I have been trying to find someway of getting an Arc Xfer/Arc good signal on my Miller Spectrum 625 machine for about a year now. The Miller 625 is the same machine as the Hobart Airforce 625. Does anybody have some exprience with either of these machines and working them with a cnc machine?

In a few days I am going to put a current transformer on my ground lead because putting it on one of the power wires did not work. Does anybody else have an idea how to get an "arc good" signal from either of these machines?

Thanks, Steve

Thermal Dynamics use a small current sensor by MicroSwitch part # CSDA1DC
Al.

The problem with using conventional Current transformers with the smaller inverter type plasma units is that the smaller units don't have PFC (Power Factor Correction) so the curent spikes are of short duration (pulses) that don't work well with conventional 60 HZ magnetics.

The Honeywell unit is a hall effect unit and just reacts to DC (and AC) magnetic fields of most frequencies. If used on the Workclamp ground Where you have DC it should work. You need an external source of DC voltage and the output is active low and only sinks 20ma (TTL levels). To make it meaningful to external electronics it needs to have more electronics on it. At the least a buffer driver to give it 50 to 100ma drive capability.

The price at 18 bucks is not a killer but the need for added electronics might be. It won't drive the Arc Good circuit directly and is the wrong polarity (is ON = high with no current and LOW with current). It also needs an independant voltage supply which adds to the complexity if it's designed to mount at the plasma unit.

There is an interesting article on building a PIC based high current readout for an auto using a hall effect sensor that could be adapted to sensing on the workclamp lead of even bigger units.

You did not say what you are inputing the signal to, if you put something like the MicroSwitch unit on, if you put it on the DC side, it won't matter if it is on the high side or ground side, it will operate up to 16vdc supply and will input directly to some low current opto inputs that require a sink input for on. Most are configured this way. It would only require a FET buffer (2n7000) to raise the current drive to 200ma or use it for TTL input or relay driver.
Some systems use this for arc transfer or to constantly monitor the Arc Good condition.
Al.

You did not say what you are inputing the signal to, if you put something like the MicroSwitch unit on, if you put it on the DC side, it won't matter if it is on the high side or ground side, it will operate up to 16vdc supply and will input directly to some low current opto inputs that require a sink input for on. Most are configured this way. It would only require a FET buffer (2n7000) to raise the current drive to 200ma or use it for TTL input or relay driver.
Some systems use this for arc transfer or to constantly monitor the Arc Good condition.
Al.

I have already attempted to use a reed switch (Hamlin P/N 59145-020 0549) which was unsuccessful. The switch would work for a few seconds and then fail closed. I started out actually using the reed switch that is used inside of the larger Miller machines. This was on the recommendation of a Miller tech rep. I had actually bought the $60 switch direct from miller and found out that its the same switch p/n that I previously mentioned. The switch from Milller comes with a cage around it for the ground lead to go through but its not needed so I just bought the switch itself from digikey. The power supply that runs through the switch is a little transformer from some household item like a small widget light or something.

BTW I am using an MP1000 torch height controller. The current draw through the reed switch from the torch height controller card is very very low so it shouldnt be burning up the switch. The dataplate on the power supply is 9vdc @ 500mA.

I have an online buddy that a plasma cam so he has the instructional video for connecting just about any plasma machine. From the clip that he sent me (relating to the Miller Spectrum 625 and Hobart Airforce 625) it looks like plasma gets its own arc good/xfer from sensing the torch voltage. Is this true?

Does anybody have some pictures of how they connected their miller machine? Thanks, Steve

I would suspect the reed switch is failing somehow due to high current causing it, the reed switch is a mechanical electro-magnetic device., the Honeywell unit is electronic and current is only limited by the size of the conductor, in fact I believe Thermal Dynamics split the conductor into 2 conductors, one goes through the sensor and the other bypasses, only because the single conductor is too large.
All you need to register is that current is flowing, voltage in itself is not an indication of transfer.
Al.

Hello,

We used to make auto THC for the machines which we sold (I'm retired now). The approach we took was to monitor the arc voltage; it is typically high, 200 or more volts before the arc is struck. When the arc transfers, the voltage drops to 90-120 volts as current flows form torch to workpiece. When this happens, a relay closure or other signal can be sent to begin the machine motion and enable the auto height operation. I designed these controls around 1990.

Rergards,
Jack C.

Hello,

The approach we took was to monitor the arc voltage; it is typically high, 200 or more volts before the arc is struck. When the arc transfers, the voltage drops to 90-120 volts as current flows form torch to workpiece. When this happens, a relay closure or other signal can be sent to begin the machine motion and enable the auto height operation. I designed these controls around 1990.


I agree the AHC can sense the Voltage, but this is for arc transfer/arc good signal, here I believe the current flow is sensed for this.
Al.

Using open circuit voltage for arc good is not as reliable as actually sensing the current. On some machines that voltage can be higher or lower so it has to be adjustable. We elected to use the method all of the plasma systems use and that is to measure current.

On the reed swtich: I can't imagine how it could be burning out unless you have the external voltage source across the contacts. The coil around the outside is wrapped around the reed body (heavy enameled wire) but makes no physical contact with the wires coming from the reed switch itself. That switch only has to supply a few mils of current. Our circuit on the sensor card draws less than 50ma total (with the relay pulled in). It basically is a rectifier (bridge) and series pass (small NPN) with a zener and pot to limit the voltage and to set the sensitivity. That drives the coil of a sensitive 5V relay. Any AC or DC voltage of 7V or greater at about 50ma will turn it on.

To get a voltage signal out of a reed circuit the reed switch needs to be in series with either the + or - (or one leg of an AC signal). One side of the external supply hooks to one side of the NO reed swtich. The other side of the power hooks to one of the Arc Good terminals and the other Arc Good terminal connects to the second (other) reed switch terminal. Basically the reed switch is a SPST switch in series with one side of the power source. Voltage should be at the ARC GOOD terminals on the card ONLY when the cuurent is flowing in the reed switch coil and activates the contacts. They won't carry a lot of current but they will handle up to about an amp.

If they are burning up they are not connected right. I used a swtich like that on my Miller for Arc Good for over 8 months until we started shipping the CT.

After my last post I finally got around to trying a different use of the current transformer. I wrapped the work lead around the CT once and put a 15ohm resistor across it. There was NO output during either operation, cutting or firing into air.

When this all started I did as the MP1000 instructions dictate and put one wrap of one input leg (220v) through the current transformer. This would give me insignificant readings. Cutting into thin air would give me about 3v and cutting metal would give about 3.2v. Those numbers are inconsistent with electrical theory. I am thinking that my sub-par equipment (Sears multimeter and Harbor Freight am-meter) could be causeing me bogus readings.

I was really hoping to get confirmation that somebody has a used the same Plasma cutter as I and actually got it to work. I really dont want to buy another machine. Thanks for any additional help, Steve

What style of CT are you using, is it a module type with internal electronics like the MicroSwitch or a passive CT with just a secondary and the plasma cable as primary?
A passive type has to be used on AC to get any meaningfull readings and also it is a Current transformer, and should not be used open circuit or lightly loaded, due to excessive high voltages appearing on the secondary, if you have a 15ohm res. as a shunt the secondary current can be calculated by knowing the ratio of the CT.
Al.

What style of CT are you using, is it a module type with internal electronics like the MicroSwitch or a passive CT with just a secondary and the plasma cable as primary?
A passive type has to be used on AC to get any meaningfull readings and also it is a Current transformer, and should not be used open circuit or lightly loaded, due to excessive high voltages appearing on the secondary, if you have a 15ohm res. as a shunt the secondary current can be calculated by knowing the ratio of the CT.
Al.


It is a secondary transformer of the 200:1 value. At one point I did the calculations to figure what kind of voltage I should expect with the given variables. Machine is rated at 40amps so I used that number along with the 200:1 reduction and a 15ohm shunt resistor. The only problem is, I can do all the math I want, all day long but I still need to see the equipment put out the numbers. No matter how many winds I did through the CT I would get relatively the same voltage all of the time.

I used a clamp on harbor freight am-meter to determine actual amp draw during cut and firing into thin air. Cutting at 40amps (as indicated on the dial of the machine) did not indicate 40+ amps at the clamp on am-meter.

Torchead has several times in this forum and other forums that machines like mine that us power correction can have an affect on actual amp draw.

I did have the reed switch connected as suggested earlier. It is a small ac-dc wall socket converter for somekind of household electronics (i dont remember what it was for). It had on leg going to the switch that then connected to the THC board. The other leg went direct to the THC board. The switch (N/O) would leave an open circuit until flowed through the work clamp. I had it setup just as Miller designed it.

The only things that i can think of that are frying the switches are spike from the crappy wall converter or high frequency from the plasma machine could be inducing an extra current into the switch. You can correct me if Im wrong, the high frequency may not be a problem.

I bought a few extra switches so I may try and experiment with a different power supply. Does anybody think that the powersupply is messing up the switch? Any suggestions on what kind of power supply to use? Some exact examples would be nice. At this point, if the power supply costs $100 i am still better off than getting a new plasma cutter. Thanks again, Steve

I am still not clear on where you are measuring the current with the clamp-on or the CT, the non-powered type of clamp on is AC only, as is the passive CT.
Whether you are measuring the AC or DC current, the correct device has to be used.
Measuring the DC current, the PF does not come into the equation.
Al.

I have attempted measureing from the ac and dc sides using the same current transformer. The transformer I used is just a secondary tranformer and know it was incorrect to use it on the dc side (work clamp lead).

Per Torchhead and CandCNC the correct way is to measure current flow from one leg coming from the wall socket. When I did it this way I could only get about .5v split cutting air or cutting metal. The additional current draw when cutting metal was not showing up on the current transformer output. I have tried taking the current measurement before and after the power correction circuitry.

No matter how many times I wrapped the leg through the CT I would always only get a negligable difference in the output voltage from the CT.

Here is a photo of the current transformer as mounted inside of my Miller Spectrum 625.

http://thumb16.webshots.net/t/53/753/5/45/6/2596545060059540656XIWlOX_th.jpg (http://home-and-garden.webshots.com/photo/2596545060059540656XIWlOX)

So I just got back from Sears with a new Fluke meter. I get .2ohms when checking the resistance between the output posts of the current transformer with the shunt connected and the wires attached to the THC sensor card. I suppose the .2ohm reading could be because of the transformer winding.

I did another test run with the new multimeter and I get 4.7 volts cutting air and 6.8 volts when cutting metal at full output from the plasma machine (40amps).

Is 2 volts enough of a difference to work with the THC sensor? Steve

200:1? Not the CT provided with the THC! It's a 200:5 (40:1).

Soo. At 40A of current with one full wrap you should see 1A across 15 ohms or
[drum roll here]...15VAC. It only takes about 7V (min) to turn on the Arc Good circuit so that should be plenty.

Describe the reed switch. Do you have one side of the plug-in supply to one of the Arc Good pins on the sensor card? Then the other side of the supply goes to one of the two reed swtich connections; then the other reed switch (opposite end ) is connected to the other Arc Good terminal?

I can think of no reason your reed swtich is burning out unless it's hooked up wrong. See PDF for proper hookup and the Arc Good circuit on the sensor card.

So I just got back from Sears with a new Fluke meter. I get .2ohms when checking the resistance between the output posts of the current transformer with the shunt connected and the wires attached to the THC sensor card. I suppose the .2ohm reading could be because of the transformer winding.

I did another test run with the new multimeter and I get 4.7 volts cutting air and 6.8 volts when cutting metal at full output from the plasma machine (40amps).

Is 2 volts enough of a difference to work with the THC sensor? Steve

Takes about 7V AC with the sensitivity set at max to make the relay pull in.
You can sense a 2 volt differential but you would need to get the voltage up a little more by using a higher value resistor or another wrap on the CT.

You can set the trip point at any spot with the sensitivity adjustment on the THC Sensor card.

I ran another wrap through the CT to get the voltage up. I havent had a chance to check out the voltage yet though.

Correct me if I'm wrong but is the 6.8volts that I'm seeing correct? I am only taking a reading off of one leg of a 220v supply so I should be seeing 1/2 of the 15volts you mentioned. Is my logic right?

Also, there is another issue because I am only seeing 6.8volts with two wraps. By the math I should see 14-15volts.

Hopefully this one more wrap will raise the voltage enough to make it all work.

If you read 40Amps the math does not change. Only thing that skews the math is that the current waveform is not a pure 60 HZ (PF = 1.0) so you may not get a linear relationship. It gets worse with smaller plasma units because they do not typically employ power factor correction. Since it's current we are measuring if it's not on but a short amount of the cycle (low PF) then magnetics designed to read or pass a normal 60HZ waveform will not give the same results on a low power factor.

Believe your meter. It is that voltage that gets used.

The problem in electronics (especially AC and ESPECIALLY with magnetics is that there are a lot of variables so in the end you have to measure things and when the measurement does not match the theory you don't adjust the measurement but you have to account for the descrepency by thinking about other variables.

I still can't account for your burned up reed swtiches when you are pulling virtually no current!

If you read 40Amps the math does not change. Only thing that skews the math is that the current waveform is not a pure 60 HZ (PF = 1.0) so you may not get a linear relationship. It gets worse with smaller plasma units because they do not typically employ power factor correction. Since it's current we are measuring if it's not on but a short amount of the cycle (low PF) then magnetics designed to read or pass a normal 60HZ waveform will not give the same results on a low power factor.

Believe your meter. It is that voltage that gets used.

The problem in electronics (especially AC and ESPECIALLY with magnetics is that there are a lot of variables so in the end you have to measure things and when the measurement does not match the theory you don't adjust the measurement but you have to account for the descrepency by thinking about other variables.

I still can't account for your burned up reed swtiches when you are pulling virtually no current!

That is some great information.

I didnt read 40 amps anywhere. The knob goes up to 40 amps so I just figured the "40 amps" would be split between the two hot legs coming into the machine.

It sounds again like using the reed switch may be the best answer. I wouldnt put it past me to have had it hooked up incorrectly and that could be why the switches burned out. I need to look around and find the rest of the switches that I purchased.

Thanks, Steve