Kempomat250

[Swedespeed]

Cheers,
I have a similar problem with my Kempomat 250. There is no welding voltage/current but gas solenoid and wire feed still operates normally.
On closer inspection I noted that two wires between the course power setting knob and the primary winding on the transformer phase located closest to the front panel of the welder have molten insulation and one of the copper wires is actually completely melted off. Both overheated wires unsurprisingly lead to the same coil pair, which is the highest power primary circuit primary winding on the course power setting knob.

I have disconnected and measured all the coils on the failing phase and compared them to corresponding coils on another presumably healthy phase and they check out well. My thinking is however that even one shorted turn on the primary winding (which would not make a notable difference in measured total coil resistance) would create a one turn shorted "ring" around the transformer core which when induced by the magnetic field of the transformer would create a huge load and therefore start melting the primary feed wires. This rather catastrophic failure mode would probably render the transformer useless so I would much prefer the problem to be with the switches where I may be able to make some repairs or a workaround and am therefore curious to investigate this further before giving up, but I do not fully understand the reasoning of eddie49 how a failed switch would cause one primary coil to get over loaded or shorted. If eddie49 reads this, could you please elaborate a bit?
Anyone else here that have experienced a failed power setting switch cause burned primary wiring? The cases of failed switches I have read about online have only caused that setting to become inoperative, caused by the switch no longer being able to power that coil.
Also, I am curious if the repaired switch on Nielsie 's machine solved his problem. One thing I noted from the picture of his transformer is that the black wiring going between the switches and the transformer looks like it has been re-routed at some point. From the Kemppi factory these wires are neatly bundled, organized and tucked above the transformer, at least they are on my machine.

The circuit drawing that eddie49 is referring to is helpful but the resolution is unfortunately so low that it is impossible for me to see the numbering of the leads which would be very helpful in further trouble shooting. Does someone here have a circuit diagram with higher resolution to share or even better a service manual for the Kempomat 250?

If the worst case scenario is true and the transformer is shot, does anyone here have a Kempomat 250 with broken switches or broken electronic board etc. but with a healthy transformer that would be interested in selling it to me? Any ideas on where to find a replacement transformer would be appreciated. I may even be able to use one phase from another shorted transformer, as long as it not shorted on the same phase. I then need the first phase winding which has a sensor winding of some sort that makes it slightly different from the two other phases.

Thankful for any advice or input on how to save this old but otherwise great machine!

 

[eddie49]

Hello and welcome to the Forum! And Gott Nytt År!

It is interesting that the problem you are seeing on your Kempomat 250 - melted feed wires between voltage switches and transformer primary coils - seems to be the same as the earlier one in this thread, just over a year ago. This is an unusual and rare fault, and I think it is again caused by a faulty switch.

My reasoning is as follows:
- transformer windings seldom fail ( especially on a large high-quality one such as this )
- you have measured and compared the coil resistances and they are similar

Rather than a single shorted primary turn, I think that a section of primary winding between feed and a tapping-point has been shorted, but externally by a faulty switch. There are many turns on the primary, and the voltage applied across or induced into a single turn is low, so shorting one turn internally would not produce much current flow.

Even if a section of a transformer winding is not connected at both ends to a supply or to a load, the "auto-transformer effect" means that voltage will still be induced in the coil, and it would flow as short-circuit current in the ( black ) feed wires if a failed switch joins two taps simultaneously.

I would suggest tracing the melted wires to the exact section ( segment or wafer ) of the Coarse power range switch, and then verifying correct operation of that switch with an Ohmmeter. It should effectively be three 1-pole 4-way switches, but the way these switch sections are stacked together and linked may make it difficult to decipher. As mentioned in Post #14, you need to check for continuity from the moving pole to one output in each of the four positions, and also confirm no false conduction to the other three output screw terminals. If in doubt, you could test the welder with just one tapping wire for that coil connected to the Coarse switch, and the other three tapping feed wires not connected and insulated.

We have not heard from member Nielsie since his Post #19 in December 2023, but it sounds logical that repairing the faulty cam switch did fix his machine. As for the disturbed wiring layout, maybe his photos were taken after he had already cut the cable ties and inspected the melted primary feed wires.

 

[Swedespeed]

Thank you for your reply Eddie, very much appreciated and Happy New Year to you as well!

I will do some more measurements on the switch to verify that only two connections per phase is made in each power setting. Come to think of it the switch does not move very easily so there may indeed be some electro-mechanical issues in the switch. If the switch is the culprit, I assume that only one of three phase sections of the switch is faulty, so I will therefore compare the sections with each other and see what I find. The fine and course switch look identical so I will also cross check between switches.

Another thing I forgot to mention previously is that one of the jumper wire flat pin connectors that has to do with the 220 vs 380 3-phase configuration was very loose. I cant imagine how this could have anything to do with the melted wires but I will make sure to tighten this flat pin connector before testing the welder again.

These welders are very thoroughly built, with each connection point marked and numbered and even the wires are individually marked. Quite impressive attention to detail and great for servicing and trouble shooting. Too bad then (due to the course image resolution) that I cannot read the corresponding numbering on the schematics shared earlier in the post which would help identifying the coils and thereby make trouble shooting easier.

Thanks again, I will do the measurements and further investigations in the next few days and get back with my findings.

 

[Nix2]

Hello.
You can check if the main transformer is working (if there is no short circuit) without the switch/switches.
Then you connect the strongest and then the weakest run "for a short time" and observe for a few, a dozen or so minutes - a working transformer does not heat up (almost at all), draws a low idle current (below 0.5A for this Kemppi) and hums quietly (does not make noise).

The same procedure can also be used for some (constant) load.

If so, the transformer seems to be working. Then the range switch/switches are particularly suspicious

Regards!

 

[Swedespeed]

Hello, here is a brief update:

I have now confirmed that the "course" selector switch simultaneously connected two primary coils in parallell and those coils as expected were supplied by the wires that had overheated and melted the insulation as well as broke (melted) one of the wires completely off. I also inspected the "fine" selector switch and found that it also is at the end of its useful life, being very hard to turn and sometimes being entirely stuck. After dissecting both switched I found the hexagon holes in some of the plastic cams being quite worn and on the verge of skipping/slipping on the selector axle.

I have therefore (maybe somewhat prematurely?) ordered two new 20 Ampere 3-phase 4 position switches that I will connect in order to test each transformer coil.

Thank you Nix2 for your input and advice. I agree that bypassing the switches and manually connecting each coil and thereby separately testing them one by one would be a good approach. Too bad I didn't think of this.

I have a slight concern that a current sufficient to melt the external wires would also cause damage to the transformer coils. I however see no sign of damage on the visible parts of two primary coil wires that were connected to the melted external wires. Could it be that the mass of the transformer including the surrounding layers of inactive coils offers better short term cooling and protection than the supply wires have? Or maybe the coil wire has slightly bigger cross-section that lower resistance in case of short circuit compared to the external wiring? Hmm, I wonder how much heat would it take to damage the protective lacquer that insulates the coils.

Well I guess I am about to find out once the switches arrive and are connected and i will let you know once I test this.

Best regards.

 

[Nix2]

It is hard to say whether the winding "inside" was overheated or whether the insulation of the wires held up and is undamaged... in fact, you should do a few, a dozen or so minutes test under load (heated transformer) and then it will turn out whether it is OK.

Regards

 

[Swedespeed]

It is hard to say whether the winding "inside" was overheated or whether the insulation of the wires held up and is undamaged... in fact, you should do a few, a dozen or so minutes test under load (heated transformer) and then it will turn out whether it is OK.

Regards

OK, thanks for the input. I have now configured and wired in the new rotary switches but am still waiting for a few more parts. Will likely be able to start testing in a week or so.

I assume that one failure mode could be that some power settings (transformer windings) are OK and some are damaged? First I plan to test the various power settings without a load, just pressing the trigger and measuring DC output voltage making sure I got everything wired correctly and that there are no internal shorts in the transformer. I will use an IR camera to monitor external wiring temperatures (between switches and transformer) making sure temperatures stay low and even when comparing the three phases. I will also take your previous advice and listen for any excessive humming. I don't have a good way to add a load other than welding so if the first test goes well, I plan to try some short welding sessions, starting with low settings and gradually moving up in power.

 

[Swedespeed]

Success!
After some waiting for parts and a few other projects that came in between, here is a brief update on the results from testing of my welder so far:

Idle voltages were measured in all combinations of power settings and the voltage increased incrementally as power settings were increased, showing that I probably bly got all the wires connected correctly. This was nice to see as the new, non-standard rotary switches necessitated all switch connections to the primary coils to be completely different from before.
There was no humming from the transformer during prolonged idle voltage testing and the IR camera showed no heating of the wires feeding the transformer.

I then tried laying some test welds on all power levels which worked fine as well!

Time and longer welding sessions will tell if there is lingering damage to the transformer that has not yet showed itself but I am now relatively optimistic the new switches and replaced burnt wiring has solved the problem.

Here is a link to the 20A rotary switches i purchased to replace the worn out original ones: https://www.aliexpress.com/item/100..._main.147.51f75e5bQinEuU&gatewayAdapt=glo2fra
These switches need to be externally re-configured by bridging the input sides of (each phase separately). This is easy to do but there are many connections and outputs to keep track of to ensure the switch will operate as intended, so unless you are used to this kind of work (which I am not!) it pays to work slowly and methodically when wiring in the new switch.

I want tho extend a big thank you to eddie49 and Nix2 who provided such valuable advice and to everyone else on this great forum that has given thought to my questions. Without your input this project would have been much more difficult.

Happy welding everyone!