Xmt remote contactor problems

[Chadv]

Hey guys
Im a new user and this is my first. I have a xmt 304 serial number kg177272 I bought used about a year ago. I’ve been using it as a stick welder since I got it and it works fine but now I’m ready to try some of its other functions and I am having trouble.
I tried to hook up the s60 wire feeder that it came with to and I won’t work. The wire feeder will feed wire and flow the shielding gas but no weld current. I tried to jump the A and B pin to fire the contactor and still no weld current. It appears the all the electrode hot welding functions work but none of the functions that use the remote contactor work
Can someone please give me some advice on how to start to narrow down my problem
Thanks
Chad

[jjohn76]

Hello Chad,

Have you tried jumping between I and J? That's the other contact closure circuit. That should help narrow it down to something wrong with the socket and/or cable. Also, is there any corrosion showing in the remote socket?

If you're looking to remove the cover and check inside, it's worth both getting a technical manual and disconnecting the power cable at least 5-10 minutes before pulling the cover (that's more than long enough). You can find a tech manual that should do for what you need on the Red-d-Arc website. Check under either resources or support, look for a manuals link, and search for the EX300 technical/service manual. The manual shows you where to check to make sure your bus capacitors aren't charged (lethal voltage, best to be careful and/or find someone qualified to work with lethal voltages).


After that... Your problem sounds like it's between the remote socket and PC3 (front vertical board right behind the panel) or it's a bad component on PC3. If facing the front of the machine and looking behind the front panel on the right side, you should see the cable between the remote socket and PC3. It takes some dexterity, but check the board connection for that cable - remove/inspect, blow it out and check continuity if you can (follow system wiring schematic, ppage58), or wiggle it and make sure it's fully seated if you can't remove it. Try it out after that.

If that doesn't resolve it, whoever fixes it will have to get to PC3, which is pain to access without removing the top main board. There are a couple components to check, and if I were fixing it, I would disconnect the board, check a few zener diodes, and likely power the board up with a current limited external power supply.

Let's hope the first couple checks work.

[Chadv]

Thanks
connecting the I and J triggered the contactor but I not sure how to proceed. Should I try to fix the xmt or rework the plugs on my wire feeder and tig controls. I would probably have to use a relay or something to trigger the 110 between I and J

[Chadv]

I checked for continuity between the A an B pin and the pc3 board and everything looks ok

[jjohn76]

Chad, it's up to you on whether to fix PC3 or put a relay outside the machine. I wouldn't booger with the machine or leads, and instead spend the $60 for plugs/sockets, 24VAC relay, and wiring. A 24VAC relay with solenoid leads connected to terminals G/K and B from the feeder, and the set of N.O. contacts connected to I and J would work, but I would strongly recommend ensuring pin B from the welder's remote is not connected to the relay - it's not needed and it could cause issues.

Before that, what resistance do you measure across pins B and K? If it's open circuit, something is wrong with or around R74 on PC3. Confirm connection from pin B to R74 and from R74 to the anode side of D13. I wouldn't expect it to be shorted given the I-J contactor control works. Let me know if it does.

Do you have corrosion on PC3?

[jjohn76]

Sorry if the relay portion was confusing. If going that route , it's worth just build a cable section with the relay that mates between the remote socket and the wire feeder plug. They're easier to sell unboogered in case you decide to sell them down the road.

[Chadv]

I think I will go the relay route, board level repair is a bit over my head. I’ve been playing around with arduinos and have been thinking of building a custom control box to control a solenoid ,spoolgun , and maybe some 4t and pulsing if I get real ambitious. Thanks for your help

[jjohn76]

Edit: I changed the inserts to be the correct ones for Miller remotes - 20-27 instead of 20-14...

Makes sense. Have you found the right connectors to put in a splice? You can usually read the connector code. You'll usually have something like MS 3106-20-27P on the connector.

The first four digits (310x) are for the connector configuration - you can get connectors that screw into a panel vs attach to a cable (difference between the panel mounted remote socket on your XMT or cord mounted remote plug) and have male threaded (like the panel socket) or female threaded collar (like your remote plugs) for locking to the mating connector. Check out 3101 vs 3102 vs 3105 (IIRC) vs 3106. There are right angle ones too, though I never use them. If you decide to put your remote in a project box near/on the XMT for future projects, the right angle may be the way to go.

The second two digits are the case size - size 20 in the case of Miller connectors.

The last three digits are for the pin configuration. I believe 27 is the right configuration for Miller, but that's off the top of my head. The P or S indicate if you have the pins or sockets for mating the individual conductors. Whatever you plug into the XMT will have a 27P at the end of the model number, and whatever you connect to your remote plugs will have a 27S.

Regarding board level repair - It's probably corrosion on one of the traces - you have a microcontroller (atmega16 instead of the atmega328 on your Arduino IIRC), that you'll want to avoid damaging because of the firmware on it, but everything else is replaceable. The issue should be obvious - corrosion on traces. If that isn't the issue, you should see a burned resistor (R74) You can also measure diode drops if that B to K resistance at the remote measured short.

The hardest part is removing the PC3 attaching screws without pulling the main board PC1. I don't ever remember pulling PC3 without first pulling PC1, but do remember a screw near the remote plug that would be hard to access.

Removing either board isn't too big of a pain because the top board is plug and play - almost every connector/plug is unique (if not every) and those that aren't can only reach their intended connections. But it's good to have the manual, give the capacitors time to discharge, and check the big capacitor voltages before touching boards.

You can put down an esd mat down for board troubleshooting and attach your esd bracelet to the XMT chassis/ground while removing boards to avoid electrostatic discharge/damage to be safe - I didn't while living in Washington state, but may do it for at least my first few projects while living in a very dry northern New Mexico. Up to you...

As for custom control box down the road, totally doable for all three, though the middle one is most ambitious (sensor less motor control). You have 24VAC available across pins A and G/K, so you can create +/- DC rails with a couple diodes, a few capacitors and voltage regulators. Linear regulators are likely good enough, provided you can find ones that can handle 24VAC rectified. Your DC common rail will not be board common unless doing half rectification (use only a single diode and connect the controller box common directly to pins G/K). If you want to do a +/- rail, your controller common is still tied directly to G/K, but you would use two diodes. Also, be sure to filter/clamp/protect all power and signal lines. Many options with PTC varistors, zener diode, crowbar circuits and fuses.

The gas solenoid is probably the easiest - and I would do it (at least initially) by having the solenoid timing by the pedal. You control preflow based on when you press the pedal (check pressed condition in your main loop or set up external interrupt if you're wanting to learn that, and use either a loop counter/delay/or timer interrupt for your post flow). Millers use a 24VAC solenoid in many machines (call tech to confirm part numbers for a 24v feeder) that you can find NOS on ebay, or you can find other solenoids on eBay as well. Be sure to put a varistor across the solenoid and a diode across the relay solenoid.

The 2T/4T and ramp/crater are a little more challenging because your controller needs to know the state of the arc - open or welding. The good news is the XMT has analog signals available that tell you output current and voltage. Most machines use output voltage to sense the state of the arc - Above ~50V and it's open, below that and the current is started (also how they time HF start). Since you're probably using lift arc, you may need to use both current and voltage feedback signals. Arduino is easy enough to do in main loop, especially since all of your timing is in seconds vs some much smaller time window.

You can find diy pulsing on YouTube using an Arduino and an analog multiplexer - just be sure you are aware of and using the right voltages for the components. Miller uses 0-10VDC for output current references. Depending on which microcontroller you use, you need to divide voltage and clamp (zener diodes will work) voltage references in. I would recommend creating your own 10VDC bus to use.for current referencing instead of using the 0-10V provided across pins C and D. This requires you to set Max and background off the machine, but saves you from damaging the board power supplies. It definitely means you need to clamp any voltage you feed to the XMT. This also may require isolation from the 24VAC circuit (I need to check in to that, my Thermal Arc 250GTSW has the current reference and board common as the same and I am thinking about a similar pulsing project).

Thinking more about the spoolgun controller, you can probably find ones that would work for a 35V, 2A(minimum) for under $30 now. The Spoolmatic 30As have a manual gas valve in the trigger. The hardest part would be figuring out the acceleration curves, and finding the right connectors. The Spoolmatic connectors are a 10-pin connector keyed differently than the common ones you find. I think I found ONE that had the blue connector inserts - it still required slotting the inserts with a dremel, but wasn't too bad. The other several pigtails I made for adapting Spoolmatics to MK or Lincoln controllers had rubber inserts, which I pressed out in a vice, rotated, and pressed back in. Not as easy nor successful.

Anyways, that's a lot to digest... I have the pulser for my TA 250GTSW somewhere in the queue, and will post when I get it working reliably.

Hope this helps,
Jon

[Chadv]

Wow that’s a lot of info
ill keep you posted
thanks

[Weld dr]

Edit: I changed the inserts to be the correct ones for Miller remotes - 20-27 instead of 20-14...

Makes sense. Have you found the right connectors to put in a splice? You can usually read the connector code. You'll usually have something like MS 3106-20-27P on the connector.

The first four digits (310x) are for the connector configuration - you can get connectors that screw into a panel vs attach to a cable (difference between the panel mounted remote socket on your XMT or cord mounted remote plug) and have male threaded (like the panel socket) or female threaded collar (like your remote plugs) for locking to the mating connector. Check out 3101 vs 3102 vs 3105 (IIRC) vs 3106. There are right angle ones too, though I never use them. If you decide to put your remote in a project box near/on the XMT for future projects, the right angle may be the way to go.

The second two digits are the case size - size 20 in the case of Miller connectors.

The last three digits are for the pin configuration. I believe 27 is the right configuration for Miller, but that's off the top of my head. The P or S indicate if you have the pins or sockets for mating the individual conductors. Whatever you plug into the XMT will have a 27P at the end of the model number, and whatever you connect to your remote plugs will have a 27S.

Regarding board level repair - It's probably corrosion on one of the traces - you have a microcontroller (atmega16 instead of the atmega328 on your Arduino IIRC), that you'll want to avoid damaging because of the firmware on it, but everything else is replaceable. The issue should be obvious - corrosion on traces. If that isn't the issue, you should see a burned resistor (R74) You can also measure diode drops if that B to K resistance at the remote measured short.

The hardest part is removing the PC3 attaching screws without pulling the main board PC1. I don't ever remember pulling PC3 without first pulling PC1, but do remember a screw near the remote plug that would be hard to access.

Removing either board isn't too big of a pain because the top board is plug and play - almost every connector/plug is unique (if not every) and those that aren't can only reach their intended connections. But it's good to have the manual, give the capacitors time to discharge, and check the big capacitor voltages before touching boards.

You can put down an esd mat down for board troubleshooting and attach your esd bracelet to the XMT chassis/ground while removing boards to avoid electrostatic discharge/damage to be safe - I didn't while living in Washington state, but may do it for at least my first few projects while living in a very dry northern New Mexico. Up to you...

As for custom control box down the road, totally doable for all three, though the middle one is most ambitious (sensor less motor control). You have 24VAC available across pins A and G/K, so you can create +/- DC rails with a couple diodes, a few capacitors and voltage regulators. Linear regulators are likely good enough, provided you can find ones that can handle 24VAC rectified. Your DC common rail will not be board common unless doing half rectification (use only a single diode and connect the controller box common directly to pins G/K). If you want to do a +/- rail, your controller common is still tied directly to G/K, but you would use two diodes. Also, be sure to filter/clamp/protect all power and signal lines. Many options with PTC varistors, zener diode, crowbar circuits and fuses.

The gas solenoid is probably the easiest - and I would do it (at least initially) by having the solenoid timing by the pedal. You control preflow based on when you press the pedal (check pressed condition in your main loop or set up external interrupt if you're wanting to learn that, and use either a loop counter/delay/or timer interrupt for your post flow). Millers use a 24VAC solenoid in many machines (call tech to confirm part numbers for a 24v feeder) that you can find NOS on ebay, or you can find other solenoids on eBay as well. Be sure to put a varistor across the solenoid and a diode across the relay solenoid.

The 2T/4T and ramp/crater are a little more challenging because your controller needs to know the state of the arc - open or welding. The good news is the XMT has analog signals available that tell you output current and voltage. Most machines use output voltage to sense the state of the arc - Above ~50V and it's open, below that and the current is started (also how they time HF start). Since you're probably using lift arc, you may need to use both current and voltage feedback signals. Arduino is easy enough to do in main loop, especially since all of your timing is in seconds vs some much smaller time window.

You can find diy pulsing on YouTube using an Arduino and an analog multiplexer - just be sure you are aware of and using the right voltages for the components. Miller uses 0-10VDC for output current references. Depending on which microcontroller you use, you need to divide voltage and clamp (zener diodes will work) voltage references in. I would recommend creating your own 10VDC bus to use.for current referencing instead of using the 0-10V provided across pins C and D. This requires you to set Max and background off the machine, but saves you from damaging the board power supplies. It definitely means you need to clamp any voltage you feed to the XMT. This also may require isolation from the 24VAC circuit (I need to check in to that, my Thermal Arc 250GTSW has the current reference and board common as the same and I am thinking about a similar pulsing project).

Thinking more about the spoolgun controller, you can probably find ones that would work for a 35V, 2A(minimum) for under $30 now. The Spoolmatic 30As have a manual gas valve in the trigger. The hardest part would be figuring out the acceleration curves, and finding the right connectors. The Spoolmatic connectors are a 10-pin connector keyed differently than the common ones you find. I think I found ONE that had the blue connector inserts - it still required slotting the inserts with a dremel, but wasn't too bad. The other several pigtails I made for adapting Spoolmatics to MK or Lincoln controllers had rubber inserts, which I pressed out in a vice, rotated, and pressed back in. Not as easy nor successful.

Anyways, that's a lot to digest... I have the pulser for my TA 250GTSW somewhere in the queue, and will post when I get it working reliably.

Hope this helps,
Jon

Holy **** BATMAN.

[jjohn76]

Sorry, I got carried away. I spent too much time reading the XMT 304 patent, and posted something almost as confusing - tough read, even with the pictures...

https://patents.google.com/patent/US5783799A/en?oq=5783799

[Chadv]

Don’t apologize that post will be a useful resource for me. I think the most challenging part for me is going to be converting the digital 5volt pwm signal from the Arduino to communicate with the 0-10 volt signal the welder will be looking for on pin E(for ramp up then down using 2/4T tig control). I can’t find a suitable digital potentiometer. And the code needed to use them is currently over my head(I2c bus).

[jjohn76]

The AD-633 is a good way to multiply two 10V Analog signals if you're using the PWM from your Arduino. You would need a low pass filter on the output of the PWM. You can either use the XMT's remote output ( pin C or D) for one reference signal, or use a potentiometer in your box to generate the 0-10V peak reference. You would then need a voltage doubler (look up inverting opamp amplifiers, the first one would be -2x, then the second one would be -1x). You could do it with a single non-inverting opamp, but the math is easier with inverting opamps and I tend to have more quad opamp chips, so it's not really adding any components. Feed the AD-633 and all opamps +15V/-15V. You can use zener diodes to clamp the output signals to 10V. The AD 633 has an integrated output voltage clamp.

If you wanted to, you could control both peak and background in your Arduino then just use the opamp low pass filter and doubler circuits ( definitely clamp and low pass filter the reference output as well to be safe). Everything would be in software. You would just need to ensure your Arduino can calculate the analog write variables fast enough and the PWM has a higher enough frequency (at least two decades above pulsing frequency, more is better) so you can pass through the right pulsing signal.

The other way is with a digital potentiometer. This works better if your pwm is two slow Some can handle TTL logic signals on the input, I will see if I can find one. If you can't find one, you can always build level shifters with 2N3906 or similar transistors. Put a pull-up resistor on the inputs of the digital potentiometer, connect it to the collector of the transistor. Connect the base of the transistor to an Arduino digital output (with an appropriate resistor between), and connect the PNP emitter to board common. This will work for serial or parallel communication.

[jjohn76]

Also, there should also already be I2C libraries (it's the wire library for the Arduino). You just need to use the right pins ( the atmegas have hardware on them to write/read via I2C). Then it's just a matter of reading the digital potentiometer data sheet to send the right commands. You can practice with your code by printing the commands through uart to the terminal. You will definitely want to figure this out, especially when you decide you want a wireless tig pedal using the NRF, zigbee, or Bluetooth modules.

Yes, this will nickel and dime your walk and free time...

[Chadv]

So I fixed my contactor the the most complicated way possible.Thanks for your advice but i did something completely different.
I decided to send a signal to an arduino when the trigger was pressed then used the arduino to fire a solid state relay connecting pins I to J. I know this isnt required but this way I have more control over the contactor so later i can program burnback, pre/post flow and spot welding and im going to need arduino control anyway to do future upgrades

So i intercepted pin B heading back to xmt and ran it thru some of a broken 24v battery charger then a voltage regulator to get down to 5vdc.
I then sent that to an octocouple that pulled pin 3 on the arduino to ground

Meanwhile I made a panel that will screw to the side of my welding cart to hold all this stuff. its like a breakout board for the 14 pin connector

The gas solenoid is nearly done just need to run gas to it and power to it. Im using a small relay to connect the solenoid to 110vac on pins I and G

I created 10.11vdc +-by running the 24vac on pins A and G through a bridge rectifier passed a large capacitor and out a voltage regulator. I use this to power the arduino and later i think i will use a tl431 voltage reference with a potentiometer to fine tune my DAC and act as a zener diode

So as predicted the amperage control is the hardest part for me being new to all this. i bought a tlc7534 digital to analog converter but could'nt get it to work. So I ordered a mpc4725 DAC that is a module made specifically for arduino so hopefully i have better luck with that. I think i can use an op amp with the right resistors to double the 0-5v from the mpc4725 to 0-10v that i can then feed to pin E.


So thats all for now. i wonder if i should start a new thread titled custom control or something now that my contactor is fixed

help me out and tell me what I did wrong