Page 1 of 1

FAQ - Turbo pump versus Diffusion pump

Posted: Wed Oct 17, 2012 2:46 pm
by Richard Hull
This will be a FAQ designed to just give a bare outline about these two different types of secondary high vacuum pumps that are needed in virtually every successful fusor.

Questions to answer:

1. How much money do you have?

2. How skilled are you at electronics work?

3. Are you just looking to make a fusor and nothing else vacuum related?

We will answer these questions later.

Advantages and Disadvanatges:

A Turbo pump will not foul a chamber with oil if mistakes are made in startup or shut down. A diff pump can foul a fusor chamber if not operated correctly, forcing complete disassembly and cleaning of the fusor and vacuum lines to it.

A turbo pump is expensive if purchased used and warranted to work and if purchased with it required controller and cabling ($450.00 to $1,500). It is incredibly expensive new, ($3,000-$10,000). Factory repairs, if needed, will almost certainly cost more than the entire price spent on a used turbo pump.($500 or more) E-bay bargains on Turbos are often not specified to be working and often include no electronic controller or the special cabling.

A Diff pump is usually rather cheap ($40.00 to $150.00) used and is just not worth purchasing new or even refurbished. The worst repair needed would be for an easily user replacable heating element. ($40.00-$60.00)

A diff pump can stand alone. It is non-electronic in nature and has no mechanical parts or bearings to fail or go bad.

Turbo pumps demand a multiphase electronic controller and cabling. They have a complicated ultra fast rotating blade system and an internal motor and special bearings.

Garbage falling into a diff pump from the main chamber will not harm it nor stop its action.

In a turbopump, the entire item can be completely destroyed in seconds if a tiny fragment of solid material falls into the turbine-like, multi-stage, blade group.

Both pumps may or may not require water cooling depending on you choice at time of purchase. (there are air cooled types of both pumps which the user should look for and demand to keep operations simple and inexpensive.

Answers to questions...................

1. If money is no object then either pump is good, but the turbo might be better

2. If you are not electronically inclined to do limited controller repairs or to build your own controller, if your cheapo e-bay turbo pump has no controller, then the diff pump is a far better choice.

3. If the fusor is your only quest and you will not continue with vacuum related work in future then the diffusion pump is your pump. No need to spend big money on a deep pumping oil free system if you are not going to need that capability in future.

Finally, in real life, most would-be fusioneer folks opt for an air cooled diffusion pump due to price, simplicity, ease of repair and maintenance, etc. Serious researchers and "vacuum-head" purists planning to go into other vacuum related research usually choose the turbo pump. Superlative "mega-neutron" fusors have been successfully operated using both types of pumps.

Search other FAQs for more info on the selection, care, feeding and operation of the diff pump.

Richard Hull

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Wed Oct 17, 2012 5:21 pm
by Doug Coulter
As the owner of a few systems of either type, this is simply spot-on.

Diff pumps work great for dirty processes, turbos for system cleanliness.
Most kinds of dirt (like dust from deposition) just don't bother a diff pump at all.

Don't drop a chunk of something into a turbo - a big one has almost the stored energy of a Buick at highway speeds. This is why they all have screens - but don't count on them being good enough.
I've had glass shatter in the tank, and that time the screens were good enough, but it really threw me a scare.

Either will be messed up by an inrush accident. In the case of the turbo, you can melt the blade tips or do other damage (bearings/shaft). In the case of a diff pump, you lose the oil into the forepump.
That oil usually isn't cheap - nor is it the right lubricity to just leave in the forepump (there are exceptions, like diffoil-20).

So, don't do that!

Many diff pump oils will be ruined or even catch fire if exposed to atmosphere or any oxidizer at STP while hot. I've found this to be rather a pain. To prevent this, you need more valves if you want to do a quick change to something in the tank, and those can eat up a lot of the price differences.

I find the turbos easier to operate in "set and forget mode". Further, you can adjust the pumping speed over a *much* larger range than with a diff pump without needing a conductance valve. We found we could get about 2::1 pump speed range on a diff pump with a dimmer control, but not much more than that. I can easily get over 100::1 on most turbos. For a flow type system, this can be the clincher.

I've done turbos both new and used, and yep, they are expensive. I have yet to have a controller issue and both have run many thousands of hours with no maintenance whatever. So, total cost of ownership has for me been competitive. For example, my smaller 60 L/S Pfeiffer pump station shows over 31,000 hours on the turbo and is fine. My larger (512 L/S) Pfeiffer system shows over 41k hours, no problems (and right now it's sitting at 7e-9 mbar in the big tank!).

While you can probably figure out a way to arrange it on a diff pump, with a turbo you can have the forepump switch on only when the turbo starts to draw extra power due to a pressure rise. This lets me, even on solar power, leave systems pumped down for months on end. I could not do that with a multi-hundred watt diff pump even if I rigged a pressure gage/switch to control the forepump.
Right now, my 512 L/S turbo is drawing 32 watts to spin full speed. If I figure in the cost of more PV panels and house batteries, the total cost to me is LESS than with a diff pump, assuming I'm going to leave things pumped down and very clean.

Either work great for a fusor, which is just past the limit of what a mech 2 stage can do - you only need a tiny bit more for a fusor, so actually, either type is bordering on overkill for the job here. For purity, though, I like to see a base system pressure about 1/100th or better lower than I'm going to run, but I'm not real sure that matters much to anything but how long it takes to get to purity in there, as a flow type system will "wash" all that stuff out anyway. The level of purity I can get with the turbo lets me run in a batch mode with approximately .02cc of STP D per run (it's a big tank), another cost saver. It's not like Walmart sells tanks full of D.

Having both, I see zero evidence that a little diff pump oil in a fusor messes up anything much.
It's more a question of what you can cobble together on your budget and how you like to operate.

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Tue Feb 24, 2015 12:34 pm
by Adam Szendrey
My take on this:
After reading a number of turbo questions, I looked up this FAQ to see what it had to say about this matter (what high vacuum pump should one choose). I firmly believe that the turbo pump is only a valid option if you have money to burn, you work at a research institute, or in the industry, or you are a lucky SOB and get a perfectly operational one , very cheap or free. I'm very surprised how some cling with tooth and nail, to a turbo even when it's such a pain in the ass (and wallet) if any part of it gets a fault.

Even though there is the possibility of fouling the system with oil if one is careless, this is actually not that easy to do with some simple precautions. For example use solenoid or pneumatic valves, (or install a servo on a hand operated type) not hand operated ones, and wire the system in such a way that there is an interlock system. I'm pretty sure you can get used valves cheaper than used, working, turbos (with controller), and they'll make your system more versatile too. For example you can't open the vent valve while the butterfly valve at the top of the diff. pump is open, and the diff pump is hot. It's really not rocket science. Although I think that's probably overkill, you could simply have some warning lights installed on the system, or whatever...or just don't operate it when you are drunk or very tired :P. You shouldn't do that anyway. Accidental air inrush can destroy a turbo as well as mentioned above, although it wont "oil up" the system, but will cost you a lot more than just cleaning the lines and chamber, and maybe an oil change in the forepump and diff. pump. You can't really destroy a diff. pump, at least not easily. Even if you cook it you are looking at some hard scrubbing and a lesson learned (yes the oil for it can be expensive, but you really don't need the high end stuff for this application...). I have worked on systems which require a clean, oil free chamber, and they had a large diffusion pump on them. Water cooling is not scary at all... They also had a water baffle above it and also a LN2 baffle, latter being a bit hard to "feed" by the amateur (LN2 not being readily available, although possible to make at home), but you could use a water baffle combined with a chiller (use an old fridge, or an AC unit...) to achieve rather good results. A turbo is just an expensive fault waiting to happen in my humble opinion, although it does look and sound cool, and as evidenced by Dough, mostly reliable, but truly beyond being completely oil free, it pretty much gives you zero advantages. Here I need to mention that if you want a truly oil free system you should use a dry forepump as well...multi-stage roots, or scroll, etc. They are a tad expensive. Yes a turbo can shut down somewhat faster than a diff pump, has a wider flow rate range, but honestly why do you want to shut it downm, and why do you want to vary the flow rate at the pump? Just install a butterfly/gate/whatever valve in between your chamber and the high vacuum pump. I think money spent on a variable valve is better spent than a turbo with its controller. In an experimental system it's pretty much a must, in my opinion, as you'll be opening and closing the vac. chamber relatively often (testing grid designs, magnetic fields, ionization methods, whatever). This is why, by the way, I don't understand the obsession with CF flanges I see here sometimes. It's expensive, it requires a new seal ring every time you open it (yes I read that you can actually use a flat plastic/rubber ring maybe?) and pretty much nobody with a fusor, needs the vacuum ranges that CF was designed for, also it's more tedious to open/close . I'm a firm believer in metal "bell jars" instead of spheres and what not. Very easy and quick access. As I mentioned, this is NOT a UHV system, and it's not a fixed industrial or other process device. It's a vacuum system for constant experimentation. Properly maintained KF is perfectly suitable down to 1e-5 - 1e-6 torr (in some cases even beyond but you have to pay attention to the o-ring material, and shouldn't use aluminium). Right okay, the diff. pump consumes more power, granted, but considering the price of the pump you'd need to run it quite long to cost you the difference in electricity, I'd say. But again, some get a working turbo (complete with controller) very cheap. Although I would guess that's quite rare. I only speak out "against" buying one without a controller, or a new one or an expensive old one...Waste of money.

There, my 2 cents, for what it's worth.

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Tue Feb 24, 2015 8:59 pm
by Richard Hull
I appreciate well considered opinions on any FAQ and the voice of experience speaks volumes. In the final act, it is the decision of each would-be fusioneer to determine their path to vacuums beyond the forepump. This FAQ is just a way to help make wise choices based on the hard won experience of those who have "been there, done that".

Richard Hull

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Thu Feb 26, 2015 9:53 am
by Adam Szendrey
I agree Richard. I thought I'd share my own little ramble about this matter, as sometimes I feel some amateur experimenters spend a lot of money on stuff that's not really necessary. Maybe some will reconsider if they read this (sadly that doesn't seem to be the general practice, reading FAQs I mean). That money could be spent on much more useful components, or extra components increasing the diversity of the system. And some turbo related threads boggle the mind, when considering all the effort, money and time spent in some cases, just to have a turbo. I really like turbos by the way, they are a joy to work with, especially for the semiconductor industry, where contamination is an issue. But there any repair is payed for by the company, and the units are bought brand new, with warranty. I just don't feel that they actually have any advantage in an amateur lab. I'd say it's better to think a bit, and design a system with a diffusion pump, than just throw money at the project to make it "simpler".

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Thu Feb 26, 2015 9:43 pm
by Richard Hull
My feelings exactly, but for some, a turbo is either a sign that you are on the top of the heap or, more reasonably, They arrive already owning one or expect to do a lot of future serious vacuum work outside of simple fusor work which really doesn't need one.

As noted by me in other posts, diffs were a drug on the market 10-20 years ago as the big operations with big wallets tossed them out like old rusty plumbing in a rush to turbos; once turbo prices fell well below $10,000. Unfortunately, now it is old turbos....often blown up turbos or turbos with no controller, no cables and no data that are common on the market and E-bay. The diff pump may become a collectable real soon now. Who would have guessed that.

Richard Hull

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Thu Feb 26, 2015 11:08 pm
by Adam Szendrey
Indeed! I still find a lot of, relatively, cheap diffs on ebay though, but that will probably change eventually. Used turbos , in various "unknown but turns freely" states are still kind of expensive, and I rarely see one that is complete, exactly as you wrote. I was pondering about this by the way...figuring out an alternative for future fusioneers, when diffs become a rarity. Problem is, that a booster pump (say, a roots atop the rotary) is not any cheaper and offers a modest improvement in vacuum. But hey, building a diffusion pump yourself, might be a fun project! Even if it's modestly efficient (and you can quite easily achieve that), it'll pull you down into the range you need for a fusor. It would pretty much be a pipe with a bigger flange atop, a small one in the bottom, a kettle heater for the oil, and a stack made of thin stainless steel sheets, or maybe aluminium (and a simple copper pipe cooling coil around the upper two thirds or so). Diffusion pumps are so shamelessly simple. One could in fact take apart an existing small or mid sized diff and just copy the stack inside. For now though I think one can still find them for a decently low price.

As for serious vacuum work...I went to one of the universities in Hungary when I was taking a course in vacuum technology, and I had a chance to look at some of their systems, among which there was a rather big, old SEM. It still used a diffusion pump (it's "forepump" was a pair of sorption pumps, quietest system ever!), and the old prof. said, they really prefer the diffusion pump over a turbo. For one they have plenty of spare parts for it, although they don't really need them, as a fault is very rare (pretty much only the heater can die). They could even build a new diff. pump from scratch, locally. And a SEM requires a very clean chamber! They used baffles/traps, of course.

Re: FAQ - Turbo pump versus Diffusion pump

Posted: Fri Feb 27, 2015 9:27 am
by Richard Hull
That's cool. Sorption pumps are cool and quiet and if you use a diff with all the required LN2 traps and baffles, you only need a beer budget to afford the LN2 which quietly drives the entire system fore to aft. As an old fart I tend to fart a lot and the noisey forepump covers the initial sound and later I explain the resultant fumes as burning oil I spilled on the outside of the diff pump's boiler or a failing oil mist filter on the fore pump.

Kidding aside, I have a sorption pump, but haven't the budget or desire to drive to the gas distributor to fill my 20 liter LN2 dewar every time I want to run my fusor. I used to fill my small 10 liter dewar for the HEAS event in October as I have a very large 3" diameter Y,Ba,Cu super conductor disc and would do a float demo of some NdFeB magnets as part of a show and tell demo, but haven't done it in the last three years.

Richard Hull