Troubleshooting ICP/MS: PE SCIEX ELAN 5000
Created | Updated Sep 15, 2008
...work in progress....
Whilst it is incredibly unlikely that anyone else on site has similar problems with their ICP/MS system*, this information has been put here to avoid loosing it. Obviously, if the beeb shut the site down, the information might have to move again, but for now, h2g2 is certainly safer than a bit of paper on a desk in the lab.
On the other hand, there are a few other people who use this researcher's ELAN 5000 who are not so familiar with all the problems which may occur, and in cases when I am not available (holidays, sickness, business trips), this information may as well be held someplace in an easily readable format, so here it all is.
The following abbreviations shall be used in this entry:
- PIS: Plasma ignition sequence
- RFG: Radio frequency generator
Turning on the ELAN 5000
Before we start with the troubleshooting though, let's just mention how to start up the ELAN from scratch, as there are a few things which differ slightly from standard new insturment operating procedure...
In order to get the ELAN ramped up, pumped down and ready to go, the following things must be done several days in advance of analysis (its the pumped down bit which is the problem here - it takes a few days for the vacuum pumps to warm up properly):
- Switch the chiller on: Hit the big black switch on the back, and the wee blue one at the front. Check the fluid levels at the back, the operating pressure (50 +/-5 psi) and the temperature (15.0 +/- 0.5 degrees)
- Flick up the 4 (four) circuit breakers on the back of the ELAN
- Open Argon bottles
- Power up the PC, hit ENTER, 'Ctrl-d', ENTER
- Type ELAN, and hit ENTER
- Wait until software screen opens
- Switch to different workspace ('Alt-F2')
- Type ELANTOOLS, and hit ENTER
- Go to Service - Force Outputs (A-N)
- Force Backing Pump ON
- Go to Service - Force Analog Inputs
- Force Main Water Temperature GOOD
- Hit the big yellow 'Vacuum' button
- Allow vacuum to stabilise to around 1E-5 Torr
- Hit big yellow 'Vacuum' button again
- Wait a few minutes, make sure backing pump switches back on again
- Swtich back to ELAN software ('Alt-F1')
- Applications Shutdown
- Switch off computer
First Things First: the Blindingly Obvious
- Pretty buttons: Note that if one of the buttons which should be lit is not, that could just mean that the lamp has blown. If you can't work out why the damn thing isn't lit, replace the lamp (take care, voltages). If you don't have any more replacement bulbs, just use the ELANTOOLS software to check instrument status.
- Extraction fan: It is on, right?
- Peristaltic pump: Is the end of your sample tubing submerged? Are the clips closed over?
- Argon: Lets hope you have plenty of gas.
- Pump cooling fans: These should both be working as well.
Hardware Issues Which Look Like Hardware Issues
Machine keeps loosing vacuum
Well who knows why? It might be the vacuum control unit died, it might be a instrument to PC interface problem. Hope for the latter. Take the screw plate below the ion focusing board off the machine. Check the number of red LEDs lit (hidden behind left hand ribbon cable). If you have three lit, hit the reset button on the interface board (NOT the reset button on System Sequencer below), now there should just be one LED lit. Reboot the PC, and pump down the system. Most of the time, everything is now fine. If the vacuum dies immediately again, seek professional help, or 100,000 euros for a new instrument.
PIS fails before attempting ignition
Check the argon pressure, and then check the argon pressure sensor status with ELANTOOLS. The sensor may be having a bad day. If there is definitely plenty of argon, play with the argon flow for a while, until the sensor wakes up. Then try PIS again, and hope for best.
Circuit breakers blow near end of PIS
Ha ha. Now you're screwed. If the PIS gets to the stage where it is trying to actually light the plasma, and then the whole machine falls down, this could well be an indication of a short circuit in the RFG. Nasty. Switch everything off, uncouple the machine from the electric, and go find yourself an alan key and a friend who knows plenty about electricity. Take the RFG generator apart, get to friend to check the board, clean the board, re-wire/re-solder the board as necessary.
Hardware Issues Which Look Like Software Issues
Software crashes after sample read delay
Hopefully this is just another instrument to PC interface problem - make sure the autosampler is on, and then reset the interface and reboot the computer (see 'Machine keeps loosing vacuum' section above)
Calibration senseless, ions/sec in millions
Check your Farady box - the springy metal thing which is squashed when the torch is in position. Clean the contacts with some fine sandpaper, and if bits have fallen off, find some good adhesive and glue them back on.
Software Issues Which Look Like Software Issues
Error Message: 'hd: No space on dev 1/40'
The PE Sciex ELAN 5000 interfaces with a PC running under an old version of UNIX which is called XENIX. The actual software then runs over a DOS emulator on top of the XENIX OS. Most of the basic commands you might be familiar with from using a UNIX-like system or an MSDOS-like system will work in the emulator command prompt, which is loaded by typing 'vpix' at login. The 'hd: No space on dev 1/40' message
is a UNIX classic, and basically means that the root directory of your system is full of junk. Just like Windoze machines like a bit of room to play with on the c: drive, so to UNIX wants space in the root directory (which is actually a partition of your hard drive, not a directory, but don't worry). If you type 'df-v' at the command prompt you will see how much space you have on your hard disk. If the root directory is close to 95% full or more, then you need to delete some of the junk. This is where it gets more complicated though, perhaps some other time.
Software Issues Which Look Like Hardware Issues
None found to date (Feb 2008). Still plenty of time though eh?
