XMT304 repair

[Bushytails]

I picked up an XMT304 on ebay that upon arrival was obviously known to not work, and the seller even cut off the power cord that had been shown in the listing. But, I'm broke and need to make it work, so dug into it. (My boss has a new policy of "doing more with less", i.e. make us work twice the hours for half the pay...)

Serial is MH432084U. CC/CV, 230/460, no special options.

I was able to find a service manual for an older version, but it quite obviously does not match the machine I have - PC1 in particular has changed from the schematics in the manual I have. Does anyone have a service manual for my serial number? It would really help!


Here's what I've done to it so far:
Initial inspection found R7 (200 ohm precharge) open. IGBTs, diodes, etc all not obviously bad. Jumpered a good resistor across it, powered it up, Help 7. No voltage on filter caps. Issue seemed to be with PC1, the top control board. Discovered VR1 was running hot, and the +15vdc rail was +3vdc. Tracked it down to an issue in what appears to be the lift-arc circuitry. Desoldered Q39, which appears to be the switch for a flyback circuit, and that took the load off the +15vdc rail. (I say "appears" a lot because the circuit in my machine looks absolutely nothing like the circuit shown in my outdated manual, so I'm guessing)

Noticed the aluminum bar to the negative output post had been arcing. Removed, wire brushed, re-installed.

Now had Help 6 on powerup. All rails low due to the control transformer being linked for 460 while running off 230. No voltage on filter caps. Two clunks from W2 on powerup, with 2.2V (at least that's how my meter decides to show the pulses) on the coil. W2 NO contacts still open. Jumpered a battery to the contactor, clunk, but contacts still open. Pulled out the contactor block, unable to manually actuate W2. Found mechanical failure of the W1/W2 interlock preventing contactor from pulling in, repaired, tested contactor, and put unit back together. The design and construction of the interlock is fairly ****ty - miller used a substandard contactor imho.

Unit now powers up, successfully links for 230, shows meters, etc. 337.2vdc on both filter caps. Capacitor charging seems correct, in that they smoothly charge to around 300V, the relay on pc1 clicks, then they suddenly jump to the final voltage. Current meter is stuck at 280A. Found hall effect sensor is putting out 2.8V with no current, and that doesn't change when either of the pots on the back of it are twiddled. Both +15vdc and -15vdc rails to it are spot on. Appears to need a new hall effect sensor. But, should operate in CV mode...

Tried welding. Will make some sparks for about a second, then shuts down with a Help 6 for a fraction of a second, followed up by a fresh precharge cycle. R7 (or, rather, the substitute I have dangling on jumper wires) gets mighty hot if you let it do this a few times. Measured line voltage at SR1 terminals, rock solid. Measured the +/-18vac output from the control transformer, rock solid. According to the outdated manual I have, help 6 is triggered by low voltage on the 24vdc rails, which are derived from the 18vac taps. Was going to measure the 24vdc rails, but it was getting dark and rainy, and my current work station is the driveway...


So, while waiting for weather that'll let me poke at it more, here's a couple questions I have:
Can anything other than the low voltage lockout on the 24vdc rails trigger help 6? For example, low voltage on the filter caps? I was going to measure them during a welding attempt next to see if they were holding solid. I suspect many things have changed since the manual I found.

Is it normal for these machines to have this many simultaneous problems? A mechanical contactor failure, a shorted fet or control issue in the lift-arc circuit (which would usually result in pc1 being thrown out, I'd figure), arcing internal connections, an open power resistor, a failed hall effect sensor, whatever is causing the unit to shut down when I try welding, and who knows what I'll find next. I bought miller because of previous good experience, but...

Any tips on where I should start looking for the help-6-while-welding issue? My current plan is to measure a few more voltages to see if anything is legitimately dropping too much while welding, and if not, try to determine if pin 14 is the only thing that triggers help 6.

Anyone have an accurate schematic for the lift-arc circuit on my serial number? I haven't poked at it to see what failed yet - just that all the current from the +15vdc rail was going through Q39. A quick glance at the board looked like it was a 555 driving Q39 with T3 and a couple freewheel diodes in a flyback arrangement, but I didn't analyze it fully. I don't see any feedback or obvious reason removing Q39 would do anything other than disable lift-arc functionality, but I'd still like to fix it, just in case I ever get the time to learn tig.

And, as I already said (at least a couple times , anyone have a new service manual?


Thanks!

[Noel]

You sir have done a lot of typing and I commend the effort. While I'm not able to offer any help or solutions to your problem, I will say your braver then most for trying, and your persistence will I'm sure pay off. I hope it does.

[Chrisheavymech]

Is there any resistance between the contacts on your main power switch?

[Aeronca41]

You know, with that list of problems it almost sounds like that machine was left plugged in and there was a lightning strike close by. While it is not practical in an industrial environment were the welders may be direct wired, I make it a habit to ALWAYS unplug welders before leaving the shop for the day. And if I happen to be using my Dynasty 200 and there's a storm, I stop and unplug it.

Besides lightning, there are other sources of short-duration high voltage transients on the power line, like an accident where a pole gets hit, and all of those sensitive electronics don't like it a bit.

Hope someone comes up with a schematic for you. I try fixing a lot of stuff without diagrams and it's a real pain!

[1997CST]

MH is still under warranty. Take to an authorized Miller repair shop.

[Weld dr]

Sounds like the warranty is void by doing what he did

[Bushytails]

Is there any resistance between the contacts on your main power switch?

As close to zero as I can measure without getting out my 4-wire meter. The uppermost terminal seems to wiggle even though the screw is tight, though.

MH is still under warranty. Take to an authorized Miller repair shop.

... I never imagined this unit could be under warranty! That's a new experience for me.

Sounds like the warranty is void by doing what he did

I emailed Miller with the serial number and the history, including the troubleshooting I'd done, and got back:
"This machine is still under warranty until December of this year. Take it in for repair. Working on it yourself will void the warranty."

So... I guess it's off to the local welding shop. That is not how I expected this repair to go. LOL.

[Bushytails]

Dropped it off at the LWS today. They say the shop they ship them to is running about a 7-8 week turnaround.

[cruizer]

Was going to say that this unit is still under full warranty as long as you didn’t happen to cut any wires excluding the power cord.

[Bushytails]

I've seen you help a lot of people with these... any chance you have a service manual for my serial that you can send me? I figure I'll need it at some point, and I'm curious how close my reverse-engineering was.

I'm also thinking of upgrading it to have the accessory outlet so I can have one fewer cords to run my spool gun, if you have a used pull of the larger, heavier control transformer that you'd want to sell cheap... thanks!

[Bushytails]

So, an update... It's still at the shop. It's been five months as of yesterday. I'm sure glad I'm not trying to run my business with that welder, or something...

The local shop doesn't do electronics repairs, so they shipped it out, with a 7-8 week expected turnaround. First apparently PC1 was backordered, then when they got them in, they still couldn't make it work. Thanks to the telephone game (you know, where a messaged gets repeated between people so many times that it gets garbled), I don't have any real info, but the guy at the lws thought they said something about it keeping on blowing boards. They have apparently given up after getting sick of unsuccessfully parts-cannoning it, and a month and a half ago shipped it to (supposedly) the miller factory service center. Haven't heard anything since. Seems to be how it goes.

So, I bought a second one, because I was sick of waiting. This one is not still under warranty. Initial inspection revealed one of the filter caps had a hole blown in the side, measured as a dead short, and most of the diodes and the scr in the input module were fried too. Couldn't find any other issues. IGBTs, bleeder resistors, diodes, etc all measured good. Looked like it already had a new power switch - judging from the plastic insulator, the old one failed rather dramatically. Put two new caps and a new input module in it today, and it welds great... up to 95A. Over 95A it shuts down with a help-6 when I strike an arc. Under it'll happily burn a whole rod with a nice stable arc. Anyone with experience working on these think that's likely to be a further problem with the welder, or (and I'm guessing this is the case) is the error likely correctly indicating my power sucks too badly? Thanks!

[1997CST]

So, an update... It's still at the shop. It's been five months as of yesterday. I'm sure glad I'm not trying to run my business with that welder, or something...

The local shop doesn't do electronics repairs, so they shipped it out, with a 7-8 week expected turnaround. First apparently PC1 was backordered, then when they got them in, they still couldn't make it work. Thanks to the telephone game (you know, where a messaged gets repeated between people so many times that it gets garbled), I don't have any real info, but the guy at the lws thought they said something about it keeping on blowing boards. They have apparently given up after getting sick of unsuccessfully parts-cannoning it, and a month and a half ago shipped it to (supposedly) the miller factory service center. Haven't heard anything since. Seems to be how it goes.

So, I bought a second one, because I was sick of waiting. This one is not still under warranty. Initial inspection revealed one of the filter caps had a hole blown in the side, measured as a dead short, and most of the diodes and the scr in the input module were fried too. Couldn't find any other issues. IGBTs, bleeder resistors, diodes, etc all measured good. Looked like it already had a new power switch - judging from the plastic insulator, the old one failed rather dramatically. Put two new caps and a new input module in it today, and it welds great... up to 95A. Over 95A it shuts down with a help-6 when I strike an arc. Under it'll happily burn a whole rod with a nice stable arc. Anyone with experience working on these think that's likely to be a further problem with the welder, or (and I'm guessing this is the case) is the error likely correctly indicating my power sucks too badly? Thanks!

OMG. It not rocket science. I could have had it back in your hands in 2 weeks. Sorry you're experiencing this much trouble.

[Bushytails]

OMG. It not rocket science. I could have had it back in your hands in 2 weeks. Sorry you're experiencing this much trouble.

I doubt it needs much more than PC1, a hall effect sensor, and a precharge resistor... I had it making sparks! Anything that makes sparks is obviously not a total basket case. ****, PC1 was probably pretty easily repairable too, though I'm sure a shop is just going to throw a new one in for a warranty repair.

For the second welder, I need to rig up a dummy load so I can do measurements at power. Roll of wire in a bucket or something. I suspect the problem is my power supply, but I can't both burn rods and watch meters at the same time. Could set up the 'scope and record things, but a dummy load would be handy to have anyway.

[Bushytails]

Well, took XMT 304 #2 to someone else's house today, and it welded great at 200A, so the problem is indeed my power. The repairs I did seem to have been successful, and it welds perfectly smoothly.

Background: The power here is a large dc-ac inverter system, that can supply >40A at 240V for welding duty cycles... but I'm guessing it's not playing well with the non-resistive input current of the welder, and either is having problems with the tops of the wave getting chopped off, or there's a slight lag when striking the arc before the inverter's voltage regulation kicks in. The inverter itself has no problem with the overall load, and isn't setting any warnings, including its low ac output warning flag. The same inverter powers my older, smaller transformer-based welder easily. I'll need to get my 'scope on the welder input to see what's up.

Anyone know what triggers the help 6 code? When I looked at the schematics for an earlier model while working on the first XMT 304, it looked like it was generated by measuring the rectified 18VAC from the control transformer. Is this correct? Is this the only thing that can set it, and not, say, low voltage on the filter capacitors? Trying to think of the easiest way to make it stop complaining without actually fixing the problem.

I think minimal effort try #1 is going to be a buck-boost transformer, and see if it works at 246, 252, or 264vac...

EDIT: Still no update on XMT 304 #1. I don't know if anyone even knows where it is at this point.

[Franz©]

Machine 1 sounds like WALL JOB. Factory can't fix to spec within budges they shove it against the wall and hope you give up.

Send an email with BCC to yourself saying you are far less than happy with the progress of this WARRANTY repair, and you expect confirmation in writing signed by a Corporate Officer the machine warranty will be extended an equal number of days it has been out of your possession.

That machine will never come back if you don't rattle cage.
Goven Suprglobalmegacorp's 80/20 policy it may not anyhow, IF it still exists and hasn't gone back to China to beBorn Again.