Electrical arc

[wraymogg]

Hello,

last week it happen a weird thing. I will describe it and please comment every aspect since I try to understand what happen. The system is an old Finnigan GCQ Plus, identical with Thermo Polaris (not PolarisQ)

After an injection with 5 ul hexane, injector at 280 celsius degrees, right before filament was about to go on, the system gave a communication problem. When I exited software and started it again to restart the communication a loud noise occurred and it began to smoke. First thing I thought is 'the diffusion pump locked', since is the only mechanical thing able to give that noise.

After opening it up I noticed the tip of the transfer line was chipped and the small piece was inside the vacuum chamber. Also the walls of the vacuum chamber were painted like an electrical arc just happen.

Also the motherboard had a fried resistor on it, driving a chip that converts the transfer line temperature sensor's resistivity into voltage. That, because later, with other chip will verify that voltage with a reference voltage to check if the sensor is OK. Discovering what got fried, I tested the transfer line. Fried as well. Both heating resistor and temperature sensor fried.

What happen ?

My idea is like this: The communication error happen exactly when the solvent was eluting. When this insane device gets into communication error for some reason will start to heat the transfer line and ion source like crazy instead of cooling it. This was tested before and is one of the stupid things Finnigan made.
I think the solvent auto-ignited and the little explosion lifted for a second the lid on the vacuum chamber - hence the noise and the chipped transfer line that holds in place the lid. The air that entered the chamber, permitted to form an electrical arc. That electrical arc went through the transfer line to the motherboard and burned the converter, plus it burned the transfer line as well.

The big question, what role the vespel tip has on the transfer line ? Can be the fact that it was chipped that permitted the electricity to go through the transfer line ? I measured it and the tip will NOT conduct the electrical current. It may be that the plasma resulted in the explosion to conduct the current beyond the tip to the metallic part of the transfer line.

The vacuum gauge indicated 40 mTorr, but I don't have an ion gauge to know exactly what was inside.

Thoughts ?

Regards,
Vlad

[dblux_]

... First thing I thought is 'the diffusion pump locked', since is the only mechanical thing able to give that noise. ...

But diffusion pump has no moving parts to get locked

I think the solvent auto-ignited and the little explosion lifted for a second the lid on the vacuum chamber - hence the noise and the chipped transfer line that holds in place the lid. The air that entered the chamber, permitted to form an electrical arc. That electrical arc went through the transfer line to the motherboard and burned the converter, plus it burned the transfer line as well.

If it was hexane, then I doubt it could ignite it vacuum chamber. Maybe it was a short circuit in resistive heating element ?

[wraymogg]

Hi,

true, I am used to the turbomoleculat pump on the other systems, this old system is the only one using diffusion pump and this doesn't have moving parts. Why I was thinking of it ? Because in 2 times, I had overtemperature in the pump. The smart guys from spectralab mounted both fans to push air inside, instead of tunneling it. I fixed that but it may damaged the peltier element.

So if you say is impossible for hexane to auto-ignite, I still miss the source of the small explosion inside. The electrical arc formed because of the air getting in when the lid was lifted.

Question: the electrical arc can form without the filament on ? I mean high voltage still exists inside from the pin connecting the ion trap. Actually when the system enters 'communication error' will shut off the voltages on the ion trap, lenses, etc. When I restarted the software, it powered them again on, and this is when the explosion happen. So after applying voltage to ion trap, it exploded.

Is it possible than the peltier element on diffusion pump to be defective and if is not cool enough, the vapors will not condese on baffle and they can escape to the vacuum chamber. If so, that enough can generate an electrical arc. I will test that peltier element today.

Question: the electrical arc alone can generate the explosion that lifted the vacuum chamber lid ?

Regards,
Vlad

[dblux_]

... So if you say is impossible for hexane to auto-ignite, I still miss the source of the small explosion inside...

I think so. There has to be enough oxygen for substance to ignite. Either in the atmosphere or in the substance itself.

As I don't have enough experience with such situations I don't want to speculate.
In my humble opinion faulty resistive heaters may explode due to short circuit high currents.