Hackers and tinkerers of HP5890II instruments help needed

[jryter]

I am at a small university. We have an HP 5890Series II GC with 5970B MS, Chemstation and 2 autosamplers. The instrument is in good condition but the MS has failed (obsolete and very old so not worth fixing for us). Surprisingly we were able to purchase a new GC-MS after 20yrs. So the old instrument needs to be either
1. Sold- It doesn't work completely and not worth a lot. Parted out. Shipping uggh!
2. Hacked. I was thinking of letting students add different detectors like EC or TC or FID to the GC. This would need to be "easy" as they are not experts in anything and cheap as we are a state school. The old MS parts are great for show and tell for classes. I do have some old TC detectors salvaged from an old GC.
Has anyone ever changed detectors in the the 5890, would the new detectors be able to talk to chemstation and which would you recommend for teaching labs-for versatility and projects they could do. We do have some simple GOW MACs with a TC detectors.
3. We could recycle it. last option/

Ideas of fun projects? Want to buy it-pick up in the Colorado mts!.

[James_Ball]

The 5970 would definitely make a good display for teaching mass spec and being able to show what a quadrupole and electron multiplier look like.

What is the problem with the 5970? Bad turbo pump or fried electronics?

I have been helping someone through email to bring back his old 5970 from the dead, has been an interesting project trying to remember all the things I did with one 20 years ago.

As for the 5890, changing out detectors isn't too bad, but it does require removing the side and rear panels to access the plumbing and electronic connections. The most difficult to connect are probably the thermocouple wires it uses to monitor and control the temps, that is a small, finicky connector with some unique square pins to fool with, but it is possible. The detector boards just slide into slots on the side and usually have a probe that extends into the actual detector for the connection of control board to the detector. Rugged and easy to connect. Gas plumbing routes to the opposide side. They attach to the gas controls with an oring and a split plastic threaded connector, which are pretty much universal among the different detector types. The controller have normal 1/8" Swagelok connectors for the incoming gas lines.

The electronics should autodetect the new detector from the chips on the control board and load the appropriate setting and menus into the front control panel, and should be easy to setup in MSDChemstation too.

If you run into any major problems just yell, I think I still have my set of complete manuals that the service engineers had, they are more indepth than the normal user manual. I just have to look and see where I stashed them.

[jryter]

Thanks for the reply. Actually fixing the instrument would be another option. It seems to be an electronics issue. The power on the detector started shutting off in the middle of runs and now I can't seem to get it to turn on after fiddling with the manual I have. Our manual isn't the engineering manual so I would be willing to try some more things that you suggest. the vacuum is around 1x10-5. I disconnected the column and just sealed it off with a nut to test. There was a little problem with the heater on the column inlet. When we were working on it the old wiring broke and was replaced. I also found an old 5970 for cheap on ebay for parts but it was damage more than they let on so the quad and ion source is available from that for teaching but i don't trust any of the electronics.

I guess adding a detector for the GS would mean finding a detector to put in. Probably some used ones.....

email me if you'd like. I would love some suggestions before we get rid of it completely. jryter@western.edu

Jarral Ryter

[itspip]

Using is as a "display model" for students would be helpful. You could also have an old column for them to cut and install to give them practice for that, as well as the injection liners and septa.

[jryter]

That is a good idea. putting in columns is not trivial! We don't have a ton of space but.....

[James_Ball]

Thanks for the reply. Actually fixing the instrument would be another option. It seems to be an electronics issue. The power on the detector started shutting off in the middle of runs and now I can't seem to get it to turn on after fiddling with the manual I have. Our manual isn't the engineering manual so I would be willing to try some more things that you suggest. the vacuum is around 1x10-5. I disconnected the column and just sealed it off with a nut to test. There was a little problem with the heater on the column inlet. When we were working on it the old wiring broke and was replaced. I also found an old 5970 for cheap on ebay for parts but it was damage more than they let on so the quad and ion source is available from that for teaching but i don't trust any of the electronics.

I guess adding a detector for the GS would mean finding a detector to put in. Probably some used ones.....

email me if you'd like. I would love some suggestions before we get rid of it completely. jryter@western.edu

Jarral Ryter

TCD and FID you could probably find easily, NPD was not so common and PID was usually an aftermarket one like the ones from OI Analytical. ECD would be a headache if you are not used to dealing with radiation licenses, and if someone is selling one on Ebay I would not buy it because shipping and transporting an ECD without the proper license can get you in trouble. We moved our lab across town and ended up having to ship the ECDs back to Agilent and have them ship them to the new address just to comply with the regulations.

The power on the detector, it is just giving you an error about the Electron Multiplier? Maybe a "no signal detected" type error?

It could be as simple as the EM horn needs replacing, or as bad as needing to replace the high voltage supply. The good thing with working on the electronics in the 5970 is that the circuit boards are color coded with the tabs that you pull up to remove them. The normal user's manual includes the schematics for the boards as well as wiring, if you need that I can email you a copy in .pdf format.

[jryter]

I replaced the detector, EM horn. And still got no power on- I did receive a message like that. I was hoping with my ebay purchase i could replace boards. But I think a rat was living in that instrument and shipping was very bad on it! Anyway yes I think that one could look at each component on the board. I did see that they are easily replaced. I can't remember which one was the power for the detector now. But our manual did list them and did include a bit of info.

Jarral

[KM-USA]

Jarral - Sorry - didn't see this post until today.

Yes, I've still got 5890 circuit boards available, let me know which part number(s) you are interested in. I may also have complete FID detector and FID circuit board as well.

I already sold most of my HPLC parts to a customer from this board, and I typically ask potential buyer to make an offer, my prices are pretty darn low.

I really need to inventory the 5890 GC stuff though.

I'll send an E-mail too.

[osp001]

TCD and FID you could probably find easily, NPD was not so common and PID was usually an aftermarket one like the ones from OI Analytical. ECD would be a headache if you are not used to dealing with radiation licenses, and if someone is selling one on Ebay I would not buy it because shipping and transporting an ECD without the proper license can get you in trouble.

I am compelled to mention the seldom-seen flame photometric detector. I have a 5890 with FID/FPD, and have no use for the FPD but its weird cool-on-column injector stumped me until I posted an image here and someone knew what it was.

OP, is there a specific error you're getting when it cuts out in the middle of a run? What happens when you try to tune with the PFTBA vial? Also check to make certain there's liquid in the PFTBA vial.

[jryter]

I have heard/read of that detector but never seen one in person. Pretty specific use it seems for certain industries!

It was a while ago when stopped/gave up on the MS. It would basically turn off the EM volts in the middle of a run at random times. I tried running a simple single compound several time to try and repeat it but seemed random. So I cleaned the source. Checked conductivity on the filaments, replaced the detector horn, and even put in a new column just to have all new parts.

Yes I did try to refill the tune vial to halfway. Little things like that can really cause a headache!
Now when I tried to run an autotune it says no MS signal detected. I double and triple checked the source set up.... and even swapped in another source we had. I would give it another go if you had some suggestions.... Like I said that was a few months ago so I could go back and fiddle some more before I give up.
Jarral

[Peter Apps]

TCD and FID you could probably find easily, NPD was not so common and PID was usually an aftermarket one like the ones from OI Analytical. ECD would be a headache if you are not used to dealing with radiation licenses, and if someone is selling one on Ebay I would not buy it because shipping and transporting an ECD without the proper license can get you in trouble.

I am compelled to mention the seldom-seen flame photometric detector. I have a 5890 with FID/FPD, and have no use for the FPD but its weird cool-on-column injector stumped me until I posted an image here and someone knew what it was.

OP, is there a specific error you're getting when it cuts out in the middle of a run? What happens when you try to tune with the PFTBA vial? Also check to make certain there's liquid in the PFTBA vial.

An FPD has some features that might be interesting for students - like the ECD it is selective (but nobody thinks you can make a dirty bomb out if it), the detection mechansim is well understood (unlike the FID), and the response is non-linear in a well-behaved and well understood way. Some of the compounds it detects have very potent odours - so you can set up very nice sniff split experiments on onions and garlic for example.

Peter

[jryter]

That sounds interesting! What gasses are needed? I would be willing to try something like that. Would you be willing to party with it?

[James_Ball]

I have heard/read of that detector but never seen one in person. Pretty specific use it seems for certain industries!

It was a while ago when stopped/gave up on the MS. It would basically turn off the EM volts in the middle of a run at random times. I tried running a simple single compound several time to try and repeat it but seemed random. So I cleaned the source. Checked conductivity on the filaments, replaced the detector horn, and even put in a new column just to have all new parts.

Yes I did try to refill the tune vial to halfway. Little things like that can really cause a headache!
Now when I tried to run an autotune it says no MS signal detected. I double and triple checked the source set up.... and even swapped in another source we had. I would give it another go if you had some suggestions.... Like I said that was a few months ago so I could go back and fiddle some more before I give up.
Jarral

Did it give a notched looking baseline? Like the detector was changing sensitivity during the run?

Our 5970 died this way several years ago. Agilent came out and pretty much replaced ever component in it and it still would not go away. If I unplugged the power over night, I could bring it back up and it would run for a day just fine, then have intermittent baseline/sensitivity drops that looked like a square wave, then the lows would be more than the highs before it finally stopped working. Unplug and repeat, over and over. Never did figure out what it was, so replaced it with our first 5973.

[jryter]

No I don't think it did that exactly. But with 20 year old electronics it could be who knows.... And after hearing several stories of people with the instrument in its prime not able to fix it even with the engineer makes me feel like the odds of getting it to work are low.

[James_Ball]

It was probably past its prime by then, well over 100,000 samples through it and the 5973 had already been out a few years. None the less, I really hated to see it go, it was the most stable instrument I had up until then, and probably better than all the ones since.