Page 1 of 1

Agilent 5973N Frustrating Leak

Posted: Tue Apr 17, 2018 4:29 pm
by Akthmps88
We recently acquired an Agilent 5973N MSD. I have made several attempts to operate the MSD, none of which have been successful. Here is a summary of the situation.

The MSD turbo pumps down to 100% within ~ 5 minutes. I have the transfer line capped with a blank ferrule for now. I received the MSD with a Pfeiffer foreline pump, and it made such loud noises that I connected an Edwards foreline pump instead. It runs fairly quiet. In any event, when I go to scan, it reports failure of the HED supply, and other times it simply reports "no peaks found" when I go to the manual tune tab. I have a 59864B gauge. Even though the turbo spins down to 100%, the gauge only reads around 4 x 10 -2 Torr. It should obviously be in the range of -5 or -6 Torr. It seems that there is a tremendous vacuum leak somewhere. I found it very difficult to believe that the manifold pressure could only be at -2 Torr when I don't have evidence of gurgling or leaking from the Edwards foreline pump (but I do hear gurgling from the Pfeiffer pump). I have another 59864B gauge, so I swapped it out in the event that my gauge was not reading correctly, and the second gauge also reads 4 x 10 -2 Torr. So, at least based on the gauge, the manifold is nowhere near where it needs to be. I swapped the HED supply yesterday, and the situation is still the same. It won't scan and reports HED supply error (most likely due to the high pressure).

Has anyone ever seen a 5973 MSD pump down to 100% turbo and only read 4 x 10 -2 Torr? I am surprised. I figured if the leak was this big, I would hear it, and/or the turbo would never be able to achieve 100% pump down.

In regards to finding the leak, I have removed the large o-ring around the manifold door several times now. No dust or lint is observed on it. I have worked with these MSDs before and have never had issues with pump down. I also cleaned the o-rings inside the front plate where the vent valve and PFTBA valve are located. Everything looks fine. I also completely removed the transfer line and inspected the o-ring that seals it. Looks good. I have also removed the clamps around the vacuum hose on both the turbo and foreline ends. All is secure.

The only places I haven't interrogated much are the ion gauge at the rear of the MSD. Feels very secure and I couldn't get the fitting to budge. The turbo looks to be seated fine underneath the MSD.

We did purchase this MSD used from a vendor we've used many times before with success. They pumped it down and tuned it just fine about 2 weeks ago.

If anyone has any suggestions, I'd love to hear them. Many thanks!

-Aaron

Re: Agilent 5973N Frustrating Leak

Posted: Tue Apr 17, 2018 7:38 pm
by high desert chemist
When you say you swapped out the 59864B gauge, did you swap just the controller, or did you also swap the triode assembly that's in the manifold. High pressure on a triode means that the current is weak due to electrons from the filament being capture, scattered, etc from gas in the manifold. If something's wrong with the triode assembly, your current could be low, which would manifest as an artificially high pressure. Maybe check the manual on the 59864B for guidance. That seems like a high pressure for the turbo to be able to hit 100% speed. Once you've ascertained whether there's a leak or not, then maybe the HED problem will be easier to solve; right now, they may or may not be related.

Re: Agilent 5973N Frustrating Leak

Posted: Tue Apr 17, 2018 8:28 pm
by Akthmps88
Thank you for the quick response. In regards to the ion gauge, I just replaced the controller box itself. I had to use the same blue connection cable to the pirani gauge/bulb. I have not replaced the pirani gauge/bulb itself.

The vendor I purchased the 5973 from is sending me another transfer line assembly to try, as well as a new o-ring for the analyzer manifold. I will know more tomorrow and report my results.

One last observation I noticed today... When the system is at full vacuum, the ion gauge reading stabilizes right around 4 x 10 -2 Torr. However, when I vent the MS, even when the turbo has gone down to around 50% speed, the vacuum reading remains fairly consistent at 4 x 10 -2 Torr. Then, when I shut off the foreline pump, turn off the MSD and open the vent valve, the reading on the controller actually goes down 10 -3 Torr range and holds there for about 15-20 seconds before it goes back up to -2 Torr range, and then eventually turns off completely. Seems rather odd?

Re: Agilent 5973N Frustrating Leak

Posted: Tue Apr 17, 2018 8:32 pm
by LALman
I have that system. I was able to repair my HP/AGILENT HP59864B Micro-ion Ionization Gauge by having a friend replace the caps mentioned in this link, on my failed unit. Maybe try out the fake ion-tube test rig he suggests for testing the gage controller and see if there is a problem with the controller.

You can get a new triode from Scientific Instrument Service for ~$200.

If the vacuum is actually OK, then maybe there is a problem with the 5973 mainboard and/or high voltage section.

Re: Agilent 5973N Frustrating Leak

Posted: Tue Apr 17, 2018 8:44 pm
by high desert chemist
Akthmps88 wrote:
One last observation I noticed today... When the system is at full vacuum, the ion gauge reading stabilizes right around 4 x 10 -2 Torr. However, when I vent the MS, even when the turbo has gone down to around 50% speed, the vacuum reading remains fairly consistent at 4 x 10 -2 Torr. Then, when I shut off the foreline pump, turn off the MSD and open the vent valve, the reading on the controller actually goes down 10 -3 Torr range and holds there for about 15-20 seconds before it goes back up to -2 Torr range, and then eventually turns off completely. Seems rather odd?
That is strange behavior. Sounds like triode assembly may be faulty, or the connection between the triode and the gauge may be suspect. You may try cleaning the contacts on the atmospheric side of the triode assembly and using the blue cord from the other gauge controller to see if it resolves the error. LALman's suggestion about a test stand to rule out triode gauge malfunctions may also be a good route to go.

Re: Agilent 5973N Frustrating Leak

Posted: Wed Apr 18, 2018 6:21 pm
by James_Ball
One place these can leak that is difficult to find in in the transfer line itself. If you remove the heater assembly you find that there is a smaller tube welded inside a larger tube and a leak can occur at that weld. We had one that began leaking there and I chased the leak for months. I was still getting enough vacuum to scan and when I sprayed some Dustoff into the heater area by putting the tube just past the nut and into the insulation I saw a response. Maybe the new transfer line assembly will solve your leak problem.

Re: Agilent 5973N Frustrating Leak

Posted: Wed Apr 18, 2018 7:47 pm
by Akthmps88
Thank you to all who responded! The 5973 is now fixed. I vented and removed the ion gauge bulb from the rear of the 5973. I noticed initially it was very blackened, but then I also observed a small crack in the gauge. I do not believe this happened upon removing the gauge. I installed a new one, and a couple of minutes I was already at -5 Torr. I scanned a few minutes later and the system now looks as expected.

So, it appears the leak was indeed at the ion gauge bulb itself. Perhaps sometime happened to it in transit.

Thanks,
Aaron