Turbos and intermittent backing

How to get to vacuum, what the classes are, and what is needed for what job.

Turbos and intermittent backing

Postby Doug Coulter » Thu Feb 02, 2017 12:16 pm

First of all, once a system is tight, you probably don't need much of a turbo. Even though my own tank is fairly huge, I really don't need the 500 liter/sec pump I have, though "it's nice to have plenty when it's time to go from 1 bar". Probably 100 l/s or even less would do for me, and in fact, nearly all the time I have mine spun down to where that should be around what its ratings are.

There is an enormous difference between turbo and turbo drag. The latter is super easy to deal with, and I have a Pfeiffer OEM pump station that uses a forepump that looks like it came from the aquarium store - they are that good in re compression ratio and usable back pressure.

A plain turbo pump needs backing more like an oil diffusion pump, and the trick I use may not work, or might require a big reservoir and more vacuum gages (my cheapo pirani gage would work for that). If you have one of those ebay specials, it's probably this kind, and they need a real decent vacuum at the forepump side to really do much - and the lower that number, the harder it is to do the intermittent thing with the backing pump. Most of the ebay specials seem to have come from the semiconductor fab biz, and a lot of them aren't in too bad a shape or are easily cleaned up and repaired, with new bearings available - but a lot of times just cleaning it up and re lubing the bearings (with the right stuff!) is all they need.

There are two reasons to do that if you can. One is that those things wear out a lot faster than turbos. The other is they get hot and waste a lot of power (matters more to me perhaps as I'm on solar and batteries). If the oil type, one has to take a lot of extra care to prevent "suckback" when the pump is stopped. There are special setups sold commercially for that.

Here, I "had too much money" and for the big setup, just used an oil free two stage piston pump that is almost good enough to do fusor by itself. Fusors (depending on your choice of operating conditions) are run right about where a really good two stage oil pump limits out...or close. The thing is, if you want purity, you need to get a lot better than that - if half your gas is "something other than intended" you'll get kinda funny results. I'd advise you to be able to hit 1e-6 mbar *at least* and 1e-8 if you can. That latter is about as good as you're going to do if there is *any* viton seal in the system on the high vacuum side. It diffuses water, our big enemy (damn sticky polar molecule that has lousy recombination times and eats things when ionized, and takes forever to get unstuck from the chamber walls). Butyl doesn't, but diffuses air instead...CF doesn't leak AT ALL, but it's a trick to reuse the gaskets - you don't tighten them fully the first time, leaving some room to tighten a little more next time, and you can get 3-4 times out of one. Not that big a deal with the cheap ones for 2.75" but huge with the expensive big ones. The reason the flanges are so thick, with so many bolts, and they tell you to just pull it in to touching is that the directions are for academics and grad students who will screw up anything requiring real-world judgement. (I'm sure that's not going to make a lot of friends, but I've been there and observed that)

Don't test run turbos with 1 bar on either side, and especially be careful about inrush accidents if one is up to speed - not only can they fly apart with the energy of a car crash, the blade tips are hypersonic and can melt...don't ask how I know this. It's the most horrible sound you've ever heard, combined with money going up in smoke.

For intermittent backing, what I use is a sensor (it's built in to pfeiffer turbo controllers) for how much power the turbo is using. Above some threshold, run the backing pump, below that, shut it off. Works a charm. If you don't have a little hysterisis, the "hunting" the controller does to regulate speed will give you some fits, but a retriggerable one shot driving your solid state relay should take care of that one. Most times it will take only a couple revolutions of the backing pump to be good again, though of course it'll run for half an hour when pumping down from STP.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Re: Turbos and intermittent backing

Postby Bob Reite » Sat Feb 04, 2017 1:12 am

I'm going to have to try that with my setup! While I don't have to worry about the power budget when the sun goes down, my power bill broke the $200 a month barrier, and I'm looking for ways to cut it. I suppose I can just let my system bleed up to atmosphere if I'm not going to do a run for awhile, but I really don't want to go though a bake out every time I want to run a new test.
The more reactive the materials, the more spectacular the failures.
The testing isn't over until the prototype is destroyed.
User avatar
Bob Reite
 
Posts: 29
Joined: Wed Nov 11, 2015 2:02 pm

Re: Turbos and intermittent backing

Postby Doug Coulter » Sat Feb 04, 2017 12:06 pm

It not only saves power and time, but wear and tear on the forepump...And noise in the lab. A side benefit is you get used to the run rhythm, and if it starts running more, you know to go check things. I like free indicators.

It's not just a bakeout either, unless you can bake to "silly" temperatures, well above what viton can take, for example....it still takes a lot of time. Keeping the system full of some noble/monatomic gas at near STP might work, haven't tried it for long periods. I did try dry nitrogen with poor results, Diatomic (N2) or molecular gasses like CO2 do "funny things" later. (I have some fun stories about that) I suggest argon (pure) from the welding store; just buy a small tank and regulator. Dry. Cheap. Not leaky. Easy to pump out later.

The key of course is your high vacuum pump. If turbo-drag, this is going to work and work well. The biggest hassle is the power draw fluctuates as the (you hope it's a) PID loop controls rotor speed. The one I have is a little on the bang-bang side of things and makes setting thresholds a bit touchy. And of course you have to vary those depending on if you're going to be spinning full speed, or say, 30% of that, since for a given back pressure, it'll draw more or less power at different rotor speeds. For my Pfeiffer setup that means when I go to or from "standby speed" I also have to remember to re-set the backing pump thresholds.

I have yet to try this with a plain turbo and a really good backing oil pump, which I have - just don't have an easy way to lash those into a system just now, and am busy with other stuff. I'd think it would work, but the forepump line would be running much lower pressure, and maybe a reservoir there would be a good thing to limit the cycle rate. We have some small oil pumps that make really good vacuum we scavenged off some leak detectors that also had small diff pumps (which we sold or gave away). Those seem to have one of the silicone oils (DC-xxx) in them - I know that's verboten (bad lubricity), but that's what they have in 'em as we got them. In a test of one, I made a tiny bell jar out of a mason canning jar and hooked it all up, pulled it down, turned the mess off, and around 3 months later you could lift the jar off. No oil sucked back. Pretty impressive. So it theory, it should work. Something like an arduino (or an ESP8266 or...) might make it easier to do. I am liking the model of some uP in "whatever the thing is" and using "any computer with software I can run" to be the front panel, with maybe an override switch here and there; I get tired of making front panels and having to go between buildings to do stuff on them. Obviously, one that can serve web pages makes *any* device able to control it, hence the ESP suggestion (or a Pi if you have excess bucks and don't need super real time things). I'm still a believer in hardware gates, filters, and one shots etc for high reliability things - they never lose their programming or crash, though I have to say, the various uP's I use here - unos, Pies, etc - only need rebooting when I want to, the uptime is fabulous. Intel not so much (even with linux), and ESP code for the common ones has to be carefully crafted to stay out of the way of the Wifi handling stuff's timing needs or it goes down...And you have to remember to re-connect if the WAP it is talking to glitches...So those are a lot of bang for the buck, but also more skull sweat to make really reliable.

FWIW, this is one reason I took some time off to develop my "lan of things" code and hardware base. I needed that sort of thing for remote control, but boy, it'll spoil ya, and keeping it on the LAN has huge benefits over putting any of it on the internet. As I sit here, only lifting a mouse finger, I can see status of my entire campus and control a bunch of things, plotted or with CGI's in the (mostly) raspberry pies that are on the spot(s), saving me endless hoofing it around. Doing it on non critical systems first was a good move to learn the edge cases and reliability issues and fixes before hooking it up to many thousands of bucks of physics gear.

A side benefit for the physics stuff is I can set the thresholds such that it's fine just waiting for me to do something, but when I run the fusor, the extra pressure makes the backing pump run continuously, so my solenoid "let gas out" control just works when I need that. I just close the manual valve in the forepump line before a run, one and done.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Re: Turbos and intermittent backing

Postby Peter Schmelcher » Sat Feb 04, 2017 3:49 pm

Bob my setup has a turbo drag and a 4 stage diaphragm roughing pump but running next to you the constant noise gets old fast.

You should consider an ion pump. I have a used 20 l/s Varian which runs at 3300V and the current is an indication of the pressure (100uA is 0.9E-6T). Mine runs continuously at 20uA which is 66mW. At my current pressure the ion pump lifetime should be 5 to 10 years of 24x7 pumping.

Buying a used ion pump has risk, so don’t pay a lot and commission it with a high temperature bake out.

-Peter
Peter Schmelcher
 
Posts: 25
Joined: Tue Sep 30, 2014 9:04 pm

Re: Turbos and intermittent backing

Postby Jerry » Wed Feb 08, 2017 6:03 am

No turbo needs a diff pump to back it. Generic non-drag turbos have about the same requirements as a diffusion pump. With drag you can gat away with fore pressures into the couple torr range, thats why you can run with a diaphragm pump. A good multi stage diaphragm pump from varian, pfeiffer, vaccubrand, etc can get down to a couple hundred millitorr without an issue.

Ion pumps can be a freaking nightmare. There are three of them on a friend's SEM and getting them fired up after sitting was a pain in the butt.

I have been thinking about putting a VFD on the roots blower on my roughing setup. The blower model I have does not have the high pressure bypass on it so you have to shut it down when initially rouging the chamber. I need to get this system back up and running for work, I am installing an ion gun inside the small system to see if I can get better adhesion thermally depositing metal onto plastic.
Jerry
 
Posts: 556
Joined: Sun Jul 18, 2010 1:07 am
Location: Beaverton, OR

Re: Turbos and intermittent backing

Postby Doug Coulter » Wed Feb 08, 2017 3:52 pm

I think that's what I said, Jerry - a non drag turbo needs diff-pump like foreline pressures, which was bad phrasing, I guess, in that what I meant was they need around the same foreline pressure as a diff pump. Hopefully reading that I have some small oil type piston pumps that will do that should have helped make it clear.
We are saying the same thing I believe.

Peter seems to be having good luck with his ion pump, but as you say it's a case of horses for courses, and frequent pumpdowns from 1 bar mean you need all the other stuff regardless. My point was that since you have that stuff anyway, why not use it smart instead of hard? Run the forepump intermittently.

FWIW, my Pfeiffer 2 stage oil free piston pump HAS a VFD in it, according to them, one adapted from a conveyer belt drive. It spins up slowly, and due to the nature of my control system, often is shut down before it comes anywhere near full speed, as about 2 rotations is what it takes to offload the turbo-drag enough for it to stop drawing power above my switch point. I just bang-bang it with a solid state relay...but it does have that gentle spin up built in, and if it's left running a long time, "knows" what the gas load is and will slow down too automatically after awhile. The thing is, that's not as cool has having it just stop, hence the relay I put in there.

Boy, if you've got a roots you're doing the BIG stuff, eh? Or a really frequent cycle. I don't get into my fusor super often, but often enough that this setup I'm using, with intermittent backing and a turbo-drag running at ~ 30% speed gets it done, seems super reliable (6 years no failures) and cost the price of one relay and one switch extra (manual override), which is kind of hard to beat. I'm not going for super hard vacuum, there would be no point in an ion pump with a setup that necessarily contains some viton seals. In that standby mode, it's sitting a couple e-7 right now (outgassing the new BN feed through) and will go to maybe a couple e-8 at full speed...there's simply no point in doing that till the outgassing is done, and even then diffusion through the couple of viton seals means e-8 will not be breached no matter what. No way I can or will design those completely out, and buy new copper gaskets for every change of any kind (most of them are copper, on things that are stable). The stray H field of an ion pump is an issue. Heck, the stray field of even the PKR-251 gage is quite noticeable, to the point where I had to put it on an extension to keep the H field in the tank down to the one gauss range.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Re: Turbos and intermittent backing

Postby Jerry » Wed Feb 08, 2017 8:04 pm

The new scroll pumps also have VFDs built into them. Mine is just before they switched to it.

I have the smallest roots blower that Leybold RUVAC makes, the WA151, it is good for 92CFM with a 28CFM backing pump, which is a Leybold D30. The big chamber is 26x30 and the other is 18x30. Blower can only handle a 98 torr pressure differential between the inlet and outlet so it has to be turned on around 35torr chamber pressure. One nice thing about a blower is it is oil free and keeps pump oil from backstreaming into the chamber/turbo.
Jerry
 
Posts: 556
Joined: Sun Jul 18, 2010 1:07 am
Location: Beaverton, OR

Re: Turbos and intermittent backing

Postby Donovan Ready » Wed Feb 08, 2017 11:34 pm

Sorry to go so off-topic, but that sounds like fun for 1/4 milers if you turn it upside-down. :mrgreen:

Good read, though. I just had to laugh. Today has been.. stressful..
Image Donovan
Donovan Ready
 
Posts: 130
Joined: Thu Apr 17, 2014 2:22 pm
Location: Austin, Texas

Re: Turbos and intermittent backing

Postby Peter Schmelcher » Fri Feb 10, 2017 12:38 am

Jerry ion pumps have a steep the learning curve. Mine is a triode vacdiode and oddly increased the chamber pressure when ever it was powered on. The trick or rather necessity is a good high temperature bake out. The collector plates are a sponge and when exposed to atmosphere they saturate with atoms. Warm the plates up with 3300V electron impacts in a vacuum, and they outgas forever.

I bake them out hot. Even melted an aluminum bracket that was holding the heater cartridge on the ion pump body and destroyed the heater element in the process (then upgraded the bracket to stainless), so perhaps a bit too hot.

My rule of thumb is the pressure before baking begins times 2 is pressure when I terminate the bakeout. The amount of gas expelled during the bakeout is mind boggling, 3 to 5 decades of pressure increase for most of a day.

My setup will run from 4 AA batteries for a couple of days if 120V AC ever fails.

Well worth the learning curve IMHO.
Peter
Peter Schmelcher
 
Posts: 25
Joined: Tue Sep 30, 2014 9:04 pm

Re: Turbos and intermittent backing

Postby Jerry » Fri Feb 10, 2017 2:35 am

Oh, the pumps on these have heaters on them to bake them out. There is a battery backup on the ion pumps to keep them going for months if the power goes out. They are just notorious for not starting up once they shut down for a long length of time even while kept under a vacuum.
Jerry
 
Posts: 556
Joined: Sun Jul 18, 2010 1:07 am
Location: Beaverton, OR


Return to Vacuum Technique

Who is online

Users browsing this forum: No registered users and 1 guest