POWER SUPPLY FIRE

Verify that fuse was as specified.

...Jim Thompson

ugh

=A0the

=A0 =A0 ...Jim Thompson

=A0 | =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

Fuse holder lid and fuse itself missing, Quote "lost" end quote during firefighting.

Missing/molten AL starts 1/16" above the fuseholder. The fuseholder was slightly carbonized on the outside but the bakelite held structural integrity and the wires were still attached. The whole rear end of the panel is missing all insulation, its burnt up, just bare copper remains. Line cord was coiled, and still plugged in. What is left of the line cord is copper surrounded by carbon junk, no insulation left at all, all the way to the plug.

Inside the box The burned hot and neutral still ran to front panel switch, where they were both switched. Returned neutral neutral and hot from switch intact. Fuse on the switcher board is intact and shows continuity.

He has pictures of the unit pre shipment, and the fuseholder terminals are down at the bottom of the holder where they should be, melting was above the fuseholder.

Front panel is intact as are the potted multipliers. Everything 1" or deeper toward the front is just carbonized or gently melted. The back of the the unit was 20 cm from a wall and the burn pattern on the wall is the paint burned off the cinderblock down to grey block. Looks like a rocket or torch burned the paint off the wall.

This just doesn't make sense. We thought we would see evidence of arcing. We do not. No spalling, no pitting, no copper balls. The edge of the missing AL is melted and is a uniform melt. It is if the top of the rear panel just vanished. Yet no charring on the top lid on the inside. I watched the fire inspector look for the parts and a chunk of AL, they are not there.

There were other flamable materials, other possible sources of ignition on the bench.

Can any body suggest a book on forensics of electrical fires?

Steve

coiled cable

_______________ outlet________ (back wall)

Above view.

Steve

osr @uakr>Fuse holder lid and fuse itself missing, Quote "lost" end quote

This reminds me of the manufacturers who have a fuse holder exposed to the world and who also have a secondary device inside the unit for the the tinfoil-inserting types. . . "Attribution" (aka leaving in your post the name of the person to whom you are replying) is standard practice on Usenet.

The lack of a GFI trip seems to rule out a lot of possibilities such as the cord being pinched by the housing cover. Was the ground pin of the GFI protected receptacle properly wired? Were neutral and hot reversed in the receptacle? What about the UL rated plug?. If receptacle not wired properly could shift some of the blame to building owner. Strange that a fuse and fuse-holder would be "lost". Makes one wonder and have "doubt". Was the fire contained to the area described? Any evidence of chemicals or powders in the bottom of or beneath the power supply?

Did any major parts in the PS burn up like electrolytic capacitors or bridge rectifiers?

I have seen thin aluminum sheet oxidize and "dust away" before my eyes in corrosive conditions, maybe a capacitor vented from heat and worked on the panel?

What does the fire inspector think happened? Any surveillance videos?

They're screwed, if you can find a competent lawyer. They obviously blew the fuse, and, not having an exact replacement on hand, stuffed a nail or somthing in its place. They took the nail out "during firefighting" to try to cover their tracks.

Good Luck! Rich

Without having any fire forensic experience, my reading of your description sounds like someone took to the box with a high-tem flame source (eg oxy-acetylene torch) to create a culprit.

The 30kV at 1mA is not likely to, but certainly can ignite nearby flammable media. The supply should be able to handle a dead short on the

That switcher should be UL approved, since it is AC fed. If it is, it MUST be able to handle a dead short condition on its output.

That said, it is very unlikely that it was the cause, BUT it is possible if the engineer, and the assembler did not examine the AC power entry to the unit carefully. Is there an 'entry module' or discreet line filter arrangement incorporated?

That would depend on the ampacity of the branch you hooked into, and the function of the breaker that protects the branch. If those were high or faulty, a short on the AC side inside the box could cause some serious high current condition damage. If it was being fed by a properly protected line, it should have snapped the protection breaker long before any such great damage could possibly occur. None of the tail end elements of the circuit can draw that much. If they fed it with an AC feed that had no protection, then THEY were at fault for running a test using an AC fed device, without assuring that said AC feed had proper current overload protection incorporated into it.

It still sounds like the feed circuit protection failed... miserably.

A single event can do that. Even though the breaker opened, it seems that it opened at a higher fault current than it is rated at. I would fault test that branch and breaker again to be sure that it did not fail too damned late.

No telling if any additional elements were added to the scenario. Especially if they are deliberately attempting to make it appear as if your client is culpable.

Then, the question remains... does the supply still operate? If so, you can be fairly sure that it was not the 30kV output that chewed the case away.

The unit failed at the entry point. The remaining supply segments were not part of the 'failure' circuit at all. The AC entry either has a design flaw, or that assembly was performed incorrectly.

ad

Short circuit protected/arc sensed on the HV end, and the litz in the coil would melt if that failed.

it

UL approved switcher.

ne

No entry module, and a QC picture taken at assembly matches the drawings , and the remains match the picture.

gh

=A0None of the tail end

Wiring was conduit from the panel to a Ul approved lab grade outlet strip built along the wall, wiring for 4 feet on either side of the outlet in use was burnt/destroyed.

=A0the

ly.

His unit was double fused, one on the back panel , one on the PSU board. PSU board intact, main fuse missing in action, Also much heat from later fire, as flamable materials were along the benchtop. Lab bench was unfused feed off the 20A breaker, had GFIs, no additional protection.

ms

Breaker being examined by 3rd party EE hired by the insurance. I'd bet the only time the breaker ever cycled since the 70s was when they installed the GFIs. 3rd party EE took the wall wiring and GFIs, too.

you

Control board and Ul Approved PSU are uniformly chared across their surfaces, but I would bet the multipliers would run if new boards were plugged in. We're talking two 30 watt HV modules systems here, and one was off by default as a DPDT switch selected a positive module or a negative module on the low voltage (15V) side. HV load side was fliped by moving a connector. The litz wire would fail in the coil long before it could source that kind of power, its only a 600 uA source, short circuit, at full power.

Steve

That transformer should be able to take such shorting WITHOUT heating the winding to the point that the wire enamel would melt. If all you are pushing is 1mA, your HV multiplier caps should be significantly small enough in value such that you do not store any more energy than the output requirement specifies.

What oscillation frequency was chosen (I assume you used a pot core transformer)?

(snip)

Theory: The outlet in use by your device had the conductors "back stabbed" into it instead of screwed down. An outlet downstream of your device (and the breaker) was used by some other load pulling high level of current. The backstabbed outlet overheated and a local fire and arcing ensued near your power supply and involved chemicals or other materials nearby. Damage to your device was secondary to the cause which was a failed outlet. I don't know if lab grade outlests can be backstabbed, but a loose screw could result in same overheating.

For a 20 Amp fault current breaker, it sure sounds as if more than 20 Amps was sustained for more than a few seconds. I would point a finger directly at that breaker, and the Fire Marshall (assuming there was an emergency response) should have noted such overcurrent evidence as being VERY abnormal as well.

To put it as succinctly and frankly as possible, I would say that without question, that 20 Amp breaker did NOT open when that fault current was presented to it.

An absolute failure mode. Hopefully, it is not a high quality US manufactured part. Alas, even some great, old US makers are shopping the labor out to nations that cost them less to get the product, despite the fact that failure numbers also go up.

Wild ass theory:

If they left the 30KV supply on for an extended time, would static electricity draw these metal powders into the power supply? I am thinking "thermite" here.

More likely spilled by a operator. 7.5 years of working with grad students (Who have BS degrees in Chem Eng) in a similar lab suggests that similar PSUs are actually viewed as cupholders, paperweights, and as useful chemical storage areas.

Yeah, the dust can move, and it would do something else nasty if fine enough.

Steve

Coworker was working on a centrifuge reworking a board. The board was exploding little pops. I forget the name of the chemical, but was highly explosive. Salt soloution and labs, will cause more problems.

greg

Was this powered up unattended?

How far away were people in the lab when this broke out?

unattended.

Steve