I found this welder at KMS apperantly it came from a school but was untested so I picked it up for $100.00 bucks. It turned out one capacitor realeased the magic smoke and I replaced the cap bank with some out of a 250 machine seen below I also put a new liner and some bits on the gun.
Nice find. I've helped fix a couple of welders in the last couple of years. The fix was usually something pretty easy on the main board. The welding shops only know how to replace a whole board - even is the problem isn't the board. If the board is not available (common for Chinese gear, or something a bit older), or is too expensive, the welder can be had for a song.
In one case, the logic power supply circuit on a Chinese design actually used the "on" led as part of the 5V supply. Whenever the welder was turned on after being off for a while, it caused a 250ma+ current spike to flow through the LED for a couple of milliseconds as it charged the filter capacitors. The LED is only rated for around 10ma continuous and probably only 50ma peak. After a few hundred such cycles over 1.5 years (enough to get the welder off warranty), the LED got tired of the abuse, blew up and went open circuit. Since the LED was part of the 5V power supply (not just lighting up when the supply was on), the 5V supply quit and the machine did nothing. I replaced the LED with a 1A power diode (good for 10A spikes), along with a zener diode to control the 5V supply and the welder has worked for several years since. I then added a LED with a current limiting resistor as the "on" LED. Total repair cost: < $2 - all from my parts bins.
In another case the problem was blown main rectifying diodes (SCRs actually) on a big TIG welder. Found the replacements on e-bay for $20 each - $150 each from retail suppliers).
Thats where it is nice to know more about electronics unfortunatly I'm not one of those although I have a basic knowledge that has worked so far.
I haven't played with the welder lately because my garage is getting reskinned and siding put on that's years overdue along with a false gable towards the alley to mount the external cctv system it should be good I'm just waiting for delivery of longer cables to reach the alley.
I saw that machine at KMS - well done getting it going Tom! All that used stuff they have at the back "Scott" at KMS says they get from the School Board. There's a plasma cutter in the pile right now if you're willing to take a risk... and some other welders too.
The old ones were 140,000 mf and the new ones were 150,000 mf he said it will run just a little bit smoother, it was a straight swap though and bolted right up the capacitor bank was only $300.00 a bit much for a test but I figured if it couldn't be fixed hmmm maybe a induction heater.
Well @Tom O and @JohnW and the rest you have encouraged me not to fix my welder, but to kindda fix my garage furnace! The unit quit working on me flashing with an error #2 code. I wasted a bunch of time, took it apart, tested and replaced the flame over switch. No help. I then had the furnace guy in, he charged me and couldn't fix it. He didn't even take the board out. Disappointing. Said the board was bad and the furnace is an odd one (Stirling brand, no one carries it or parts), and board if he could find it would be $500. His only anwer is $1300 for a new one. Great. Repairs are so hard these days, gotta be a renaissance man for everything.
Anyway since JohnW said board problems were common and often easy to fix I pulled the board out and started looking at it. Hey I found cracking on some pins. So I resoldered - the furnace now lights and runs. Kinda. The flame won't stay on for very long (30s-60s) and it then cuts out. Then it cools and starts up again. Warms the garage right now but I'm not so sure about winter. I think it used to run for longer... Any ideas???
Is there any error code now? Obviously you know how to read the flashing LED to get error code #2.?
Did you re-solder all of the pins? If one has cracked, there may be others that are not as obvious. Like the slightly out of focus one down and left of the obviously bad one in your picture also has a crack. Reflowing every pin will sometimes fix a non-obvious bad connection. If it is double sided board, carefully check all the vias (connections through from one side of the board to the other). If one or more pins had cracked the odds are not bad that there may be others that are not visibly cracked, but may fail once they heat up a bit.
Thermal cycling is often really hard on through hole boards like this. If you don't heat the shop all the time, the board is cycled from at least -30C to +30C on a regular basis. The board material expands at a different rate than the copper wires, traces, and solder, so it can break the connections over time. If there are vias, it sometimes helps to solder in a short piece of bare copper wire in case the plated via has cracked where you can't see it. It also makes it less likely to break again in the future.
Look for any electrolytic capacitors that show any sign of bulging at the top (or leaking). If the top is not flat or slightly concave, think about replacing them.
Finally the flame sensor is a common problem. Does it have a short stainless steel wire sticking out into the flame? That is the most common type of flame sensor. The plasma of a flame is a reasonably good conductor of electricity, while air and natural gas are not. The flame sensor attempts to pass a small current from the flame sensor wire to ground. If no current flows (it should flow in the range of one to several uA), it thinks there is no flame, and so shuts off the gas. The sensor wire often gets a thin coating of corrosion on it and so it does not conduct properly, and the controller thinks the flame is out. Fix it by polishing it up with steel wool to remove the corrosion. I asked about the error code since this type of failure should flash out an error code.
Thanks John. I'll check on these things. The board is one sided. I soldered those two suspect pins. I'll do more and check for bad caps. The flame sensor I did clean and the igniter rods too. The furnace guy did do that too. There's no error code flashing now just normal code. Great advice John thanks.
I don't think that most furnace controllers that have current based flame sensors have thermocouples. They both serve the same purpose - sensing the flame and shutting down the gas if there is no flame. Thermocouples are usually used in pilot-light based systems where the thermocouple makes sure the pilot light is working by generating the current that keeps the gas valve open. The newer microcontroller control boards tend to use the current based flame sensor for the flame, and other sensors measuring vacuum and or air / exhaust flow to monitor the system. I could be wrong. If it has both I'd be interested to know.
The furnace is the new type with the flame sensor but no thermocouple. It has a vacuum sensor on the exhaust system and two sensors for over heat connected in series. One mounted in the flame box the other mounted near the heat exchanger. There is also the flame igniter in the flame box - two stainless bars separated by a gap where a high voltage spark jumps to ignite the gas. There are quite a few parts...at least compared to the thermocouple and pilot light type. Here is the board. And the other shot shows the flame box (?) in the top right. It has the flame sensor, ignitor, gas nozzle going in.
It is poorly designed for servicing. You have to take all kinds of stuff apart to get at anything, no room for tools and fasteners are hiding behind stuff.
There are only a couple of reasons it would be shutting down.
The first is easy - the thermostat has requested that it turn off. It won't immediately turn off. It will shut off the gas relatively soon (within a few seconds), the burner air / exhaust fan (usually the same motor with a fan on each end) will turn off next after 5-20 seconds as it wants to be sure that the exhaust has in fact been exhausted. Finally, it will take 30-60 seconds to cool off the heat exchanger shut down completely. Could the thermostat wiring have a bad connection?
The next reason is that it has determined that something is wrong. It could be the flame sensor or of the other sensors is not doing what is expected. For instance if there is gas flowing, there should be a flame. There should be a vacuum indicating that the combustion air / exhaust fan is operating, the heat exchangers should be heating up, but it should not get too hot. If any of these conditions shut it down it should start flashing the error LED. It was apparently doing that when it has the bad solder joints, but that is fixed now, and as I understand it, there is no flashing LED, so it is not intentionally shutting itself down for a safety reason. When these sorts of errors occurs it will often run through one of more cycles to try to run again, and will only fail solidly after two or three failures. The retry cycles will often involve running the fans for 30-60 seconds with no ignition to try to clear any gas that may be hanging around.
The third reason is that something non safety related is not actuating correctly. There could be a bad relay or some other bad connection that is not actuating something properly. It could also be a bad transistor on the board that is not actuating a relay properly. The CPU can usually only control 10-120ma on each of its pins, so a transistor will be used to switch the relays that probably need 50-100ma to be turned on. The trick is to find what it is that is not working.
The way to troubleshoot this stuff is to put a meter across something that you think should operate in some way and watch it as the systems cycles. Particularly look for intermittent signals, or weak signals. Sometimes a connection will not fail until current has flowed through it for a few seconds and caused it to heat up, or the connection is being made, but it has too much resistance. This is where the troubleshooting starts to get a bit more complicated. On high-impedance circuits very low currents are used, which are very susceptible to bad connections and noise. Make sure the spark igniter wire (orange?) is kept away from other wires. None of this is easy to do in something like a furnace which will refuse to operate at all when all the subsystem are not connected up, and is probably hanging from the ceiling in an awkward position.
If you crank the thermostat to far above room temp (40C) does it function properly? Could be that you're just too close to the goal temp, so it only kicks on in short stints?
I had a similar issue, it turned out it was two part: The blower motor was weak, and the vacuum switch (shown in your second pic, top center) was tough to switch.
--> Thinking back, the issue was near identical, the furnace would kick on for a bit, then turn off, and cycle like that. In the winter my house went from room temp, down to 14C, but took a period of days to do so. Becuase it cycled on and off just enough to almost maintain temp.
As a workaround while diagnosing, I jumpered the two leads coming out of the vacuum switch once the furnace was actually on (20 second window). I used the appropriate amperage fuse to make the jump, just in case.
--> For testing, try the same, jump the vacuum switch leads at difference stages in the cycle!
----> When are you available to hangout, I could bring a fuse and try with you.
Edit: Jumper goes in to replace the vaccuum switch when there is actual fire in the furnace. If you jump it before hand, then start the furnace, it will error out that the switch is stuck. Need to start like normal, then switch when you see fire. If it stays on for a full cycle (1/2 hour or w/e) then you have found the issue!