Restoring a Mass Spectrometer - maybe

[TerraHertz]

Recently I scored a mostly complete Mass Spectrometer / Residual Gas Analyzer. Which *might* be possible to restore to working order, and would be very useful for a project of mine. The only thing actually missing was the rotary backing pump and plumbing. The turbomolecular pump was still there.

But I didn't want to start a thread in repair until it seemed fairly certain it could be restored. Which isn't clear yet.

However an unexpected problem appeared, where you guys may be able to help me.
The system is a Model: DCQ300RF, made by Anglo Scientific Instruments.
The OS on the control PC is Win98SE, to give an idea of the vintage.

Google finds:
  http://www.angloscientificinstruments.com/
  Telephone: +44 (0)1805 625 205
  Postal address
    Anglo Scientific Instruments,
    3, Devon Units, Hatchmoor Industrial Estate,
    Torrington, Devon. EX38 7HP. UK.
  email: angloscientific@live.co.uk

The web site exists, though it's kind of minimalist and the most recent date I can find anywhere on it is from 2011.

Then, while evaluating the condition of a critical PCB (high input impedance amplifier for the quadrupole sensor) I found one of the key parts is broken. A tiny resistor... which unfortunately is the DC feedback path to the input FET amplifier summing point. A SMD resistor, but suspended in free space on two wires to avoid PCB surface leakage. The wires had been left under tension, broke the solder contact off one end of the resistor. It will be a very high value, and is unmarked. My HP 3456A says it's open circuit, ie > 1 G-ohm. Most likely the resistor is actually broken and open circuit. I have no way to even guesstimate the correct value. It's the bit that sets the loop gain for the tiny current of ions reaching the central sensor electrode in a vacuum. Could have been almost *any* other part in the system and I'd have been able to work out a probable value. But not this one.

So I tried emailing angloscientific@live.co.uk
Asking what is the correct value for that R7. (Also if schematics and calibration procedure are available, but I never had much hope of that.)
They don't respond. At all. Not even a "sorry, no." (Edit: 20170202: received a response. Still in progress.)

My question here is, can any forum member in England, please find out if that company still actually exists, and will respond to anything. They appear to have been quite a small operation. Maybe they just closed down recently, and the web site is still up? Or they are still on Christmas holiday or something.

The arrow in the pic indicates the stress-failed solder joint. After removing the wire it looks like there is still metalization on that end of the resistor, but I can't measure any value.
They should have used a very fine, soft wire for one side of the resistor.

The tall teflon post is actually a feedthrough to the sense pin of the guadrupole unit. The round IC is an AD549.

[CJay]

Companies house don't have an entry for Anglo Scientific Instruments and their phone number is dead, they do list an Anglo Scientific in Hertfordshire though

https://beta.companieshouse.gov.uk/company/03914022

[TerraHertz]

Companies house don't have an entry for Anglo Scientific Instruments and their phone number is dead, they do list an Anglo Scientific in Hertfordshire though

https://beta.companieshouse.gov.uk/company/03914022

Thanks!
Oh oh.
https://beta.companieshouse.gov.uk/company/03914022/officers  suggests a company that just went under.
Some chiefs still on board, others and all the indians resigned.  'Indians', bearing in mind one of them lives in a castle.

Not an email address to be seen.

Bah. My luck, probably somewhere in Devon someone recently threw boxes of schematics and spare parts for this system into a dumpster.

[samofab]

SMD resistors over 1Gohm do exist. And if the resistors was suspended away from PCB it's likely that it is a very high value. Try to measure it with something that goes up to 10Gohm.

[CJay]

Companies house don't have an entry for Anglo Scientific Instruments and their phone number is dead, they do list an Anglo Scientific in Hertfordshire though

https://beta.companieshouse.gov.uk/company/03914022

Thanks!
Oh oh.
https://beta.companieshouse.gov.uk/company/03914022/officers  suggests a company that just went under.
Some chiefs still on board, others and all the indians resigned.  'Indians', bearing in mind one of them lives in a castle.

Not an email address to be seen.

Bah. My luck, probably somewhere in Devon someone recently threw boxes of schematics and spare parts for this system into a dumpster.

Doesn't look too bad to me, none of those resignations are particularly recent and the accounts look fairly healthy and up to date don't they?

What I would question is if they are the same company. Would be worth a phone call if I can find a number.

[TerraHertz]

SMD resistors over 1Gohm do exist. And if the resistors was suspended away from PCB it's likely that it is a very high value. Try to measure it with something that goes up to 10Gohm.

I don't have anything that can do that.  Oh wait... nanovoltmeter. Hmm.
I've yet to look at the end of the resistor with a microscope. Fingers crossed it will be just the solder that fractured, leaving the hard metalization on the alumina.  In which case I'll just solder on a fine wire, clean the resistor well, and try it in operation.

With the sliiiight complication of having to get the high vacuum system working before that will be possible.
At least I have some nice test gases. http://everist.org/spacejunk/sell/helium-3.htm


I had posted a few random pics elsewhere:
https://www.eevblog.com/forum/testgear/list-your-test-equipment-score-here!/msg1121700/#msg1121700
https://www.eevblog.com/forum/testgear/does-anyone-know-how-to-get-the-handles-off-this-case/
But will do a more complete tale here later.

[TerraHertz]

Doesn't look too bad to me, none of those resignations are particularly recent and the accounts look fairly healthy and up to date don't they?

The last one was for 2015. Maybe 2016 is still coming.
Mysterious to me. Looking through the docs there, I can't find anything that even hints at what that company does.
Debtors 344,539, but only 250 in bank? How do they pay wages?

What I would question is if they are the same company. Would be worth a phone call if I can find a number.

Might be completely unrelated. But strange to have two different companies existing for a long while with the same name.

[CJay]

Doesn't look too bad to me, none of those resignations are particularly recent and the accounts look fairly healthy and up to date don't they?

The last one was for 2015. Maybe 2016 is still coming.
Mysterious to me. Looking through the docs there, I can't find anything that even hints at what that company does.
Debtors 344,539, but only 250 in bank? How do they pay wages?

Well 2015 was filed October 2016, so I'd not expect 2016/17 until October this year.

It's a bit odd that they don't seem to exist according to Google and yeah, info on what they do is sparse, didn't have time to dig more though. Happy to give them a call if you find a  number though.

Lots of companies are debt financed, they use the bank's money. Can't last forever and it's not ideal but...

What I would question is if they are the same company. Would be worth a phone call if I can find a number.

Might be completely unrelated. But strange to have two different companies existing for a long while with the same name.
[/quote]

It's not not that uncommon, one is Anglo Scientific, the other is Anglo Scientific Instruments, there are lots of companies with almost identical names.

 If there's a company reg number on the Devon company web site you could track them that way through companies house and that'd give you the definitive status of ASI but that's not necessarily going to tell you if they were bought out, liquidated and re-emerged from the ashes a day later etc.

[Alex Nikitin]

Hmm, I work for a company that makes RGAs and very familiar with this kind of circuitry. Most likely the resistor is 20Gohm (my current company is apparently distantly related to the Anglo Scientific Instruments - both originated from the same company long time ago, so the electrometer circuit for their and some of our old RGAs were most likely designed by the same person). I am pretty sure that you won't be able to restore the broken SMD part. Check Farnell for suitable replacements, these are available and not expensive (unlike better through-hole ones). Main problem with an SMD part would be to keep it clean, you need to solder the (preferably much thinner) wires to it, then wash the resistor thoroughly with IPA, dry it in an oven at 60-70C for few hours and only after that carefully solder it on the board, with minimum solder and flux on the Teflon support. There is a slight chance I might find some more info on the unit you have, PM me if you wish.

Cheers

Alex

[TerraHertz]

Update: I received an email reply from the angloscientific email addess, with some information and a schematic of a very similar amplifier module. Conversation in progress, asking their views regarding public discussion of technical details of the system. (Meaning how much I can say here, without annoying them.)

Don't know yet if the company is still trading, or whether more tech information may be available.

Positive:  replacement filaments are still available. Cross off one potential project killer.
Also yes, that resistor is 10G or 20G, he's not sure.

Alex, thanks. Will do.

[joeqsmith]

I've always been curious about buying equipment like this.  How would you know what it was used for, what was ran through it?  Seems easy enough to dump the stuff on eBay.   I wonder if you contacted some of the parts suppliers if they would even allow you to return them for repairs because of the risk of exposure.  Maybe they would have you sign something stating what was ran through it but with surplus ebay, there would be no way to know.    Maybe it's a non issue. 

[TerraHertz]

I've always been curious about buying equipment like this.  How would you know what it was used for, what was ran through it?

It definitely is an issue with some such gear. For instance I have a rotary vac pump stashed away, from an ultracentrifuge. The oil itself used in that is highly poisonous (I found out by research after taking the pump, before touching the oil.)
In general used vacuum systems should be treated as hazardous.

This one though, I'm told was used in some research into coal processing. Pretty innocuous. Also the RGA itself should be quite clean inside, since the only path into it is via a couple of meters of fine stainless capillary and two very fine needle valves in series. This is about as good as it gets.

Seems easy enough to dump the stuff on eBay.   I wonder if you contacted some of the parts suppliers if they would even allow you to return them for repairs because of the risk of exposure.  Maybe they would have you sign something stating what was ran through it but with surplus ebay, there would be no way to know.    Maybe it's a non issue.

Well, not a problem for me, since ebay vacuum gear is usually too expensive for my micro-budget anyway.
This unit was effectively a dumpster find. Free.
The missing rotary pump is not a problem since I have a few already.
There's also no facility for pump-down/up automation and sequencing, so that will have to be added.

[Kleinstein]

Depending on what you want to do with the mass spec, it might not be such a good idea to use a classical rotary pump. For a clean system and if the turbo pump allows it, an oil free membrane pump might be the better choice.

Contamination can be a problem. If you send in things for repair, one usually has to sign paper-works to shown what you have done with it and that there is no hazardous contamination.

At the university is was quite a trouble to dispose an old system as it had used Hg diffusion pump and was used with Hg containing stuff.

[TerraHertz]

Depending on what you want to do with the mass spec, it might not be such a good idea to use a classical rotary pump. For a clean system and if the turbo pump allows it, an oil free membrane pump might be the better choice.

That's a good point, thanks. Though, I'm constrained to using what I have, since I'm poor.

Contamination can be a problem. If you send in things for repair, one usually has to sign paper-works to shown what you have done with it and that there is no hazardous contamination.

As for 'repair', given the costs of commercial repair the rule is, either I repair something myself, or it doesn't get repaired.

Speaking of which, one of the three turbomolecular pumps that I was given at the same time as the RGA, is a
RIGAKU   Model RTP 300 VRC    SN: RD0793-05  Date: 1995-6
It's the biggest of the three, but has no controller or cable. Water cooled.
Has some dirt inside on rotor face. Not adhering, due to oil film. Needs cleaning.
Ding on bottom rim - dropped on concrete at some time? Effect on rotor balance unknown, still spins freely.
Dim: H: 275mm, Throat: 100mm.

Googling, it turns out http://www.rigaku.com/en  no longer make turbo pumps. Not even listed among their products.
Very few mentions to be found. A couple of controllers on ebay:
http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.Xrigaku+rtp300vrc.TRS0&_nkw=rigaku+rtp300vrc&_sacat=0
Only the $355 one has the necessary cable. Not to mention the *other* proprietary sockets on the controller.

And this system, which contains the same pump and controller: http://www.ebay.com/itm/X-ray-diffractometer-XRD-Rotating-anode-Rigaku-Rotaflex-RU-200B-HBC5-/400567351527?hash=item5d43acb8e7:g:~XUAAOSwYIxYAOg-
(Ha ha, someone has pinched the rotary backing pump out of that machine, but that isn't mentioned in the listing.)

Given all that, I think I'm unlikely to ever be able to run this pump, even if it works. So, might as well take it apart for fun.

But first, just in case, I'll ask here if anyone knows of:
 - User and or service manuals for the pump and controller?
 - Any ideas of where I might find a controller, for free or up to $100?

[Kleinstein]

Getting a controller for the pump could be difficult. Maybe you can get some from Iran, including the stuxnet virus.   

Even if the rotor is rotating free at low speed, there is still a chance the bearing may need service / oil change.

Normally the mass spec should not need a big pump. Small ones (e.g. 40 mm inlet) can get away without water cooling.

[TerraHertz]

Getting a controller for the pump could be difficult. Maybe you can get some from Iran, including the stuxnet virus.   

Ha ha. Actually older TMP controllers tend to be analog; no software at all.

Even if the rotor is rotating free at low speed, there is still a chance the bearing may need service / oil change.

I know. Hence finding a manual is essential too. Oil type, level, procedure, etc. Chances are extremely low.

Normally the mass spec should not need a big pump. Small ones (e.g. 40 mm inlet) can get away without water cooling.

Oh, I should have mentioned. The RGA is not _the_ project. It's a nice addition to the main project. Which _does_ need a large turbo pump.  And a rapid pump-down rotary (this: http://everist.org/NobLog/20141212_racked_serendipity.htm ) as well as a not-so-big rotary pump for operation.

Btw, the mass spec has its own small turbo already. A Varian TV-301 Navigator, complete with controller. Too bad someone beat me to that system and took the backing pump. But no matter.

Another part is a Varian 936-60SP He leak detector (which I'd hoped to modify to serve as a mass spectrometer for low atomic number gases, but won't don't need to if the Anglo Scientific one can be restored.)

[TerraHertz]

Since the Anglo Scientific RGA can't function without a working vacuum system, and I have a pile of old vacuum gear to go through and see what works, I may as well turn this into a general 'vacuum stuff progress' thread.

And so... *another* thing in that recent score, has been sitting on the bench untouched due to expectation of it being ruined junk, and having too many other things to do. Yesterday I was conducting a vicious cull of pointless junk that was accumulating in the  workshop (running out of space) and finally got to that bit of gear.

It's a HP 5973 Mass Selective Detector. Google it.
http://www.aimanalytical.com/Manuals/5973N%20Inert%20MSD.pdf    hardware manual

ebay, web
  http://www.ebay.com/itm/HP-AGILENT-5973-NETWORK-MASS-SELECTIVE-DETECTOR-/361417325460
  https://www.gmi-inc.com/agilent-hp-5973-mass-spectrometer.html

In use it's supposed to sit next to a HP gas chromatograph, to separate different fractions for analysis in the MSD. But it can work alone too.  This one was made in 1999, according to an internal cal label.
The reason I expected it to be useless, is it's missing some bits (both hi-vac and backing vacuum pumps), and also had been dumped outside in the rain for a long time - worse than the Anglo Scientific unit.

Anyway, regardless of whether there's any chance (probably zero) of getting it working, it's very cool looking.

I like the Anglo Scientific one because it's more likely I can maintain and reverse-engineer it due to its simplicity.
The HP one is obviously orders of magnitude harder, even though the principles are the same - its still a quadrupole sensor. Also the boards have dust and water trails across them, though at first glance they look like they should be restorable with a good cleaning. Leaving whatever original faults existed, if any. Fingers crossed it just got thrown out because the software is for WinXP. Or the mating chromatograph unit failed, or something.

The whole side of the vacuum chamber swings out, with just a couple of thumbscrews and 4 connectors to undo.
Someone had left the thumbscrews undone, but the cables were still plugged in. Which was a bit odd.  Anyway, no visible damage inside the vacuum chamber, though it was open to the weather for a while via the underneath port to the removed hi-vac pump, and the non-compressed side panel seal (not held down by the thumbscrews.)
Probably would need a good solvent clean and bakeout at minimum.

The sensor though... wow! Even if it's utterly dead, this is by far the nicest thing in my spacejunk museum. The photos totally fail to do it justice.

Notice the wires that are insulated with glass beads. This works best for LV signals, because there's no plastic to outgas, and the beads don't result in any trapped voids, that leak air or whatever to the chamber for a long time. A couple of wires with teflon (probably) insulation would be for signals high enough to ionize low pressure gas, so beads can't be used.

Another fun thing - photo 6252. It's the transformer developing the antiphase high voltage signals to the electrodes of the quadrupole. How do you like the physical arrangement of the three windings? Not intuitive? I love it! The small thick wire air coil is the single primary. The two litz wire segmented stacks are the secondaries, in antiphase. There are internal adjustment slugs, accessible from outside the case.

Edit to add: Oh, if anyone happens to have/know of the following, please PM me:
 - Install media with the drive software suite.
 - User manual for the software.
 - Service manual with schematics and check/cal procedures.

What kind of PC HPIB card it expected (if it mattered), I don't yet know.

[edpalmer42]

I don't know if you've already found it, but this collection of brochures mentions the DQC300M RGA from ASI/Larimax.

http://orionte.ipfn.ist.utl.pt/files/HugoAlves/Catalog.pdf

[TerraHertz]

I don't know if you've already found it, but this collection of brochures mentions the DQC300M RGA from ASI/Larimax.

http://orionte.ipfn.ist.utl.pt/files/HugoAlves/Catalog.pdf

Thanks. Looks like a slightly later version of one I found on the hard disk of the system. Nice for historical tracking.
The manual and windows HLP file are also there, along with the application software.

It doesn't look like I'm going to hear anything more from the person who replied that one time. Who hasn't responded to later questions about whether there's any reason I shouldn't post whatever I find. But give it some more time.

[Alex Nikitin]

Thank you for the photos of the HP quad - very nice! The coils arrangement in the RF transformer is familiar to me  . The quad itself is a thing of beauty - gold plated monolithic hyperbolic.

Cheers

Alex

[TerraHertz]

Thank you for the photos of the HP quad - very nice! The coils arrangement in the RF transformer is familiar to me  . The quad itself is a thing of beauty - gold plated monolithic hyperbolic.

If you'd like more detailed closeup photos of anything, just ask.

I still haven't hunted for a detailed theory of how quadrupole detectors actually work. Think I understand it, but my guesses have been wrong often enough.

Need to get some decent disposable vacuum-cleanliness adequate gloves, (and rtfm) before I try undoing the obviously removable parts to have a look.
What do you use? Nitrile rubber surgical, with no talc? I assume latex surgical gloves would be complete fail.

I noticed the inside of the vacuum chamber has a light misting of the diffusion pump oil. None I can see on the quadrupole, but of course it must be there too. Presumably that isn't good.

[TimFox]

I don't have a reference to cite, but I remember investigating the quadrupole mass spectrometer 40 years ago and the theory behind it is very elegant, involving Mathieu functions as a solution to differential equations (originally for waves on the surface of a circular lake).
For the analyzer, see https://en.wikipedia.org/wiki/Quadrupole_mass_analyzer, and for the Mathieu functions see https://en.wikipedia.org/wiki/Mathieu_function.

[CatalinaWOW]

The glass bead insulation is neat, but given that I think I see 10 insulated wires and only 4 with the glass beads they must have been desperately close on their outgassing requirements to make it worthwhile for less than 30% reduction in that element of the residual gas sources.

[Messtechniker]

I still haven't hunted for a detailed theory of how quadrupole detectors actually work.

Some info here around p. 96:
https://www.leyboldproducts.com/media/pdf/01/ab/0a/FVT_Fundamentals_of_Vacuum_Technology_EN57d6d275b6c8b.pdf

Yours Messtechniker

[TerraHertz]

I still haven't hunted for a detailed theory of how quadrupole detectors actually work.

Have been onto that since.

Some info here around p. 96:
https://www.leyboldproducts.com/media/pdf/01/ab/0a/FVT_Fundamentals_of_Vacuum_Technology_EN57d6d275b6c8b.pdf

That's a really useful book, thanks! Saved.  So good I looked for a 2nd-hand physical copy - alas none found yet.