Problem with MSD 5973

[James_Ball]

The EM is 10 years old, but its voltage hasn´t been increasing before problem (normally 1800 V).

I have had an EM before that ran at about 1200V for years then died completely between runs. I would change the EM and see if that fixes the problem.

They all seem to behave differently, some die over night at low voltage and some you increase a few volts at a time for several years until you are running it near 2700V and it keeps going but you have lost all linear range. 10 years is a long time on an EM that is certain.

[cleh]

Agree with James. Try a new EM.

[awoods]

Had the same problem last year (peak shapes and EM voltage high)
Your problem is with the electron multiplier, the EM voltage should be between 1000v and 3000v. Engineer told me when your MS is tuning with the EM voltage in excess of 2900v it is old and needs to be replaced.

[lmh]

Just in case anyone comes across this thread, I'll admit I've created an almost identical situation recently, but this was pure stupidity on my part, and was easily solved:

After a routine filament change and some column maintenance I accidentally replaced the blue and orange wires that go to lenses at the back of the ion source in such a way that the wires dropped a bit, and got very close to the cone on the source that lines up with the transfer line when you close the door. As a result when I shut the door, I trapped the wires, damaged them slightly, and quite probably messed up the gas flow around the source's inlet cone. I can't believe I did something soooo stupid. (I did at least get them the right way round!)

They symptoms were extreme difficulty getting it to autotune, a very ragged peak-shape (particularly for the 502 mass), enormous EM voltage and a gain factor that's beyond belief. Moving the two wires to a sensible place where they're not in the way, and where the damaged insulation wasn't pushed up against a metal surface reduced the EM voltage on tuning by more than 1000V and restored nice peak shape for all three tune masses.

Proviso: At the same time I had also changed the filaments - again - because in a 2nd piece of stupidity, the first routine filament change had been triggered by a user complaining both filaments weren't working, and when removed they were completely trashed, beyond a normal dead filament. Then the instrument wouldn't tune, but it was because the column had broken about 1cm in from the transfer line, meaning that vast amounts of air was being drawn into the source, giving a terrible vacuum (hence no tuning), which I only noticed after the first filament change, so I'd operated the filaments in bad vacuum for a bit before realising there was the problem with the wires too, which meant they were a bit blue-looking, and I wanted to bias everything in my favour before restarting the instrument. I am new to GC-MS, and making every mistake imaginable.

[James_Ball]

Don't feel bad, we have all made similar mistakes

On the 5971 there is an interface cone like the one on the 5973 except much bigger that screws into the source, and it is what conducted heat from the transfer line interface to the source before they used heaters on the source. Myself and the Agilent service rep were working on getting this thing working one evening and had it all put together and pumping down when we noticed the part still sitting on the bench. Had to cool the diffusion pump, open it up and put it in, not so much fun on those lol.

I have also seen the draw out plate on the wrong side of the spacer ring, not great for the sensitivity either.