CO2 laser power supply fail.

[slowpoke]

Anyone here had to deal with high voltage power supplies driving CO2 gas lasers ?
I have one blow up on me. Just trying to figure out why it failed.

Full disclosure: i dont do repair, dont do electronics other than miniquad racing drones, minor repairs with obviously blown bits, all low voltage or low signaling (rc stuff). I dont touch anything beyond 4s lipo stuff. To be clear - I WILL NOT BE FIXING THIS HV PSU. Just trying to understand things.

As for this power supply i'm just trying to understand what went sideways to make it go pop. It is fairly fresh with only ~50 hours on idle and ~25 hours on firing laser. It is definitely not this PSUs MTBF, nowhere near. People run these for years. Model is ZYE MYJG100W, runs 80W CDWG co2 tube.

So, obvious things first. 1st stage flyback (~500Vdc) blown its side out, 2nd stage is fine (~30K Vdc). The conecting input wire to 1st stage flyback burned, completely disintegrating, taking what i assume to be current sensor wire loop with it. On boom tripped 16A rcd in the workshop and 16A breaker in the house. At the time of boom psu was pulling about 600W out of the wall (230V, UK). Aftermath is obviously blown flyback, burned wiring.

At this moment i'm just looking for explanation and general insight what went sideways.

[ace1903]

Are you sure that this has 1st stage flyback (~500Vdc) ?
To me looks like two secondaries are connected in parallel to obtain increased output current.
Current sharing between two transformer is done poorly and that is why one secondary blew(potentially weak spot was created during potting).
Can you share pictures of the bottom side of the PCB? 

[slowpoke]

Here's some more pics:

[Teledog]

eh,
Not sure if it's worth trying to repair a potted Xformer.  bit of a nightmare..
Any shorts/loose wiring that may have caused the smoke show? Power surge? Connected to a dedicated breaker or shared?
Might check fleabay for a replacement PSU.
Bought one for ~$100 a while back (variable current & all)

[ace1903]

Primaries are definitely connected parallel. Will change my guess and now I think that secondaries are connected in series.
Middle point between two transformers secondaries is grounded. Now output nearby single turn coil is on several tens kv with respect of the coil.
Think that single turn coil on other units is close to the core but on this particular unit due to crappy assembly one ends was close to high voltage output.
Anyway seems that isolation on HV wire was enough for that voltage only marginally.
Please take pictures of the HV wiring. If my theory is right seems those units are not safe for everyday use.
Or assembly shall be checked before installation.
 

[james_s]

At a glance it looks like a case of flashover, it may have started with a bad connection on the ground side of the HV winding but it's hard to say for sure. It has carbonized though and carbon is conductive so if the transformer has not failed you will need to scrape off all of the carbon or it will flash over again.

[jh15]

Interesting, last CO2 laser I worked with was water cooled and produced a whopping 5 watt beam.

[slowpoke]

Well, i suppose things moved on since then

My thing is 900x600 bed size, at the moment 80W tube, that actually outputs 80W out of the business end full tilt, can be overdriven to ~100W but kills it in a month doing so. People using tubes up to 180W at home these days.

Tube is water cooled, using just a bucket with room temp water, but ideally i should get a chiller for that...

Still inefficient though, ~70-80% power setting of max 80W out pulls 600W out of the wall (240V). Can cut up to 10mm acrylic, engrave stone/glass, cut/engrave wood, engrave anodized aluminium, etc. Useful. Until power supply blows up, that  is.

[james_s]

Still inefficient though, ~70-80% power setting of max 80W out pulls 600W out of the wall (240V). Can cut up to 10mm acrylic, engrave stone/glass, cut/engrave wood, engrave anodized aluminium, etc. Useful. Until power supply blows up, that  is.

That's incredibly efficient for a gas laser. The common red HeNe lasers will consume about 35W to put out 5mW of optical power. Ion lasers are even worse, a multiline argon laser will typically consume 1500W to put out 150mW optical.

[Per Hansson]

Still inefficient though, ~70-80% power setting of max 80W out pulls 600W out of the wall (240V). Can cut up to 10mm acrylic, engrave stone/glass, cut/engrave wood, engrave anodized aluminium, etc. Useful. Until power supply blows up, that  is.

That's incredibly efficient for a gas laser. The common red HeNe lasers will consume about 35W to put out 5mW of optical power. Ion lasers are even worse, a multiline argon laser will typically consume 1500W to put out 150mW optical.

Yes 7.5% is not shabby for a CO2 laser.
Even some of the best industrial solid state CO2 lasers (no turbine = no moving parts and no laser gas pumping during operation still only attain about 16% efficiency at best).
https://www.rofin.com/fileadmin/user_upload/COHR_DC_Series_DS_0420.pdf

[T3sl4co1l]

There's a cable coming out the power end, right..?

That contraction being "was".

Looks like whatever's coming out of the one transformer, it arced to chassis probably, then burned back for some time (less than a minute?), along the way hitting the red wire (some sort of sense winding?) which in turn arced to the chassis, hence the extra burn points on the board.

I would guess the two transformer design is to generate balanced AC (+/- around ground), which is better for a variety of reasons (less voltage with respect to ground for the same total drop across the tube; potentially lower EMI; potential simplification in design, though with two separate transformers it seems they didn't go for that in this case).  Neon sign transformers typically do the same thing, for the same reason.

Tim