Turbo pump paranoia

[Chris Roberts]

I am now at the point where hooking up my chamber and drawing vacuum are visible on the horizon, and have been reading up a lot on turbo pumps following my recent lucky acquisition of one (Pfeiffer tpu060) from surplus. I have never used turbo pumps before, and am starting to get a little antsy about keeping said pump safe, and free of horrific disasters. So, the question I am posing to you all is, just how fragile are these things?

Say I accidentally twist my gas input valve way too far and bring the pressure up to a hundred microns, or I am showing the fusor to others and somebody bumps the table the fusor is on? Does a situation like that automatically yield tears, profanity, and cleaning up metal shards from my chamber? I am not expecting the above situations to happen specifically, I am just trying to get a good feel for this.

I enjoy a statement posted by Michael Kan a couple years back on the forums, "the turbo pump runs on fear". I guess I am merely trying to figure out approximately how much fear is needed for one of these to run... Thanks for any insight.

-Chris

[Jerry Biehler]

One thing you need to know is that some of the Pfeiffer turbos and I believe your is included require the oil cartridge to be replaced at least every two years, preferably every year. They also say if the pump has set more than three years without running the bearings need to be changed. The oil cartridge is about $80.

Put an orifice in your gas inlet to prevent things like overpressure for excessive volume.

Also using something like a capacitance manometer or a convectron gauge with a set point output to shut things down when you get an overpressure can help reduce the risks of damage. The setpoint type manometers seem to go pretty cheap on ebay.

If you want something bulletproof go with a small diff pump with DC705. At worst you will have a small mess to clean up if it gets vented to Atm.

[Doug Coulter]

I run turbos here, pfeiffer ones in fact. Not that much to really get worried about. Just don't drop chunks of anything into one that is running. I run intake screens (a pfeiffer part) on mine just in case, and in fact they're probably saved my larger one when a piece of ceramic in the tank shattered due to arcing -- I just picked up the pieces off the screen, no blood, no foul.

I have further accidentally vented my small one (about the same as yours I think, 65 l/sec) to air a couple of times. You won't forget that scream of the hypersonic rotor tips in air for as long as you live. In all cases I was able to fix that up fast -- within a couple seconds, and it lived fine and no permanent damage other than to my nerves (I paid new prices for these). In one case, something broke, and I just put my hand over the hole (easy since it was there at the time and the reason the thing broke). Once the turbo had slowed a little more, letting it all the way up to air didn't hurt it, and the controller sensed the high power and shut off -- nice.

Don't put a huge air inlet valve you can make a big mistake with. These all have vent valves on the pump itself, which will get you to air safely and quick enough -- they vent between turbo and drag stages which is a place the thing can stand the gas inlet to be even at full speed, no problems other than that you should of course turn it off first.

If it's not real hard, arrange things so that gravity isn't your enemy re things falling in there, and do use a screen, that only cuts pump speed a little, but can save the thing and has here I'm pretty sure. Do some early tests with the turbo not running, bang on stuff and if it breaks or suddenly leaks, well, no risk to the turbo at that point. Once you get the technique in hand, it's not a big deal at all.

Both of mine are over two years old, perfect results, no oil or bearing changes, and have really racked up the runtime hours as I leave chambers pumping down between runs a lot. Just nice trouble free things, no need to be paranoid, just careful.

[AllenWallace]

Chris,

When my son was working on is fusor, we saw a bugle jet electron beam travel down the vacuum piping to the Turbo Pump - and then the pump spun down never to start up again.

You should consider a metal screen to cover the pump from debris and hopefully prevent errant electron beams which wander through copper piping. The downside of a screen is that it reduces the flow rate, taking longer to pump down your chamber.

FYI, we really did not need the Turbo pump. After bake-out and using a larger Welch 1397 pump, it only took 10 minutes to pump the chamber down to a vacuum where the plasma would extinguish.

Allen Wallace

[Doug Coulter]

Yes, I've seen that kind of thing too. And yes, you don't need a turbo, but if you have one, then well use it, they are great! I have seen beams enter my ion gage, which has a big H field at times, for example, but never have run a turbo without the screen, which Pfeiffer says only cuts you down in the 10% range, not a big deal, and their screens are very fine indeed (but more hole than metal).

[Chris Roberts]

It looks like I should easily be able to avoid most of these potential accidents then. My vacuum connection is on the side of the chamber, which goes to a right angle valve with the turbo underneath. That alone prevents anything in the chamber itself from falling in, and plasma will not have a line of sight path either. The pump also came with a metal screen so that will make a nice last line of defense.

As far as over-pressuring the chamber goes, the only gas inlet valve will be the deuterium feed which is one of those fine-tuned sapphire leak valves. I can see myself accidentally over-pressuring the chamber by a couple of hundred microns, but I feel it would be pretty intentional to go much higher. Of course, if one of my quartz feedthroughs implodes that will be another story, but there is only so much you can prepare for, and with a little of Doug's luck my pump could possibly survive that too.

Many thanks to everyone for their comments. So I figure as long as nobody knows of any tales of a turbo disintegrating from small bumps or vibrations to the table it is sitting on, I should be pretty well in the clear...

-Chris

[Chris Bradley]

Chris Roberts wrote:
> I can see myself accidentally over-pressuring the chamber by a couple of hundred microns, but I feel it would be pretty intentional to go much higher.
Then you have nothing to worry about. In my recent experience, a turbomolecular pump will begin to 'bog down' and take the controller to its current limit (and shut down) at a few 10's of mbar (!), but nothing more dramatic than that. Lower than that at 100's of microns is where it begins pumping!

[Dustinit]

Another good reason to have a screen is that the turbo blades are electrically isolated to ground by ceramic bearings. This means the blades will charge up and eventually only pump out neutrals. With the screen in place this charge does not effect fields within the chamber. Something the guys here spent 2 weeks finding out on the MS.
I have seen an instrument being shaken by the turbo manufacturer to show the turbo's robustness but they do make some interesting noises even with very slow angular movements which can be disconcerting. Opening the gate valve with no cones installed does implode the vanes however. Not a mistake you do twice.
Dustin.

[DaveC]

Chris -

The subject of gas inlet control has been discussed before....using a capillary is probably the absolute simplest approach to avoid major unintentional pressure excursions. I've also had good success with a gas supply tank of a couple liters volume, that is maintained at a low pressure, not much about 1 Torr, and often less. With only needle valves - even the micrometer sort - this approach is a big help.

The other simple protection scheme is to have a 90 deg. bend in the intake pipe, in addition to the screen. I've also had to use a large grounded baffle plate... conductance around the edges... to intercept errant e beams and other charged stuff.

Needed more for the ion gages' benefit, than the turbo in my case.

Dave Cooper

[Doug Coulter]

Dustin,
that info is truly new and interesting, thanks. It never occurred to me that the rotor might be isolated, but yes, from other experience, isolated/floating things in tank == unexpected things happen, mostly un-good things. I've even had insulating mica or glass washers explode when they picked up more charge on one side than the other, enough to arc through them. Moral, don't put them in there unless they can stand full applied voltages even if you don't think they'll see that much. Even a bunch of thin layers does not equal one thick one for this, it seems.

In normal use I guess I've never shaken a turbo enough to notice any of that, but these setups of mine aren't easy to shake either -- they tend to be heavy and most of the vibration is the forepump shaking the table/cart if I don't isolate that. I have heard the "groan" from sudden inrush accidents a couple of times (to maybe 10-50 mbar before I got the hole covered, once with the palm of my hand, while using the other to hit the kill switch), and in that case a 90 deg bend won't do it by itself if there are flying bits, but it is sure good in the normal cases.
One of my systems has a 90 deg valve (not a gate) in that spot, and it catches most stuff, it's usually not fully open anyway.