Where to turn? Bringing a 5890/5970 GC/MS back on line

[Don Baker]

Greetings -

I have acquired a HP/Agilent 5890/5970 GC/MS system. It is claimed to have been working when taken out of service. I believe that all essential pieces (short of carrier gas and expendables), including vacuum components, are present.

Despite having the manuals, I am not clear on the start-up procedure. The manual starts off as if all equipment has already been hooked up. I need help in getting the GC column connected to the MS and that sort of thing. From there the manuals look as though they'll be OK and I have someone near by who has experience in using the instrument.

The unit has been off-line for more than two years, so I already know that I'll need a new electron multiplier and I'm prepared for a couple of other non-trivial expenses as well. The unit has been at atmospheric pressure, but has been kept in a temperature controlled, fairly clean place.

Can someone help me out or point me in a direction to get going?

Thanks,

Don Baker

[Peter Apps]

Hi Don

Have you had a look on the Agilent web site - they might have downloadable instructions ?

If I were you I would initially plug the transfer line to the MS and then start it up to check that it pumps down properly and can tune. Once the MS is sorted you can connect the GC and trouble shoot the two together.

Peter

[rp18]

There's a service manual online:
http://ipes.us/used/58901.pdf

[Ron]

That is a really old system. If there are parts needed they will probably not be too easy to come by. If the system has been sitting for a couple of years there is a good chance that there will be issues with the high vacuum pump.

Good luck, but remember that the system is over 20 years old, and you will not have the sensitivity, mass range, scan speed, and user friendly software of a newer system.

[rp18]

Not always...
Our 5890/5971 easily beats our brand new Varian IT GC-MS in sensitivity.
The old HP's are nearly indestructible work horses. The diffusion pump (I assume there is one) will work even in 100 years.
Sure, the software needs Win95 or lower and it's not that fast, but for most uses it will do great.

[Don Baker]

Peter - I was able to get some very useful documents from Agilent's site, but not the low level sort of stuff I need at this point. Unfortunately, what I need to know is what "everyone knows".

rp18 - Thanks. That was helpful. I have the user's manual, but I didn't have a service manual.

Ron - Yes, it's old, but I'm not doing cutting edge research either. Because the 5890/5970 was a popular combination, parts seem to have decent availability. Agilent still supports it.

High vacuum in this system is provided by a turbo pump - no diffusion pump. Yeah, I'm concerned about bearings and that sort of thing. If the turbo pump goes I may be looking to see how I can flange up a diffusion pump or else scouring LabX for a replacement GCMS.

Thanks for your replies. I'd be happy to collect a few more as well.

Don Baker

[Suresh Seethapathy]

Don,

I just volunteered to get a 5890-5970 instrument up and running (if possible) in our undergrad lab. I am pretty much in the same state as you are. I am going to open the MS and check if the filaments are fine and clean the source as much as possible to start with. After that, I am looking at doing a leak check without turning the heater elctronics on, but with the pumps on. The worst expenditure i am looking at as of now is the Electron Multiplier. I have no clue to how old it is.

good luck to us
Suresh.

[JI2002]

I have been using a 5890/5970 system for almost 4 years, I'm always amazed at how well it works. It needs periodic maintenance like ion source cleaning and filament replacement, but that's about all it needs. To me, the HP engineers who made the 5890/5970 system more than 25 years ago were far better than the engineers making the 7980/5975 today in terms of the software as well as the hardware.

[Don Baker]

Suresh -

Please let me know how things go as you progress.

One of the things perplexing me at this point is that the 5970 instructions say to fire up the MS and test it prior to hooking it to the GC. But if the GC is not connected, then when I start up the roughing pump, all I'm going to do is suck air through the vacuum system. It's not possible to do the test as they say, unless I have simply missed reading something (entirely possible). I just don't see where it says "block the inlet to the MS using a piece of chewing gum" or something to that effect.

[Don Baker]

Suresh -

I failed to mention that you may as well plan for a new EM. They have a lifetime of only two years, or at least that's what the lifetime is supposed to be. I have no idea how long you can stretch them.

There's a quick kilobuck out of your budget right there. I'm delaying my EM purchase until I'm sure I have everything ready to go.

Regarding the availability of parts - This machine was apparently popular enough that third party suppliers for it have sprung up. Here is one for example:
http://www.5890.com/servlet/StoreFront

[Peter Apps]

Don, you are on the right track, just don't use chewing gum. Plug the transfer line to the MS using the nut that holds the column, and either a ferrule with no hole through it, or a short piece of wire the same diameter as the column held in place with an ordinary ferrule.

Peter

[carl.nott]

Ah, it's been a while. Hoping you're not looking to run it with RTE, hehe.

On sealing the system for pump down I would suggest as others using a nut and no-hole ferrule or using a bit of column and capping the end of the column with a septa. You can go ahead and install a full column and cap the end as well.

If I remember correctly (and I may be mistaken) the 5970 does not have a fail-safe check for vacuum level when you turn on the turbo, so be sure you've got a decent-ish seal and the rough pump running before you hit that switch or you may be in for some excitement.

I'm not sure what you'll be tuning for but be sure to close the tune valve (it's manual) when pumping down.

[Bigbear]

Alpha Omega Tech out of Brele NJ is one good source for parts and information. 800-842-5742

[Don Baker]

Well the computer appears to be fun ctional, though the tape drive's capstan has decomposed into a ball of goo. Apparently that is a common situation for a 9144 tape drive. Fortunately, I found a cheap and easy fix. Cross your fingers that it works.

"Hoping you're not looking to run it with RTE"
No, but I _am_ giving it a go with HP-UX. The software upgrades are pretty expensive and I don't want to invest in that until I'm fairly sure I have a working instrument.

"On sealing the system for pump down I would suggest as others using a nut and no-hole ferrule or using a bit of column and capping the end of the column with a septa. You can go ahead and install a full column and cap the end as well."
For some reason, the manual specifically instructs not to connect the GC prior to pumping down the MS (for testing).

"I'm not sure what you'll be tuning for but be sure to close the tune valve (it's manual) when pumping down."
Hate to be dumb, but I'm not sure what you're referring to here.

I should be able to do the pump down test this weekend.

[carl.nott]

For some reason, the manual specifically instructs not to connect the GC prior to pumping down the MS (for testing).

Huh. Must be concern about having the GC heating up the MS interface when you're not expecting it.

Hate to be dumb, but I'm not sure what you're referring to here.

It's only dumb if you don't understand and don't say anything. I should have been more clear anyway. The 5970 uses a standard calibration fluid (PFTBA) which is introduced into the MS for tuning. Unlike later models the 5970 uses a manual control for the opening of the valve which contains the PFTBA. Basically a gas ball valve. It's a common mistake for 5970 operators to leave that valve open at 'bad' times, leading to hilarious (wait... hilarious? I meant painful...) situations like 'I tuned my instrument, loaded a calibration, went home, next day realized I left the valve on and my calibration was garbage'. So check to make sure it's closed when you're doing pump down to avoid weird peaks and to avoid potential leaks.