60W Power supply keeps failing

[audiomath]

I just lost another 60W power supply. This makes a total of 3 in one year, 2 of which are the "Jupiter" brand power supplies I bought from Light Object. (the first was a very cheap one I bought from another vendor with the tube)

When I install the power supplies I put the anode wire inside a piece of silicone rubber tube, and I cannot find any arcing after a failure. I happened to be standing in front of the machine today when it failed. I heard a loud squeal, then a "whuf" sound. Now there is a burned smell and no HV output.

My primary use for the machine is cutting 3mm clear cast acrylic. I use 12mm/second at 60% power to cut the plastic. I don't use the machine a lot, but when I do it will run for several hours. I have a 600W refrigerated chiller cooling the tube, and the coolant never gets hotter than 21C.

What am I doing wrong? These failures are expensive and time-consuming...

Thanks.

[Techgraphix]

I think you are just unlucky.. If the environment ain't hostile (moist, dust, very high or very low temperature etc), i can't think of anything that can cause it. If all HV-wires are kept at quite a distance of the metal frame and the connection to the tube is solid, it should be fine.
If you read the current during cutting (60-70%), it should read something like 15-18mA.. is that correct?

Kees

[audiomath]

I think you are just unlucky.. If the environment ain't hostile (moist, dust, very high or very low temperature etc), i can't think of anything that can cause it. If all HV-wires are kept at quite a distance of the metal frame and the connection to the tube is solid, it should be fine.
If you read the current during cutting (60-70%), it should read something like 15-18mA.. is that correct?

Kees

Yes, around that. The machine doesn't have a cathode current meter, but I built one into a box and put a normally-shorted jack in the laser machine sheet metal for the cathode connection. I only plug the meter in occasionally. I say "around that" because I doubt the cheap Chinese-made panel meter I put in the box is very accurate, and I don't remember the exact reading. Less than 20mA, in any case.

The environment is very clean, with humidity controlled at 34-40% and almost no dust. We just replaced the HVAC system this spring, and installed a new MERV 16 air filtration system. It isn't a formal "clean room", but since we work with surface-mount electronics it is cleaner than most labs, I think.

This is frustrating. The expense is a factor, but the biggest issue is the time it takes to change the power supply, do test cuts and "tweak" the job settings. There is enough variation in the PWM/output between different power supplies that I can't just change the power supply and go back to work, unfortunately. After the first one failed, I put a HV connector in the anode wire to the tube, so at least I don't have to disturb the tube when changing the supply.

Thanks for the reply.
--jim

[Tech_Marco]

I think you need to pay closed attention to your laser tube. 3 power supply broken in a year sounds unbelievable. When the laser power supply arcing, either the transformer of LPSU defective or the laser tube is leaking gas!

Not sure why you unplugged the current meter. To us, it is one of the most important device that needed for job and troubleshooting. Remember that an analog meter is cheap and no reason to be expensive, regardless Chinese made or Japan made. We have those meter installed on all our machine so far I have 'none' defective.

If you have a 'coupling' connector on the HIGH side, you better double check the connector. It seems OK but internally could be bad. When the surface of the contact got oxidized you would get in trouble like arcing. That's why all our power supply comes with a single wire, direct, and no coupling because of this high risk case

Marco

[Techgraphix]

When the surface of the contact got oxidized you would get in trouble like arcing.

That's why the better quality HV-connectors are silver or silverplated . Ag2O or silveroxide, conducts even better than silver
But, still, the plug needs to be mounted on an isolating (preferably PTFE) bracket, a few cm's from the metal.

Kees

[audiomath]

I think you need to pay closed attention to your laser tube. 3 power supply broken in a year sounds unbelievable. When the laser power supply arcing, either the transformer of LPSU defective or the laser tube is leaking gas!

Not sure why you unplugged the current meter. To us, it is one of the most important device that needed for job and troubleshooting. Remember that an analog meter is cheap and no reason to be expensive, regardless Chinese made or Japan made. We have those meter installed on all our machine so far I have 'none' defective.

If you have a 'coupling' connector on the HIGH side, you better double check the connector. It seems OK but internally could be bad. When the surface of the contact got oxidized you would get in trouble like arcing. That's why all our power supply comes with a single wire, direct, and no coupling because of this high risk case

Marco

You may have seen my other post showing what I assume is arcing in the "output end" of the tube.
You have a good point about using the meter. The reason I unplug it is because having the cathode circuit extended by a cable is potentially dangerous. I need to cut the sheet metal on the laser case and mount the meter permanently.

Now the decision is whether to replace the 60W tube and PSU with 60W components, or go back to 40W. I converted the machine to 60W to save time on the big cutting jobs, but scanning and engraving quality isn't as good. I assume that's normal, but since it appears the tube is bad I could be wrong. I need to decide and get parts ordered today; I have a customer who needs product.

--jim

[Tech_Marco]

60W is great for engraving. Even a 80W is still doing a very good job!
For 40W laser tube, it sometimes may not get you enough power for deep engraving when speed like 500mm/s or up

[audiomath]

60W is great for engraving. Even 80W is good!
40W sometimes may not get you deep engraving enough with huger speed like 600mm/s

That's interesting... I haven't ever tried running the machine faster than 200mm/s.
Ok, you changed my mind. I'll stay with the 60W components. You'll see an order in the next hour or so.

Thanks!
--jim

[Tech_Marco]

To me it is not acceptable for a machine not able to run at 400mm/s for engraving! Our machine require to run at least 550mm/s as a minimum standard before we allow it to leave our shop.

I used to do my engraving work for at least 550mm/s in most case. 600mm/s or 650mm.s is not uncommon here in our shop. For cutting, we run from 5mm~35mm/s. Occasionally, we ran 100mm/s for outlining.

[audiomath]

The machine seems to move fine at 400 mm/s. I can scan a 100x100mm rectangle at that speed and it doesn't appear to lose any steps.
(it returns to the correct origin when through) I'll have to wait until the tube and PSU are replaced to really tell for sure, though.

I suppose I never thought to run it any faster because of my experiences with CNC routers. 400mm/s would be impossibly fast on most routers, even smaller ones. Of course the laser engraver has much less mass to sling around, and doesn't have to push a tool through the workpiece. I hadn't really thought much about it before, though.

Thanks,
--jim