Re: Strange control problem -- SOLVED!!! -- 3 phase "firing system"

Glen, I solved the mystery. (about the firing system tripping on overcurrent with small currents, on a 200A welder)

Here's what it was all about:

The current sensor (connected to DC shunt) was NOT, contrary to a statement by their engineer, a differential sensor.

What it was actually measuring was difference between the negative bus and the plus side of the shunt. Even though I properly wired two ends of the shunt where they belonged on the firing system, the sensor was not taking the difference between them, it was taking the difference between the + side of the shunt and the negative bus, ignoring the minus side of the shunt.

It is completely contrary to the impression that they were trying to make.

The problem was that there was too much wire between the minus bus and the old shunt in the welder, which was connected to the current meter and was located near the commutator. Between the negative bus and the shunt, there was an interphase transformer and also the "reactor". Too much resistance was all a part of the "virtual shunt" that was in reality being sensed.

My solution: put another shunt right on the negative bus. It is a

300A, 50mV shunt, so the system can produce more current than allowed, I will have to adjust for it, but it happily puts out A LOT of amps.

I pronounce this firing system to be perfectly functional.

I am very happy right now.

I think that I can adjust the firing board (there is a trimpot for current sensing), so that it would produce exactly 200A when the current control potentiometer is at the max. I will do that before going to be tonight.

When everything works, I will buy a huge box of donuts for PCTI.

> >>OK, another guess... I found two pins on the firing system that are >>marked OVSET and OCSET that are not mentioned in the manual that I >>have (someone from that company emailed it to me). >> >>The manual does not quite match the system, there are two more >>internal trimpots not mentioned, and I think that the manual also does >>not mention this OCSET pin. It is blank in the manual. >> >>Some mentions of OCSET pins in other systems like this >> >>
formatting link
>> >>seem to be quite related and explanatory of my problem, which is that >>OC trips at lowest currents. Maybe I need to supply some voltage to >>OCSET. I will call their expert on Monday to discuss this, I think. >> >>i > > Sounds like you are on thr right track, the correct manual should be a > big help :-).
Reply to
Loading thread data ...

I got the right trimpot tuned, so that it will produce exactly 200A when the potentiometer is turned rightmost. I did it at the pot at 25% of max voltage, and moved the trimpot to get the current needle to exactly 50A.


Reply to


Hmm, how did the interphase transformer get in the circuit; it is there to balance current in the 6-phase configuration and is of no use to your 3-phase rectifier, which should not even be using the neutral ???

Also you should be doing your voltage sensing on the rectifier side of the reactor, to provide some hope of control stability while welding. Consider the voltage swings on either side of the reactor during short-circuit mode welding, where a drop of liquid metal contacts the workpiece before separating from the wire. Looking at your figure 3, the inductor needs to be in the load circuit; you should consider it to be part of your load as far as the regulator is concerned. You do *not* want the capacitor connected directly to the work leads after the inductor as shown on your Figure

3, which I must not have looked at very carefully the first time.

MIG welding sometimes is called a "constant voltage" process because the I/V curve for the power supply at normal welding currents is closer to constant voltage than to constant current. The MIG welding process (as opposed to power supply) is however actually approximately constant current, with the current adjusted with the welding wire feed speed control. You might want to find some MIG welder I/V curves, and possibly consider a means of adjusting voltage droop with current. I am not sure if droop is necessary for good arc stability or just an artifact of the construction of the old welders I have seen curves for.

Is that when it works into your test load, or into an arc?

Reply to
Glen Walpert

Thanks. I am going to get a full night of sleep tonight.

It is of no use, but it simply provides some convenient wires to hook up the negative side. It is merely useless, but not harmful, right?

Remember, I wired the rectifier the way you suggested, using only half of the secondaries, wye connected. The common point of the wye is connected to the interphase transformer, simply because "the wires are there".

Maybe I should take it out and try to sell on ebay, but if I do not take it out, there should not be any harm from just having it.

You are 100% right, as I am now realizing. Thank you.

That is indeed the case, voltage sensing is done right off the copper bars bolted on the rectifier output.

Do I need a capacitor at all?

Thanks. I know very little about MIG, but I heard that people weld from car batteries. I will try to read up about it.

My plan is to make a super ridimentary control panel with two pots for voltage and current control, a on off switch, and a RESET pushbutton.

My first goal is to get stick welding to work.

I was thinking of setting voltage to 90 volts and current to desired current, and try to weld.

Should I try to drop voltage as current increases, even when it is below the desired current set by the pot? (that would make the machine not quite a CC power supply, but, perhaps, it is necessary for proper welding?)

Or should voltage be constant until current reaches the p[reset amount, and then should steeply drop to maintain the desired current through the arc?

When I get a sensible looking weld with 6013 electrodes, I think.

A side comment here is that what I am building is not just a welder, it is a universal DC power supply capable of 12 or so kilowatts of power, with range from 0 to 300 volts and 0 to 200 amps. I may revisit the amp issue later, but right now reality is that 200 A welding is the max that my 60 amp garage circuit would allow.

Reply to

Oops, what a brain fart, it is the negative side of the rectifier that is connected to the interphase transformer.

I am very sorry.

Reply to

It can't do any harm just being there, but I don't think it is a good idea to have your DC welding current going through it.

I can't see any reason for one.

Right, try that first.

Your original controller had a provision for adjustable current droop; you can certainly weld without it but you might want to give it a try. This would be a circuit to adjust the current limit as a function of voltage for stick or TIG.

Voltage will automatically drop to to maintain the desired current, when you are in current control mode, that is, operating at the current limit you have set at the controller. You want to adjust the current setting of the controller only and leave the voltage control out of it except possibly to limit open circuit voltage. In most CC welders the open circuit voltage is set by the transformer winding ratio and not otherwise limited.

For MIG welding you switch to voltage control by lowering the voltage set point and running the current limit up "out of the way" so that it has no effect except for short circuit limiting if you stick the wire. Wire feed rate sets the current, not the controller. In a similar manner to CC current droop with decreasing voltage, CV welding can use voltage droop with increasing current by adjusting the voltage set point as a function of current; not necessary but perhaps desirable for better arc stability.

Sounds like fun. Just don't burn the place down, it is amazingly easy to set things on fire while welding, and fires are hard to detect while you have the hood down. I always have both a fire extinguisher and a pressurized garden hose within reach while welding in my garage, and I usually remember to move flammable materials out of bouncing slag ball range and not to wear pants with frayed cuffs :-).

Reply to
Glen Walpert

Glen, do you know why? Would it overheat or something? I can surely take it out, and sell, I trust your judgment, but would like to know better.

Someone also said something interesting, that I likely have too much inductance in the reactor for MIG. Does it make sense to you?

Will do tonight.

Would be indeed very interesting to do something like that. Again, I want to first cover the basics.

That's very simple and nice to know.


Alright, that's good to know.

Yes, thanks. I did a little bit of welding before. I have a little extinguisher.

Reply to

In the original application the magnetic field from each half of the windings in the interphase transformer cancel; the iron only has to handle the field from current mismatch between rectifier sections. With full DC through one half the iron will likely saturate at high currents, plus the extra inductance and resistance, it just seems like a bad idea. I can't imagine anyone wanting to pay even the shipping across town for an interphase transformer however; what use could it be outside of its original intended use with its matching transformer?

That is indeed possible and would show up as an erratic arc and problems sticking the wire as current fails to increase fast enough when the arc shortens or vice-versa. I would try it first and if it dosen't work well try winding a smaller inductor. Possibly on your interphase transformer iron?

Trade it in on a big one, an 18 pound dry chemical extinguisher would be good and guaranteed in stock at reasonable prices at any welding supply company. Good for electrical fires, and will slow down wood fires while someone gets a hose to really put it out. Burning pants can usually be put out with a glove :-).

Reply to
Glen Walpert

Good question. I think that I will just replace it with 1 gauge cable. Maybe not right now, but eventually.

Did you see a picture of my first weld?

formatting link

Well, I could, I suppose, just tap into the middle of my reactor, right?

You know what, yes, I will definitely buy one. Thanks.

I redid some things more nicely, now there is extra 120V power in the welder, switched by the same switch as the main 3 phase contactor. It will power the water cooler outlet on the welder, the fan, 12VDC power supply, the PCTI controller, etc.

120V used to be produced by the welder's controller, but it cannot work this way, now that I converted to a three phase rectifier. I could use some 230v->115V transformer, but it would need to be relatively big (run the fan motor, the tig cooler, etc) . So I simply get 120V from an outlet. It is switched by the same switch as the main power contactor. i
Reply to

Yes, those look at least as good as your practice welds with the original welder. Did you notice any difference in the arc between then and now?

Sure, or you could put in several taps for an adjustable reactor if you can get at the windings.

That works. If you were expecting anyone else to ever service it you would want a "CAUTION - Powered from multiple sources" label on the cover, to let the service tech know to disconnect all of them.

Reply to
Glen Walpert

It is hard to say as I cannot compare them side by side. I also think that using only half windings in wye configuration also limits my current a little too much for 1/8" electrodes.

My impression is that I have a little bit more "sticking" than with the original welder. That, _if true_, which is hard to say with certainty, may be explained by a need for a DIG setting (increase current when coltage drops). It is more pronounced when I try to drag the rod.

Other than that, I think, it welds just as good.

Any idea how I could implement a DIG setting, electronics wise?

(I am sure that you know, but I feel that I should clarify that by DIG, I mean increasing amperage if voltage drops below X volts, to prevent electrode sticking. That amperage increase "burns off" a piece of electrode quickly and resolves sticking)

Got it. Thanks. Yes, I can get to the windings, they are all open and made of copper bar stock.

I will ABSOLUTELY do that. The mitigating factor is that the welder is controlled by a single switch and both cables enter through the same entrance.

Thanks Glen. I would not be where I am, without your guidance.

Reply to

An increase in amperage when the voltage drops is highly desirable in stick welding, but implementing the DIG function as you have described it is probably not the way to go for a first cut. You should first implement droop, reducing current as voltage increases in a uniform manner. I think this function was called BOOST in your manual but I could be thinking of something else; increasing current as voltage drops (BOOST) is tha same as decreasing current as voltage rises (which has been called droop probably since Edison figured out how to parallel DC generators), but marketing prefers to claim their product has BOOST of course.

To implement droop you need replace the current limit setting potentiometer with the 0 - 5 VDC control option mentioned in the controller manual, and produce that signal with either an analog or digital computer. Considering the analog case, you will need to convert the welders DC voltage to perhaps a 0 - 5 VDC signal and use op amps to subtract an anjustable function (potentiometer) of voltage from the current limit set point. This is probably how your original controller worked, so you could just reverse engineer it and copy their circuit:-).

You will want to do the same thing in CV mode, subtracting an adjustable function of current from the voltage set point, so you will need to convert the current shunt signal to a voltage suitable for your analog op-amps or digitize it also.

In a modern welder all this is no doubt all done with DSP. But the easiest stick welding machine I ever used was a resistor bank welder, which merely put a big power resistor with about 25 taps between the welding leads and a big DC rectifier which fed about 50 resistor banks with near constant DC voltage. Lots of droop means the arc intensity varies with arc length strongly - you could easily hear when arc length was optimum (~1/8" for 7018) while running a bead, then taper off gradually at the end by pulling away slowly for good crater fill. If you have no droop, when you pull away the arc gets hotter (same current, higher voltage, more watts into the arc) and conversely it gets colder as it gets shorter, the exact opposite of what you want.

Reply to
Glen Walpert

Glen, the discussion leads me to asking, is it possible that what is actually desirable is to make a "constant power" welder, rather than "constant current" welder?

That makes perfect sense. Sounds also like a simple, modular change.

Sounds interesting. Have you ever tried to do anything with PLCs, would they be suitable for this application? Are they too expensive if I look for a used one?

Reply to

Possible, for instance you could measure power with a multiplier (volts * amps) & compare to a power set point, or just calculate the current set point as power set point / volts, but I am not sure if it would be useful.

You could probably pick up a cheap PLC and use it to replace all of the control and timing relays you might otherwise use, but the PLCs I am familiar with are too slow to be useful for controlling arc current or voltage.

Reply to
Glen Walpert

I asked here yesterday about it, looks like it is not hard to implement a power sensor based on AD633 analog multiplier. I will do that once I get other things done. I made a control panel yesterday (preliminary, on a junk aluminum piece), will try to wire it in tonight. The panel includes a "current feedback/power feedback" selector.

Well, I think that the firing system should control current and voltage to match C/V settings, and the PLC would control relays and such, maybe adjusting requested voltage and current for DIG adjustments. Anyway, I tthinkk that I will just do everything with timing relays and other relays, it is simplier conceptually. A software bug could have very bad consequences.

I have 50 4pdt relays that cost me 60 cents apiece, so I can make all kinds of logical things.

Reply to

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.