Announcement

Collapse
No announcement yet.

XMT304 Help7

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Just for giggles..takes just a few minutes... pull the back vent panel off.... pull the insulation boot off of the wire connection back there and make sure that connection is sound...

    Comment


    • #32
      https://youtu.be/VnlEEpkl7Og

      Okay, I took a small video of the display and the clicking that I hear. Hopefully this works. I’ll do more troubleshooting in the morning

      Comment


      • #33
        You might want to actually test the IGBTs, if there's something odd with their measurements. I had an issue on a dynasty where the IGBTs tested OK with the diode check on a multimeter, but one didn't actually work, and never turned on even with a proper gate drive. Or, if you have an oscilloscope, see what kind of waveform (if any) it's actually making.

        Comment


        • #34
          Thinking more about your error code, a voltage imbalance across the two caps doesn't make sense because they're in parallel - they would both sink together. That imbalance error makes more sense on 460V, when the capacitors are in series. Just my thoughts, as I have never troubleshot that code. I if you're getting something on the output, both IGBTs have to be firing (or at least conducting) because this welder runs in a full bridge configuration. It sounds like you're dealing with an overvoltage, which my guess is because one of your snubbers isn't working properly. Or like popscott mentioned, it could be a bad connection at the transformer that's causing arcing/voltage spikes. You won't always see the spike with a multimeter because it's a transient spike, though you may see a slight uptick in voltage (compare the 332V DC bus to your service AC voltage time 1.414, minus about 1.8V drop for the rectifier diodes and ace) depending on how the spike is handled. On another note, R8 and R9 happen to be 150k ohms each, and if they were for some reason in parallel with your gate resistor, would give you 75k ohms instead of 100k ohms. If you take off the interconnect board, you will be able to isolate what's what. They shouldn't be in parallel with the gate resistors, so you will probably need to separate the board from the components to test each indiviudally. Don't remove the components from the heatsink unless they're gone.

          Comment


          • #35
            Hrmm, was the serial number ever posted? I don't see it just re-reading the posts.

            Bad contactor contacts can cause them to be unbalanced.

            At least on the schematic I have, over/under voltage detection is based on the 24V outputs from the control transformer, and has nothing to do with the main caps, scr/rectifier module, or any such associated circuitry. Bus imbalance is an entirely analog circuit with a pair of opamps.

            The precharge timer is tied to the help-6 and help-7 lines. If help-7 is shown, it'll reset the timer and go back into precharge mode. This is most likely the repeated clicking you hear.

            Looking at the unbalance detection circuit, an issue with the +15 and -15 rails could cause a spurious unbalance signal. Have you checked these to make sure they're both good? Are any of the larger heatsinks on the top board abnormally hot?

            Comment


            • #36
              Originally posted by Bushytails View Post
              Hrmm, was the serial number ever posted? I don't see it just re-reading the posts.

              Bad contactor contacts can cause them to be unbalanced.

              At least on the schematic I have, over/under voltage detection is based on the 24V outputs from the control transformer, and has nothing to do with the main caps, scr/rectifier module, or any such associated circuitry. Bus imbalance is an entirely analog circuit with a pair of opamps.

              The precharge timer is tied to the help-6 and help-7 lines. If help-7 is shown, it'll reset the timer and go back into precharge mode. This is most likely the repeated clicking you hear.

              Looking at the unbalance detection circuit, an issue with the +15 and -15 rails could cause a spurious unbalance signal. Have you checked these to make sure they're both good? Are any of the larger heatsinks on the top board abnormally hot?
              Serial number is MA410436a. The heat sink on the right is about 30F hotter than the left one. Haven’t gotten a chance to test anything else yet

              Comment


              • #37
                That's a lot newer of a unit than I have a schematic for, unfortunately.

                I'd count that as abnormally hot, as I seem to recall they both ran only a tad above ambient in a working unit.

                Something could be loading down one of the 15V rails, which could trip the unbalance signal without an imbalance on the main capacitors. It's worth unplugging HD1 to see if that's the issue, or if not, using your thermal imager to look for anything else running unusually warm. I worked on a unit with a shorted mosfet in the lift arc sense voltage generator that was pulling down one of the 15V rails, which was next to a little transformer on the top board, might be worth imaging it too.

                Comment


                • #38
                  Originally posted by jjohn76 View Post
                  Thinking more about your error code, a voltage imbalance across the two caps doesn't make sense because they're in parallel - they would both sink together. That imbalance error makes more sense on 460V, when the capacitors are in series. Just my thoughts, as I have never troubleshot that code. I if you're getting something on the output, both IGBTs have to be firing (or at least conducting) because this welder runs in a full bridge configuration. It sounds like you're dealing with an overvoltage, which my guess is because one of your snubbers isn't working properly. Or like popscott mentioned, it could be a bad connection at the transformer that's causing arcing/voltage spikes. You won't always see the spike with a multimeter because it's a transient spike, though you may see a slight uptick in voltage (compare the 332V DC bus to your service AC voltage time 1.414, minus about 1.8V drop for the rectifier diodes and ace) depending on how the spike is handled. On another note, R8 and R9 happen to be 150k ohms each, and if they were for some reason in parallel with your gate resistor, would give you 75k ohms instead of 100k ohms. If you take off the interconnect board, you will be able to isolate what's what. They shouldn't be in parallel with the gate resistors, so you will probably need to separate the board from the components to test each indiviudally. Don't remove the components from the heatsink unless they're gone.
                  Its definitely a pretty new unit I think around 2010? I’ll definitely try to make sure all the connectors are secure to the transformer and I’ll check HD1, but I doubt I have the expertise to remove the board and test individual components. The again,I probably can’t break it worse, may give it a try.

                  Comment


                  • #39
                    It seems daunting, but not too bad. You can check pretty much everything on that interconnect board with your multimeter and walk through the schematic. Two things you may also want is an IGBT tester (you can make them too, but they're pretty cheap) and maybe a torque screwdriver (I use my wheeler screwdriver that I bought for mounting scopes) to get the torque on the screws right. Take pictures before and after. Most of the leads should already be labeled in marker from the factory.

                    Comment


                    • #40
                      Bushy, good point. The balance detection measures the bus capacitor voltages and raises OV1 high if there is an issue. Which VR is that on the heatsink?

                      Comment


                      • #41
                        Originally posted by jjohn76 View Post
                        Bushy, good point. The balance detection measures the bus capacitor voltages and raises OV1 high if there is an issue. Which VR is that on the heatsink?
                        The second part of the balance detection uses the +15V and -15V with some dividers as reference voltages, and if either of them is wrong, it might trip the unbalance detection even with perfectly balanced capacitors. The hot heatsink could be one of the rails being overloaded, and the extra load of actually switching the output on causes the regulator to go into protection (those 3-terminal regulators have both overcurrent and overtemperature protection), the rail to drop, and an imbalance signal to be generated... which activates the help-7 line, which causes the controls to turn the inverter back off, allowing the rail to rise again, and the precharge timer to begin another precharge cycle...

                        _If_ this is what's happening, then the task would be determining why the rail is overloaded. It could be a constant load, like the hall effect sensor or a shorted component randomly on any of the boards, or it could be something that's only active when trying to fire up the inverter, which is why the code only appears when selecting a hot output mode. Having a thermal imager might make this pretty quick, just see what's getting hot...


                        Comment


                        • #42
                          Makes sense. Unless the layout changes, it looks like that is the +15V power supply, so the overcurrent could come from any board. Since there is a test point across the imbalance circuit (first opamp) you could get an indication if that is your issue as well. You should be able to find a component on the board to check the 15V bus. I didn't see one on the test point connector, but I may have missed it.

                          I keep going back to the pre-charge SCR. For some reason I thought the gate drive was powered by PC1, but looks instead like it's driven by whatever voltage sits across SCR. C5 holds the charge to power that gate, and since it's toast, I wonder what that SCR is doing, particularly when the output is on and the bus capacitors are draining faster.

                          Comment


                          • #43
                            Easiest place I can think of to check the +/-15 is HD1, since it has both... and as they seem to be a common failure, a good place to check for an abnormal load as well.

                            PC1 just switches dry contacts for the SCR with the little relay.

                            Comment


                            • #44
                              Ah, pulling HD1 makes sense now - not to check HD1 just to check the +/- 15V circuits... It makes sense the +15V is drawing more current - it powers the bulk of the components on that board, but maybe not to the point of getting that hot. The only warm spots I remember on that board is a couple of the ICs about 1.5-2" in from the edge just about above the rear bus capacitor to rear contactor. A5 if I remember right , possibly also U7. The gate drives (back corner just about above C1 tank capacitor) didn't show any temperatures over the surrounding area. Are there any hot spots or temp differences between the components? Just a thought if it's.in the gate drive related to the PM2 off measurements.

                              Tracking the SCR relay is dry contacts. I don't think it is affecting the PC1 power supplies, but the charred board/components on PC2 are likely causing some issue that would need to be addressed.

                              Comment


                              • #45
                                Okay I’ve attached a thermal image of the entire top board. Bottom right corner hot spot is the 20k ohm resister r3. The other hot spots are R250 and I believe it’s hs1/vs1(right heatsinked component). Nothing else is showing up as hotter than ambient.

                                Picture 2 is what I tested assuming it’s HD1.Red was +15v dropped to 12.9 occasionally as the error code cycled. Black was -14.9 constantly. I tried pulling HD1 to see what would happen without it but still error coding. I never saw red jump over 15v but also my multimeter may not have the refresh speed to catch it if it did.

                                A multimeter and a thermal camera are the extent of my diagnostic tools right now I will look into an IGBT tester but don’t see myself getting an oscilloscope any time soon.

                                I guess I need to see about ordering a new c5 and c6 that should fix that board, granted they may fry again if I don’t figure out the underlying issue.
                                Last edited by Metro424; 08-24-2020, 08:35 PM.

                                Comment

                                Working...
                                X