Contrlling the height of electrode.

[kiran]

Hi Folks,

I need a suggestion from the PLC gurus out there. I am not a full time programmer and it is a part of my duty.

I have written a code using RSLogix500 for SLC5/04 for a welding line.

This program is for controlling the height of the electrode as the welding goes on. The electrodes are moved using a small stepper drive. I have the voltage feedback from the electrode to know how far the electrodes are. But this voltage fluctuates constantantly and the range of fluctuation is as much as 0.1 to 0.4volts.

To deal with the fluctuation I have written the program to read the voltage feedback every 5 sec and calculate the average every half a min. and the compare the set level to the calculated average feedback value to determine if the electrodes need to be moved down.

The program works but I have noticed a problem. The electrodes are eanbled to move till the feedback is equal to the set point since the feedback is varying the amount of movement is not the same every time ie every half a min even though the error is same.

This is not causing an problem as far as the working of the machine is concerned but I am sure there is a better way of doing this.

Please suggest if any body has a better idea of doing this.

Thanx,

Kiran

 

[Thomas]

variations...

It seems to me that more data points being averaged would smooth the response. For instance if you check the voltage feedback every second (rather than every 5 seconds) and still do a 30 second average your average should be more consistent. One abnormally high or low reading would have much less effect.

If you did a running 30 second average (fifo the values) and average constantly you could move the electrode in when the average voltage feedback reached a set value. This might work smoother for the equipment.

$0.02

 

[Steve Bailey]

Rather than tell the stepper motor to run at a specific velocity until the instantaneous feedback voltage indicates you're within range, why not use the average feedback voltage to calculate a target position? Then command the stepper drive to move the appropriate number of steps.

 

[Thomas]

actually

Thanks Steve for clarifying. What I think I meant was determine an average voltage feedback value that will initiate a predetermined amount of movement.

 

[Terry Woods]

I think some clarification would be nice.

Kiran said...
"This program is for controlling the height of the electrode as the welding goes on. The electrodes are moved using a small stepper drive. I have the voltage feedback from the electrode to know how far the electrodes are. But this voltage fluctuates constantantly and the range of fluctuation is as much as 0.1 to 0.4volts."

"...from the electrode..."

I take this to mean that you are reading the "Arc Voltage".

The rod-consumption-rate determines the "closeness" of the rod to the work-piece. And the voltage applied to the rod determines the rod-consumption-rate, assuming that the rod is close enough to be consumed. That is a given.

The closer the rod is to the work-piece the lower the voltage-drop across the gap. As the rod is consumed, and if the rod is not repositioned closer, the voltage-drop across the gap increases. The typical voltage-drop is in the near-zero range (maybe upto 2-volts).

The arc is subject to atmospheric conditions... local atmospheric conditions... really local.

If the local area is not contained by an inert gas then the local conditions are come-as-come-may.

So, it seems to me that you are talking about using the Arc-Voltage to control the feed-rate. I haven't seen anything to the contrary.

Considering that rod-consumption is generally constant, but then, it ain't really that constant, I think you should consider "leaning into it". That is, continue feeding the rod until you have reason to stop feeding it.

As you advance the rod, if you read the minimum voltage (Set Point), then stop feeding the rod. As soon as the voltage moves away from Set Point, start feeding the rod again.

If you end up moving closer to the work than normal you will have lower voltage longer (on average - without doing any math).

Now, if it is the case that your feedback signal is related to something other than the Arc-Voltage, it would be nice if you said so and then explained what this feedback voltage really is.

 

[kiran]

controlling the height of electrode

Hi Terry,

You are right it is the the arc voltage that I am sensing.

Thanks for the suggestion.

 

[shooter]

suppose the ideal voltage is 0.25 volt.
if higher them move one step
if lower move back one step
check every 1 seconds.
this will give closest possiblke control
if you need average then make it moving average:
every loop take sample subtract the setpoint from it.
check every 30 seconds or whatever if higher then zero push the rod closer. if not leave it like it is and wayt.
on the input use a small filter to take out the strange effects a little or use oversampling a lot.