the same 100 lbs of pressure and the amount of water going onto the fire will be like 10 times that of the garden hose, despite being the same pressure.

Mixing volts and amps on the same chart is like mixing PSI and GPM....doesn't work. Frequency only relates to volts and the number of times the volts crosses zero. Amps relates to the overall volume of electrons moving. Pulse modulates the amps. Overlaying pulse on top of AC cycles really doesn't work. it is confusing because we talk about pulse frequency and AC frequency, but in this case the word frequency is not addressing the same things.

Well to be precise, you're actually wrong. Amps would not be equivalent to volume, it would be volumetric rate. Amperage is a rate (Coulombs/sec), so you cannot create an analogy with a scalar quantity such as volume. Your analogy of gallons per minute is correct, but that is not volume, it's volumetric rate. The scalary quantity that would be related to volume (in this sort of analogy) would be Coulombs of charge, because it itself is a scalar just like pure volume, not a rate of any sort.

Frequency only relates to volts? With all due respect, I beg to differ. Frequency can relate to any periodic waveform. If the waveform that describes current can be graphed as a periodic function, with a define-able wavelength, then frequency can be used to describe it. I can use frequency to describe anything that repeats itself periodically, with identical individual cycles that repeat themselves.

[cj737]

Take a square wave, if set to 60Hz, and pulse is at 300, then my understanding is that the amperage for the arc on that side of the wave is cycled 5x the waveform on both sides,

What exactly do you mean by that underlined part? I'm not sure I follow what you're saying.[/quote]
My point is the pulse cycles (in this instance 300) is 5x that of the wave cycle (60Hz) so the pulse amperage will cycle between peak and background amperage 5x for each wave cycle, on each side of the waveform (+ to -).

[Oscar]

My point is the pulse cycles (in this instance 300) is 5x that of the wave cycle (60Hz) so the pulse amperage will cycle between peak and background amperage 5x for each wave cycle, on each side of the waveform (+ to -).

I see. I guess this is where we agree to disagree. I agree that the "pulse amperage will cycle between peak and background amperage 5x for each wave cycle", but I disagree that while that specific pulsing is occurring, that it is switching "on each side of the waveform (+ to -)". I agree with the OP that the pulsing is staying within it's respective polarity because it is the AC frequency that determines how quickly (with respect to cycling) the main welding amperage switches polarity, not the AC pulse frequency.

In my mind I am thinking of a generalized situation (that mirrors the OP's view on it):

1. Take a very low AC frequency. Numerically this would be depicted as a low Hz. I mean low. 0.1 Hz. It's still AC, it just happens to have a very low Hz.
2. At this AC frequency, the square wave output would repeat itself every 10 seconds. I hope everyone here knows that Period = 1 / f, therefore Period (or cycle duration if you want to call it) = 1/0.1 = 10 seconds.
3. This implies that the square wave output would spend 5 seconds in one direction (lets just say EP), then the other 5 seconds in EN. Those two halves obviously complete 1 cycle.
4. Say I hire someone to change the welding amperage knob at the machine (aka: he is the pulser). I tell him to change it from 100A down to 50A, and then back to 100A, over and over again, each time instantaneously, and to repeat this every second (assuming he is able to do this accurately for this example, as an actual pulser would on a machine). He has no knowledge of what is happening at the arc side (IE: he doesn't know that the arc is spending 5 seconds at EP, then 5 seconds at EN, and then repeating that cycle due to the 0.1Hz AC frequency). For argument's sake, he doesn't even know what the knob is for, he just knows what to do, and is able to do it.

So lets evaluate what is happening at the arc side. During those first 5 seconds of EP (from our initial assumption in line 3), will the current at any time switch directions? Remember that the AC frequency of 0.1 Hz already determined that the current would not switch direction until 5 seconds has elapsed; so what then of the pulsing? It will go down to 50A, then back up to 100A, every second, but since the current is still operating in EP for 5 whole seconds, at no point will the current switch directions, until after 5 seconds have elapsed. Exactly mirroring what the OP has depicted. The pulse frequency here (1pps, or 1 Hz, whatever you want to call it), is out-pacing the AC frequency that controls when the polarity switches, and thus the cycle length of 10 seconds.

[Poland308]

My only point was you can change the frequency of the pulse rate even on DC voltage. There for the frequency of the pulse has nothing to do with the frequency of the electron flow.

[cj737]

My point is the pulse cycles (in this instance 300) is 5x that of the wave cycle (60Hz) so the pulse amperage will cycle between peak and background amperage 5x for each wave cycle, on each side of the waveform (+ to -).

I see. I guess this is where we agree to disagree. I agree that the "pulse amperage will cycle between peak and background amperage 5x for each wave cycle", but I disagree that while that specific pulsing is occurring, that it is switching "on each side of the waveform (+ to -)".

Actually, I think we agree. I submit (or tried to) that the pulse cycles occur on each side within the AC polarity changes. Meaning, 300 cycles while in EN, and another 300 cycles within EP. Is that your position as well?

And Poland makes my (our) point, that without AC pulse is all about how the amp changes occur, not which side of the wave. If there is an AC wave, then the pulse occurs equally on both sides. At least that’s what I think I understand to be happening

[jwmelvin]

OP, Oscar, cj737, and Poland308 all agree. As do I. I’ve had a similar question about when the two frequencies are very similar such that one would observe a beat frequency between them; I think that could product odd results.

Still some question remains regarding the usefulness of the configuration OP presents. Perhaps it depends on the time constant for puddle melt/solidification such that you could reduce overall heat input but still benefit from the peak current input?

[cj737]

Still some question remains regarding the usefulness of the configuration OP presents. Perhaps it depends on the time constant for puddle melt/solidification such that you could reduce overall heat input but still benefit from the peak current input?

I am sure there are scenarios where a very high pulse rate is not only beneficial, but needed, I have not discovered one personally. It’s a feature I can’t find an application for, just like welding at an AC frequency above 150. I just don’t get that type of work. An probably couldn’t do it anyway

[icedvolvo]

Still some question remains regarding the usefulness of the configuration OP presents. Perhaps it depends on the time constant for puddle melt/solidification such that you could reduce overall heat input but still benefit from the peak current input?

I am sure there are scenarios where a very high pulse rate is not only beneficial, but needed, I have not discovered one personally. It’s a feature I can’t find an application for, just like welding at an AC frequency above 150. I just don’t get that type of work. An probably couldn’t do it anyway

And this I think is the crux of the answer: I have not fond a use for very high pulse rates in AC either I!!

Is the following comments what most people have found or an I missing something:

1: I have played with AC frequency on my old machine with Al and low AC (say 20Hz) gives you a nice warm fuzzy arc and balls the electrode nicely whereas high AC (150Hz) gives a much sharper arc and keeps the tip sharp (I use La).

2: I only use low frequency pulsing (0.5 - 2Hz) to allow me synchronise the "cool/move - hot/weld/feed - cool/move - hot/weld/feed ..." cycle so I get that nice dropped penny look.

Has anyone else found any other use for pulsing or any reason to pulse at say 300 Hz ???

EDIT: I just found a web tutorial which basically said that some welders like to give up to about 20 pulses per weld pause (for want of a better term). So the pulsing is used to "set the weld" without over cooking it. So they set the pulse to about 20Hz and move at about 1Hz so each weld pause spot gets about 20 heat pulses before moving to the next position. If you can feed faster than 1 pe second then up the pulse frequency to match your capability. Jodie suggests he uses about 40Hz pulses probably cos he can actually weld (and the sound is maddening at 20Hz!!)

According to the blurb this allows for better penetration of the weld pool without burning through thinner materials ...and the thinner the material the higher the pulse rate .... but it basically said in not so many words there is no reason for pulse rates like 300 Hz when using AC ....

EDIT of EDIT and I just found another tutorial about DC welding of stainless using high pulse rates to get less distortion of low thermal conductivity metals for critical work .... so the answer is:

a) for some metals welded with DC with low thermal conductivity there is a use for high pulse rates (some up 5000pps!) to reduce heat distortion

b) for AC welding there is no reason (and no benefit) to use high pulse rates which is probably why some of the better welders wont let u set the pulse frequency higher than the AC frequency

[Poland308]

EDIT of EDIT and I just found another tutorial about DC welding of stainless using high pulse rates to get less distortion of low thermal conductivity metals for critical work .... so the answer is:

DC voltage doesn’t change polarity. Polarity frequency is different than pulse frequency.

[icedvolvo]

EDIT of EDIT and I just found another tutorial about DC welding of stainless using high pulse rates to get less distortion of low thermal conductivity metals for critical work .... so the answer is:

DC voltage doesn’t change polarity. Polarity frequency is different than pulse frequency.

yes thats true, but the suggestion is that there is a use for very high pulse rates in DC welding, especially for materials of low thermal conductivity and therefore more susceptible to distortion ... which is why machines have this capability for high pulse rates ... that was the original question ....

[Poland308]

I agree but I only brought up the dc because I wanted to show there is a difference between frequency of polarity and frequency of pulses.