Leaking Edwards E2M1.5 vacuum pump

[cwehinger]

Hi,

The vacuum pump for our Agilent 5973 Mass Spec is leaking oil. Can I safely use an older Edwards 2 stage E2M2 pump till the E2M1.5 pump is rebuilt.

Thank you in advance

Chuck

[MSCHemist]

Yes if you plug it into an outlet as the MS's power supply can't handle the power requirements of a more powerful pump.

[cwehinger]

Thank you

[James_Ball]

I actually have all of my 5973 and 5975 instruments running on the E2M2 pumps. They last longer(decades longer for us), run quieter, and run cooler and have no problems with foaming oil. If you run purge and trap, they also handle the higher moisture better. The pumps I am running on my instruments I took from our old 5995, 5970, 5971 and 5972 instruments that the new ones replaced and non have even been rebuilt yet, a couple are almost 25 years old. All of my small pumps died within three years.

[twranger]

New update to old post- have to agree with lifetime of the 1.5s, we have been putting Edwards RV3 on the instruments, quieter and longer lifetime.

We just had a Agilent (Varian) DS-42 fail on a 5977, its 6 years old so no complaints. Looks like Agilent is now selling either a DS-40M or their new IDP-3 scroll pump. Anyone had experience with these models?

[benhutcherson]

I think Agilent likes the quiet little E2M1.5 pumps, but it seems like they all start leaking oil and have other issues eventually. They run incredibly hot, and from what I've seen a get a higher hydrocarbon background in my 5975 than even my 5971(which uses a massively oversized Varian DS102).

The last time my E2M1.5 sprung a leak, I got away with just a replacement of the big gasket between the pump body and motor. In a few years, it will probably be full rebuild or replacement time(our guy at work who normally does that stuff looked at the tiny replacement vanes I had and said "no way-send it off"-we do have a pump shop in town fortunately).

My Varian 300-TQ has an Edwards RV3 on it, which I have zero complaints about. With the turbo(which helps a lot) backed by that pump, I see low 10^-7 torr or sometimes even high 10^-8(single quad EI mode-it hops up to high 10^-6 when I'm doing CI with CID). One independent service engineer I've worked with says he always keeps a new/rebuilt RV3 in his trunk for when he's going to sites. He says he won't replace like-for-like on a failed E2M1.5, the Pfeiffer equivalent, or the new scroll pump-he puts the RV3 on it and calls it a day.

The E2M2 is still a great pump. I'm a big fan of the Varian DS102, which Agilent still sells branded Agilent, also. That was the pump Varian would ship with everything from their budget ion trap models(which generally don't need as hard of a vacuum as quads) and even up to some GC-TQs(the LCs and convertibles on the same basic design used either two pumps or sometimes one really big one). It's a step up from the E2M2, and IIRC without looking now in some specs is better than the RV3 and in some is worse.

As an aside, the 5971/72 are capable of Ci(with a proper source/.gas manifold) and they were only available with diffusion pumps. The 5973 and later require a "performance turbo" for CI use. I posted the question as to why to-again-a former HP/Agilent CE/FSE who is now independent, and he gave me an answer that makes sense even though I probably wouldn't have thought of it. The E2M2 backing a diffusion pump is able to handle the extra flow from CI just fine, while the E2M1.5 is just not up to the task. That's also why Agilent won't pair a diffusion pump with a scroll pump-the pair can't get a good enough vacuum even for EI.

BTW, my 5871 that's now a 5972 reliably pullls high 10^-6 with analytical flows through a .18mm column when backed by the DS102. I have everything I need to do CI on it(source, manifold, plumbing to connect the two together) but even then it seems a bit jerry-rigged and CI is so much easier and more versatile on my Varian 300. Even so, I think it could easily handle doing CI.

[James_Ball]

I think Agilent likes the quiet little E2M1.5 pumps, but it seems like they all start leaking oil and have other issues eventually. They run incredibly hot, and from what I've seen a get a higher hydrocarbon background in my 5975 than even my 5971(which uses a massively oversized Varian DS102).

The last time my E2M1.5 sprung a leak, I got away with just a replacement of the big gasket between the pump body and motor. In a few years, it will probably be full rebuild or replacement time(our guy at work who normally does that stuff looked at the tiny replacement vanes I had and said "no way-send it off"-we do have a pump shop in town fortunately).

My Varian 300-TQ has an Edwards RV3 on it, which I have zero complaints about. With the turbo(which helps a lot) backed by that pump, I see low 10^-7 torr or sometimes even high 10^-8(single quad EI mode-it hops up to high 10^-6 when I'm doing CI with CID). One independent service engineer I've worked with says he always keeps a new/rebuilt RV3 in his trunk for when he's going to sites. He says he won't replace like-for-like on a failed E2M1.5, the Pfeiffer equivalent, or the new scroll pump-he puts the RV3 on it and calls it a day.

The E2M2 is still a great pump. I'm a big fan of the Varian DS102, which Agilent still sells branded Agilent, also. That was the pump Varian would ship with everything from their budget ion trap models(which generally don't need as hard of a vacuum as quads) and even up to some GC-TQs(the LCs and convertibles on the same basic design used either two pumps or sometimes one really big one). It's a step up from the E2M2, and IIRC without looking now in some specs is better than the RV3 and in some is worse.

As an aside, the 5971/72 are capable of Ci(with a proper source/.gas manifold) and they were only available with diffusion pumps. The 5973 and later require a "performance turbo" for CI use. I posted the question as to why to-again-a former HP/Agilent CE/FSE who is now independent, and he gave me an answer that makes sense even though I probably wouldn't have thought of it. The E2M2 backing a diffusion pump is able to handle the extra flow from CI just fine, while the E2M1.5 is just not up to the task. That's also why Agilent won't pair a diffusion pump with a scroll pump-the pair can't get a good enough vacuum even for EI.

BTW, my 5871 that's now a 5972 reliably pullls high 10^-6 with analytical flows through a .18mm column when backed by the DS102. I have everything I need to do CI on it(source, manifold, plumbing to connect the two together) but even then it seems a bit jerry-rigged and CI is so much easier and more versatile on my Varian 300. Even so, I think it could easily handle doing CI.

The main reason for the E2M1.5 was the smaller form factor so they would slide up under the MS and save bench space. I would rather deal with the larger pump sitting beside or below the instrument just to get the extra pumping capacity and lifetime.

I have the RV5 on my 7000C QQQ and it started leaking after a couple years. I still have it running just it does its own oil changes every few months. There is a larger scroll pump for those, but I can't justify the cost.

[benhutcherson]

The main reason for the E2M1.5 was the smaller form factor so they would slide up under the MS and save bench space. I would rather deal with the larger pump sitting beside or below the instrument just to get the extra pumping capacity and lifetime.

I have the RV5 on my 7000C QQQ and it started leaking after a couple years. I still have it running just it does its own oil changes every few months. There is a larger scroll pump for those, but I can't justify the cost.

Agreed with the general sentiment of using an uprated pump-as you said both for longevity, and in my experience better vacuum. At one time, when I was waiting on the RV3 to arrive for the Varian 300, I repurposed a MASSIVE 220V Edwards(either an older 8 or 12 series) that came with a now idle(and maybe never to run again) Finnigan LCQ Duo LC-MS. The set-up was a bit awkward and it was noisy, but that pump would pull it down like crazy and get me to 10^-8 repeatably.

Even back in the early 90s when the 5971 as an affordable benchtop mass spec was coming into common use, SIS published a couple of articles about upgrades they'd done to both improve sensitivity and reliability. At the time, they reported a 10 fold increase. HP incorporated some of their suggested changes, but never did some of the others(straight-through transfer line tip, for example). One of their big ones was they felt even the E2M2 was inadequate. They recommend an Alcatel UM2005 at 100L/min, which is roughly equivalent to the RV3. They then go on to say that they like to remotely locate their pumps to reduce noise, and specify 300L/min, or about an RV8 size pump.

https://www.sisweb.com/referenc/articles/sensit-1.htm

My favored Varian DS102 is 3.5cfm, which is in the same 100L/min range-it's not as good as an RV3, but still a big step up from an E2M2.

[Consumer Products Guy]

We had the very first 5971 system in our state. Our CEO strangely demanded that EVERY item ordered that year HAD TO arrive on site by Dec. 31, I actually asked Hewlett-Packard/Agilent to ship out an empty container, but they assured that it would arrive exactly on Dec. 31, and it did.

It ran on Windows 286 or 386 (can't remember) and I'd never seen a computer mouse before. That Windows version did not even have File Manager, has MS-DOS Executive.

The original vacuum pump for that unit was upgraded for free by the HP service engineer a few months later, and he gave me the old pump to take home to use on vehicle AC applications. Pump still works.

[James_Ball]

We had the very first 5971 system in our state. Our CEO strangely demanded that EVERY item ordered that year HAD TO arrive on site by Dec. 31, I actually asked Hewlett-Packard/Agilent to ship out an empty container, but they assured that it would arrive exactly on Dec. 31, and it did.

It ran on Windows 286 or 386 (can't remember) and I'd never seen a computer mouse before. That Windows version did not even have File Manager, has MS-DOS Executive.

The original vacuum pump for that unit was upgraded for free by the HP service engineer a few months later, and he gave me the old pump to take home to use on vehicle AC applications. Pump still works.

Sounds like you had a similar experience to me with the 5971. I learned both Windows 3.1 and MSDChemstation while working the night shift because everyone else was afraid of it. Ours had originally been slated to run through RTE-A using the PC as a scanbox type of setup. Gladly that was dropped after I figured out how to run it through Windows.

[Consumer Products Guy]

If I remember correctly, the cost was about $50K without an autosampler, so we decided we'd get the autosampler the following year. We had $30K capital remaining, got two other departments to throw in $10K each, and got it done. A couple of months later we added the autosampler.

There was a built in tutorial, how learned. The first time we needed to change a filament, we wheeled a VCR/TV from the audio visual department and did that step by step.

We later updated to a Windows 3, got a 486 computer. When the company later got computerized, we had a little head start because we were familiar with Windows.

The big wooden box for the 5971A with all the non-tip indicators became my dog's doghouse for a few years.

[twranger]

New update to old post- have to agree with lifetime of the 1.5s, we have been putting Edwards RV3 on the instruments, quieter and longer lifetime.

We just had a Agilent (Varian) DS-42 fail on a 5977, its 6 years old so no complaints. Looks like Agilent is now selling either a DS-40M or their new IDP-3 scroll pump. Anyone had experience with these models?

Back to topic, I've also heard you can't put an RV3 on a diffusion pump mass spec. Any truth to that?

[James_Ball]

New update to old post- have to agree with lifetime of the 1.5s, we have been putting Edwards RV3 on the instruments, quieter and longer lifetime.

We just had a Agilent (Varian) DS-42 fail on a 5977, its 6 years old so no complaints. Looks like Agilent is now selling either a DS-40M or their new IDP-3 scroll pump. Anyone had experience with these models?

Back to topic, I've also heard you can't put an RV3 on a diffusion pump mass spec. Any truth to that?

We run a 5973 with diffusion pump and use the E2M2 which is close to the RV3. All the 5971/72s came with the same diffusion pump and the E2M2. I can't imagine how it would cause problems. Has anyone else used that combination?

[LALman]

Agreed with the general sentiment of using an uprated pump-as you said both for longevity, and in my experience better vacuum. At one time, when I was waiting on the RV3 to arrive for the Varian 300, I repurposed a MASSIVE 220V Edwards(either an older 8 or 12 series) that came with a now idle(and maybe never to run again) Finnigan LCQ Duo LC-MS. The set-up was a bit awkward and it was noisy, but that pump would pull it down like crazy and get me to 10^-8 repeatably.

My E2M2 gives me the following vacuums at the pump inlet on a 5971 (diffusion pump turned off) at various helium carrier flows (20m x 180um x 1mm DBVRX column.
Flow mL/min Vacuum mTorr
0.20 6.6
0.30 8.2
0.40 9.9
0.50 11.6
0.60 13.2
0.70 14.8
0.80 16.6
0.90 18.3
1.00 20.1

With the diffusion pump running and a carrier flow of 0.6mL/min I got 3.3xE-5 mTorr in the 5971. Is that 10^-8 vacuum in Torr instead of mTorr?

I've rebuilt this E2M2 several times. it will pull down to about 5-6 mTorr with inlet closed but performance under flow is more important. Should I conclude my E2M2 is still working well? Should I get a RV3 do you think?

[benhutcherson]

Agreed with the general sentiment of using an uprated pump-as you said both for longevity, and in my experience better vacuum. At one time, when I was waiting on the RV3 to arrive for the Varian 300, I repurposed a MASSIVE 220V Edwards(either an older 8 or 12 series) that came with a now idle(and maybe never to run again) Finnigan LCQ Duo LC-MS. The set-up was a bit awkward and it was noisy, but that pump would pull it down like crazy and get me to 10^-8 repeatably.

My E2M2 gives me the following vacuums at the pump inlet on a 5971 (diffusion pump turned off) at various helium carrier flows (20m x 180um x 1mm DBVRX column.
Flow mL/min Vacuum mTorr
0.20 6.6
0.30 8.2
0.40 9.9
0.50 11.6
0.60 13.2
0.70 14.8
0.80 16.6
0.90 18.3
1.00 20.1

With the diffusion pump running and a carrier flow of 0.6mL/min I got 3.3 E-5 mTorr in the 5971.

Should I conclude my E2M2 is worn out? Should I get a RV3 do you think?

That honestly doesn't look terrible to me. I don't have a 5971 w/E2M2 handy(Between my work one and the ones I service, I only have one on a since-expired contract running an E2M2 and I don't have notes handy on what it usually runs). With the DS102, I can pull down to 5-6mTorr at my resting flow of .5mL/min and usually run ~16mTorr at a typical 1mL/min analytical flow.

My 5975 sitting next to it with diff pump and an E2M1.5 runs about 25mTorr at .5mL and in the 40mTorr range at 1mL/min. Needless to say, I really, really think Agilent underspecs foreline pumps now.

And yes, I'd have zero reservations running a bigger than spec pump like an RV3. In fact, one independent service man I know/have worked with fits an RV3 to any single quad HP/Agilent system he works on at the first sign of trouble with the included pump. He's enough of an advocate of that pump that he always keeps a new/rebuilt one in his service truck so that he's ready to swap on-site if needed.

It's probably best not to go nuts with it, but my Varian TQ specs a DS102 and I run an RV3 on it. It didn't come with a pump and I used a few others to get it going initially, and I honestly thought the DS102 let pressures get a little higher than I'd like with the collision cell going, and even worse when doing CI. It would run up into the high 10^-5 range doing both, and I'd see a definitely increase in noise/lower LODs over what the instrument said it should do(that's an awfully long distance to travel at that pressure). With the RV3, it will pull down into the high 10^-8 range at resting flows, and will stay at high 10^-6 with collision cell+CI(and will hold at high 10^-7 or low 10^-6 with one or the other, depending on just how I have it set up).