High Frequency Arc Stabilizer filtering/protection circuit for AC/DC Welder

[Aderondacker]

I have two welders Power craft AC 230 (own for 30+ years) and a recently salvaged Lincoln AC/DC 225/125 and a Dayton High Frequency Arc Stabilizer 3AC01. The Lincoln was in sad shape and I stillneed to repair the lead cables. I'm going to install TWECO 238TSF panel receptacles on the welder and make up adapter cables so I can attach the HF unit easier. I believe I need a filtering circuit to protect the diodes on the Lincoln. I have it working and have been TIG welding steel (20 ga.) without using the HF. I would like to set it up to start the arc on DC. I have search the web and this forum for info on the components needed for a filtering circuit. I did find mention of filtering circuit and bypass caps by no value or type. What I have learned is that capacitors are attached to the output terminals of the welder. I assume that the + side of the cap to terminal and - side to the welder frame (ground). Does it matter what type of capacitor (ceramic vs metal can) and is there any other components such as resistors, etc? I have tried different combination of words to find this info. If someone can help me I would greatly appreciate it. Lots of info on this site and the answer is probably there but I'm just not using the correct terminology to find it.. Thanks John

[tanky321]

I believe you need a low pass filter, it lets frequencies below your set amount to pass. Its pretty much just a capacitor and resistance in series to ground. Then pick off the voltage in between the cap and resistor.

http://en.wikipedia.org/wiki/Low_pass_filter

Use the formula Fc=1/(2piRC)


Andrew

[awright]

See the discussion at http://weldingweb.com/showthread.php?t=27030 in the "General Welding Questions" section. That discussion was mainly about a Bobcat engine-driven welder, but the same principles apply.

While I agree with the recommendation of filtering, I dispute the suggestion that a resistance is a practical filtering component. It is fine and commonly used in signal-level filters but I doubt that the math would work out for a high-current welding application. Resistance has to be low to avoid excessive voltage drop and power dissipation at hundreds of amps and a resistance low enough to be practical would not help much in the filtering. An inductor should be used instead.

Your HF box probably has some filtering at the welder lead inputs. Take a look at that and maybe enhance what is there. My vague recollection is that the capacitors may be just steel plates riveted to the inside wall of the HF box with a mica insulating layer.

Sorry, I don't have specific component values to offer. I'd check out what the manufacturers use.

awright

[Aderondacker]

Andrew, Do I have this correct for the circuit. At the output terminal I attach the capacitor, then I attach the resistor to the capacitor,and then the resister to ground. Or is the resister connected to the terminal first.

John

[Aderondacker]

awright,

I have some circular ferrite cores that would fit over the welding cable that I will be attaching to the panel receptacles. Would these be enough for an inductor and/or do I need to wind wire around these cores, if so do I connect the wires to anything? Then I would just add a capacitor between the terminals and ground. As you can tell, I'm not that good with electronics but I try to learn. Also I did open the HF unit last week and was looking for a filtering circuit. I believe there are two Metalized Polyester Film Capacitors (some type) on metal plates on the bottom of the unit. I assumed that this might be a filter???.

I watched a video two weeks on the net about HF arc stabilizers. They show a spark jumping from the torch to a rubber coated pair of pliers on a grounded welding table, they had the welding circuit off. If I just had HF unit turn on and not connected to the welder. Then took the leads of HF unit that connect to the welder and brought them together to see if an arc would jump the air space. If there was no spark with the leads about 1/8 inch apart, would this be and indication of no back feeding of HF. Or am I out in left field on this.

John

[tanky321]

I agree with awright, I kind of spaced out the resistance portion. Say your using 250A, if you use a 0.001ohm resistor youd be at 62W using the P=I^2R formula. So its pretty unrealistic.

If you use a inductive low pass filter it would be a better idea. As far as inductive values, im not too sure as to how those would be selected. You should be able to get some pretty high current inductors, or wind your own. The inductor would connect between the output of the welder, and where you connect the HF.


Im using a Miller HF unit, and I didnt need any filtering at all. So does this unit take the output of the welder in, then output the HF+welder output? Kind of like a VCR? (bad analogy)

Andrew

[Aderondacker]

Andrew,
I believe Century made it for Dayton. It has the Century style plugs and receptacle 3/8" male plug with a taper. Yes I plug the HF unit into the welder and plug my TIG torch and ground cable to the HF unit. I think I could leave the HF unit off and weld with it connected.

John

[awright]

Aderondacker, your test makes sense in principle, but I don't know how to calibrate a 1/8" gap against voltage. For specific materials and shapes (pointed vs. hemispherical tips of various radii) the gap that allows a spark is pretty closely related to voltage, but I don't know the values. I think it would still take voltages in the KV range to jump a 1/8" gap, but your rectifiers are probably rated at a couple of hundred volts. Thus, I think the test is appropriate procedurally, but possibly not sensitive enough.

A really good test would be to gingerly bring the knuckles of ONE hand close to the two input cable ends with the HF on and see how comfortable that is. If you get no reaction, you're good to go. If you jerk away, we still have no calibration. I suggest the knuckles so that any reflexive reaction can't result in your grabbing the leads uncontrollably. Be sure to report your results.

Can you post any pictures of what you suspect are filters in your HF box? Can you read any values off the caps that you see?


Sorry, I'm full of great generalities but a little short on actual knowledge. If I was doing this I would try to find schematics on commercial units and duplicate them. However, here are some more generalities:

Even a poorly designed filter (that follows the appropriate topology) might do some good and is extremely unlikely to do any harm.

Apropos ferrite ring inductors: They offer fairly low inductance but can't possibly do any harm and they are easy to apply. If you are talking about ring cores that you slip over the end of the cable, that would be fine. Can you see any model numbers on the cores? If so, you can contact the manufacturer to get the inductance they would provide. Ferrites are formulated for specific purposes over a very wide range of formulations so it wouldn't hurt to look for the numbers.

If you are talking about split shells that you clamp around the cable, they MUST be large enough to fully encircle the cable and have the flat mating faces meet. An air gap between the mating surfaces of even a mil will significantly reduce the resulting inductance. However, they can be oversize with no major problem except that the longer magnetic path will result in moderately lower inductance per dollar and make it harder to fit into available space. You can epoxy them together but be sure to squeeze them tightly together to minimize the effective air gap caused by the epoxy film thickness. This results in a weak joint but is necessary to maximize inductance. Use heavy tie-wraps or whatever proprietary system they come with to hold them together even after epoxying. Two will give you twice the inductance of one, etc. The ferrite is non-conductive, but leave the insulation on the cable anyway.

Use a high voltage capacitor from the cable from the welder to ground as close to the point of entry of the cable into the HF box as feasible. I'm not sure, but I seem to recollect that ceramic disc capacitors have the lowest self-inductance of commonly available capacitors. You don't want inductance in the capacitor to ground. I don't really know that a high voltage cap is required here, but it seems appropriate. As a wild guess, I'd try a 0.01uF 1KV cap (about $3) as a minimum. Here's one source:

http://search.digikey.com/scripts/DkSearch/dksus.dll

I might also consider a 0.1uF 1KV cap for about $6.

http://search.digikey.com/scripts/DkSearch/dksus.dll

I originally stated that there should be a capacitor at each end of the inductor, but I think that is probably wrong. A cap on the HF side of the inductor (forming a "pi" filter) would probably short to ground the HF energy that the device is trying to energize the output line with unless you have an inductor on the HF side of the cap. If you were to do that you would end up with a two-section filter, which would provide double the attenuation rate at high frequencies at the cost of more space required. I have no idea if it would be worth the extra bulk and hassle. Any experts out there? We can't really do a very rigorous filter design here since we don't know the frequency of the HF (in addition to not knowing how without some research).

Good luck.



awright

[tanky321]

Your HF box sounds pretty much like mine. Something makes me think that they would have filtering built into it. I mean it seems pretty logical, Just to be clear your system looks like the one below right? If so thats pretty much what ive been using for a few years now, with no problems.

I mean if you think about it, would they sell a HF add on that wouldnt incorporate filtering? Or leave it up to the welder to figure out? High power/HF filtering isnt straight forward, as we have found out, so something makes me think that they would have thought about this issue during design of the HF box.


Andrew

[Aderondacker]

awright,
I'll get those pictures and info on the capacitors next week when I get back home. I was put on lone to do a honey do list. Grand kids are great, if I new they were this must fun I would have had them first.

John

[Aderondacker]

Andrew, Mine is hook up that way, except the gas does not go through the HF unit. It has a separate hose to the gas. I have a P-17 torch and Harris regulator, it came from Grainger back in early 90's.

John

[Aderondacker]

I checked the leads of HF box to welder and there is some feed back. I also talked to a repair tech that works on different manufactures machines. He said one manufacture uses .0047uf Ceramic Cap (high voltage) as filters for stray electric arcing. The caps are connected to panel plugs and grounded to frame of the machine. I have cut the welding leads and installed Tweco 238tsf panel plugs on the Lincoln AC/DC 225/125 welder. Now I need to find the appropriate capacitors Thanks again for everyone's help, John

[NYS3]

I have an older Century High Frequency Arc Stabilizer. You can weld with the welding leads connected through the stabilizer and have the HF OFF. Also, with the HF ON and the welder OFF, I do not have any spark at the torch end, although you can hear the spark gap making noise within the machine. Hope this helps. I also made a hand drawn schematic of the stabilizer circuit with capacitor values if anybody needs it.

[awright]

I suspect that a number of people tracking this discussion would like to see your schematic. I would.

Thanks.

awright

[tanky321]

Out of curiosity what did you measure the feedback with?

[Aderondacker]

I didn't try the knuckle test. I just brought the leads that go to the welder and saw faint sparks when they touched each other.I do get sparks at the torch with just the HF on (no argon on). What would you use to test for value? I have 0-scope Eico 460, Multimeter VOM(, analog and digital). John

[awright]

DO NOT use your 'scope or DVM to test the HF unit output unless you have a "High Voltage" probe! You will probably damage the input stage of your equipment if you do.

It is likely that neither the DVM nor the Eico 'scope would have sufficient bandwidth to respond to the HF output anyway. You might be able to see the peak HF voltage with a specialized probe involving a high voltage voltage divider and a high frequency diode detector circuit but you would have to twiddle the frequency compensation to get an accurate indication of voltage. It's not rocket science (in fact, it's quite a simple circuit), but I'm not sure I could help you design one without hitting the books. Everything considered, I think the knuckle test is simpler and will probably tell you what you need to know.

Who'd a thunk that mother nature would equip primitive man with a high voltage detector?

awright