#2 FAQ - Mechanical Pump - Diffusion pump?

If you have a question about this topic, the answer is probably in here!
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

#2 FAQ - Mechanical Pump - Diffusion pump?

Post by Richard Hull »

The beginner to vacuum within these forums is bewilderd at first. Let me start by saying that we in the fusor world are different from detailed scientific vacuum technology, but not free from many of it basic tenats.

The kind and quality of the amateur fusioneers vacuum system will be determined not only by his wallet but by where he or she wants to go in the effort. Where you want to end up, if you will.

What kind of pump do I need?

#1 The mechanical pump

For the youngest enthusiast looking at making a simple demo fusor for a science project a simple, even rather weak or overused mechanical pump is all that is needed, provided he plans to end the endeavor here.

Completely out of the picture are any type of diaphram pumps used as so called vacuum pumps. Also out are any water pumps or any fluid pumps. Most all piston pumps are out as well.

There are "single stage" and "two stage" mechanical pumps. You really want a two stage pump. The difference is a longer pump down on single stage units and a somewhat higher low end pressure. The cheaper pumps are often single stage. Make sure of what you are buying.

You need, at minimum, an old air conditioner repairman's vacuum pump. These can be found used for anywhere between $50.00 and $150.00. The cheapest new two stage pumps of this type typically run a minimum of $200.00, but a good new one of higher pumping capacity might run nearly $300.00 or more. All of these pumps are called a "direct drive" type, in that the motor's shaft is directly coupled to the pump's shaft. Such units are usually rather light weight, (under 40lbs), and have a carrying handle. Direct drive pumps are usually rather noisey and run at high speeds.

The best mechanical pump might be a used "Scientific Vacuum" pump. While there are many modern scientific direct drive type pumps, the classic scientific pump is termed a "belt drive pump". These are made to run continuously for scientific and educational use. They are all very heavy and weigh, with their metal base, well over 50 lbs with large pumps weighing over 150 lbs. Oddly, these pumps can often outperform a direct drive type and cost far less, provided they are found and purchased, "used".

These pumps are generally very quiet and less rough on the ears that the high speed, often straining, direct drive units. They run at half to one quarter the speed of direct drive pumps. A gentle lub, lub, lub sound is the norm.

The above are the normal mechanical pumps used by virtually 100% of the amateur community.

Note*** If you buy a used pump, be prepared to have to clean and refill it with pump oil at a minimum. Some used acquisitions may need minor to major overhauls and replacement of parts. Be aware of what you are buying. Few, if any sellers can tell you what the lowest vacuum pressure at the pumps head is currently capable of.

The mechanical pump can usually only reduce the chamber pressure to 1/100,000ths of an atmosphere or to about 10-40 microns. (the more modern term is millitorr). This ia perfect range for demo fusors. This is roughly equivalent to the atmospheric pressure found 50-60 miles above the earth. This type of pump in a modern, complex vacuum system is refered to as a "roughing" or "foreline" pump.

What do I look for?

You will need to get at least a 2-4 CFM, (cubic foot per minute), pump with a 5 or more CFM unit being better. There is little need for more than a 15 CFM pump.

This completes the quick rinse on mechanical pumps. While a mechanical pump is a minimal "must have" to play in this game, there is a second pump for those who are serious about going on to do real fusion work..........

A Caveat....................

Real fusion work CAN BE DONE with only a mechanical pump... However, that pump must be PERFECT and fully capable of single digit micron pumping. Almost no used pump is capable of this and none of the cheaper air conditioning repair, direct drive pumps can do this even when new. If you do have a pump that can do this extremely deep pumping, fusion is possible. However, you will be wasting a lot of your valuable deuterium gas as such a system will have to run constantly and a lot of gas will be wasted to bring the fusor's gas atmosphere up to over 90% deuterium. This is needed inorder to do fusion. You are on your own here.

The second pump that most all successful fusioneers have is a diffusion pump.......

#2 The Diffusion pump

The diffusion pump or "diff pump" is an add-on or "following pump" to the mechanical pump. This pump uses high speed jets of special oil to blast the last of the resistant air molecules out of this pumps exhast port into the inlet of the mechanical or "roughing" pump so that they can be passed out of the system, at last.

Pressures attained at the inlet of the diff pump can be as low as 10e-7 torr or about 1,000 to 10,000 times deeper that just the mechanical pump can do alone. This amounts to a pressure in the chamber of about that found 180 miles above the earth which is considered outer space. This pump is referred to as the "high vacuum pump"

Note**** In systems with a diff pump, you will need two high quality valves. One to separate and isolate the mechanical pumps inlet with the diff pump's outlet and another valve between your fusor chamber and the inlet of the diff pump.

Diffusion pumps can be found very cheap, considering what they do. Most all must be purchased used. They can be found on e-bay or at Lab-X auction sites. Some few vacuum companies like Dunaway can supply used pumps that have been reconditioned and cleaned with many new parts in them. On e-bay, expect to pay between $25.00 and $100.00. A rebuilt might cost $300.00. You will need to buy the special diffusion pump oil. Silicone oils made for such pumps can cost about $30.00 for 100milliliters. (Dunaway, Lesker, etc.

What do I look for?

You need to look for a pump between 2" and 4" inlet diameter. You really want an air cooled version, with a fan on it, otherwise, you will be forced into a more common water cooled version where you will have to have a source of running water and a drain.

The ideal is about a 2.5 inch, air-cooled unit.

Your fusion chamber will need some sort of valve between it and the intake of the diff pump. Mating the two with a valve in between can create a bit of a headache. This can often be a daunting task, but if you locate the pump and valve before you build your chamber, the job is a lot easier as you just design the chamber with the proper mating flange to the valve and the valve interfacing flange to the diff pump.

By using a diff pump you will be able to immediately begin to do fusion once you have reached 10-5 or 10e-6 torr since at these pressures, admitting deuterium to raise the pressure to about 10e-2 torr for real fusion work you have a chamber filled with 99%+ deuterium.

Note** There is a pump called a turbo-molecular pump that can be used in place of the diff pump, but few amateurs have them as they are prohibitively expensive. New, they can cost over $3,000.00 with controller. Most offered for sale used are only the pump with no controller. Be very careful here. I suggest you pick up a diff pump rather than burden yourself with possible "error buying" of a "turbo pump" right off at the beginning of your fusion quest.

I will let this conclude this quick rinse buyer's guide to vacuum pumps used by the amateur fusioneer.
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
George Schmermund
Posts: 262
Joined: Tue Aug 15, 2006 8:51 pm
Real name: George Schmermund
Location: Carlsbad, CA

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by George Schmermund »

Anything obvious in high vacuum is probably wrong.
adrian.f.h
Posts: 39
Joined: Tue Feb 17, 2009 1:04 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by adrian.f.h »

Hello
The pump on the link looks very similar to mine. I don't know if there are big differences between the different types of these cheepo-pumps. Anyway my pump looks like that and caused lots of trouble with oil vapour (boiling quite fast). It needs to run hot before it can pump down to a demo-pressure-level and this takes about 10-15 min. I always turn the pump off not long after that because of being afraid of it to die from overheat and it contaminating my chamber what happened quite often in the past. Mine works for demoplasma but it's hardly suitable as backing pump.
I’m still working on a phase converter to run one of my other pumps.

Adrian
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by John Futter »

Nice FAQ Richard

maybe you should add that a minimum of a two stage baking pump is required

this in light of the previous two replies to this thread.

Air-conditioning vacuum pumps are usually single stage and the pumping speed is unimportant due to pumping long lines of small diameter. All these pumps have to do is get below the vapour pressure of water. To back a turbo or diffusion pump these would have to be in top condition and even then a diffusion pump of 2" or above will present a load greater than the backing pump of this type is capable of supplying especially the high gas loads presented by a fusor.

Any Welch duoseal (slow rotating) or Edwards {{varian}high speed} two stage pump would be ideal with Leybold tristage pumps {high speed} being the pinnacle of backing pump requirements for fusor. The tristage pumps reach 1 by ten to the minus 4 millibar ultimate vacuum in new condition which would give a high pumping speed in the 1-2 millibar vacuum range as used in a typical fusor.

As for the diffusion pumps
there are two types commonly referred to as either diff pumps or diff stacks

Its easy to tell the difference, in diff pumps the pump body is parallel, in diff stacks the body swells towards the top.
In both cases the body will be covered with a water cooling tube if water cooled-- note Richards comments, the water load for cooling is not insubstantial, a cubic meter of water an hour for a 360L/min pump.

Now for a money cruncher
diff stacks work best with Santovac oil charge -- it is very expensive

diff pumps can use Santovac or ordinary silicon oil such as Dow corning 704 , 705 fluid. The silicon fluids are much hardier at high gas loads (read lousy vacuum) than the Santovac which tends to turn into a nasty very hard black mass when mistreated.

The third type of diff pump that maybe had of the likes of ebay is the mercury diffusion pump. Size for size these offer the best pumping speed vs diameter but because of the mercury they should be avoided at all costs. Some mercury pumps can be cleaned and operated on the silicon pump oils albeit at reduced pumping speed. This type of pump usually has a liquid nitrogen trap above and was very common in older mass specs.

To avoid the mercury pump make sure you see the rating plate on the pump giving oil charge and type of oil. If the plate mentions mercury or grams/ oz's of charge then it is likely a mercury type
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Richard Hull »

Thanks for all the comments.

I have added a small blurb on the need for a two stage pump in an ideal system. A number of the more expensive air conditioning vacuum pumps are two stage. Even a fairly used one with a new charge of vacuum oil can hit 15-20 microns in a 6" chamber. I have three 5-7 CFM air conditioner pumps that have little trouble doing this.

The little 69 dollar new pump from harbor freight makes no bones about it. It clearly states that it can be expected to do no better than 75 microns, which is pretty poor. You get what you pay for.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
derekm
Posts: 141
Joined: Thu Dec 06, 2007 11:25 am
Real name: Derek Mitchell
Contact:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by derekm »

May I be presumptious to suggest an addition to the FAQ
-A mention of Critical backing pressure - The pressure at which a Diff pump can start and its typical value e.g. 0.4- 0.9 torr. for a small Diff pump (Diffstack 63/150). Thus a newcomer can get a feel for the range a diff pump operates over compared to a mechanical.


Derek
derekm
Posts: 141
Joined: Thu Dec 06, 2007 11:25 am
Real name: Derek Mitchell
Contact:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by derekm »

John Futter wrote:
>note Richards comments, the water load for cooling is not insubstantial, a cubic >meter of water an hour for a 360L/min pump.
>

John,
Which make of pump is that ... the Edwards Diffstacks seem to quote a much lower figure...
Squidhat
Posts: 30
Joined: Mon Mar 30, 2009 10:13 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Squidhat »

Hey, when using a diff pump, how do you switch from mechanical to diff? Do you have two separate openings that each pump is attached to, and once you've got the pressure you want with your mechanical, open a valve on the diff pump line and close off the mechanical's line and let the diff pump take over?
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Starfire »

Squidhat - the mechanical pump is a fore-line pump - it assists the diff pump and runs in tandem with the diff.

Diffs can not start until the pressure gets low enough.
Squidhat
Posts: 30
Joined: Mon Mar 30, 2009 10:13 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Squidhat »

Oh, they run at the same time! I knew that you needed to have a low enough pressure to let the diff pump start working, I just thought you would turn the mechanical pump off when the diff pump was running. I was going to get a trap for the mechanical pump before, but knowing that they'll be running at the same time definitely solidifies that need. Thanks John!
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by John Futter »

Derek

yes I know the data sheet says less water in litres / min. But the cooler the outer jacket the better it works. We use a closed circuit water system at work and the figure used is what we have found to make these pumps suck really well. Ie inlet temp 8 degrees Celsius for the water and the flow rate at what I said. I know that when the circulation pump needs servicing ie flow rate goes down pumping speed also goes down. Maybe we are pumping at a far greater rate than the pump name plate suggests. I also Know how big a puddle on the floor is formed when the diff pumps are serviced ( disconnected from the system) ie buckets of water for a few tens of seconds.

To put this in perspective we have a few air cooled 50L/sec Diff pumps on mobile pumping stations. These are slow compared to 50L/sec water cooled Diff pumps.

IE the cooler the condensation surface for the pump the happier it is---my thoughts


FWIW
daghike
Posts: 5
Joined: Tue Mar 01, 2011 5:43 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by daghike »

Hello,

I have a problem with a diffusion pump and mech pump that I need some help with.
I read previous posts but couldnt find a similar problem.

the mech pump brings the pressure down to 1*10-2 torr. but when the diffusion pump is on it rises to 5*10-2torr.

the diff pump is cooled with a chiller and is sufficient
mech pump oil was just replaced. and the diff pump temps are not that high ~200F for the diff oil to degrade.

would be grateful for some advice on the potential culprits.
- the christmas tree is faulty and the diff pump vapor process is not working as it should?

i have elimnated a lot of causes by now and it really has to be something with the diff pump oil vapors not going up the tree as it should but going up from the sides ..

Thanks for your time
Don
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Richard Hester »

This is true up to a point. If the outside walls of the diff pump are too cool, the oil gets too viscous to return quicky to the heater portion and do its job after condensing out on the pump walls. This is especially a problem with the the more viscous silicone oils like DC-705. Using a chiller sounds too much over the top to me. O'Hanlon mentions an optimum wall/inlet water temp. If I remember correctly, it's somewhere in the vicinity of 40C. I'll look it up tonight if I remember and have the time, and if someone else doesn't beat me to it.

I just had another look, and if your diff pump heater temp is only 200F, the pump won't do anything of note. Proper operating temps are more like 200-220C.
daghike
Posts: 5
Joined: Tue Mar 01, 2011 5:43 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by daghike »

Thanks Richard.

yes the temps are ~ 230F.

the pressure rises when it heats up and didnt go back down till its cooled down.

this leaves the tree and internal pathway of the vapor and condensate not behaving properly.

is it possible it could also be a leak? (ive checked by spraying methanol in most places not too close to the heater)

im pretty stumped on this but will keep at it till its overcome.

Thanks for your help. will post what i find.. if theres any more suggestions they are welcome !
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Richard Hull »

The other Richard is correct 200 + degrees C not F is needed.

Check your voltage on the stack. Many are 220 volt. If you are feeding 120 into a heater made for 220 then you will never hit operating temps. I operate a 220 volt stack

One characteristic of all pumps is that the pressure tends to rise against the mechanical pump as low temp volatiles vaporize as the stack heats. It is always minor, but a rise nonetheless. Finally, once the stack's boiler nears operational temps near 120-150 deg C the pressure plunges rapidly and deeply. My system with Dow silicone 704 runs at about 195 deg C. My normal back pressure yield by the mechanical pump is about 20 microns ( 2X10e-2 torr) and the diff pump drops the chamber down to the 10e-5 torr range or ~ 3 orders of magnitude.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
daghike
Posts: 5
Joined: Tue Mar 01, 2011 5:43 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by daghike »

Thanks Richard *2

a few things to try. I have fresh mech pump oil and mech pump by itself was pumping down fine to its usual limits.
the voltages are ok. this system has worked before so its definitely something with the oil or oil levels or leaks ( although leak tests didnt show anything ).

it heats up to ~230F (110C) and for this pump it has worked well before down to 10-8torr using santovac 5. (other diff pumps definitely contributed)

i do hear a popping sort of noise inside the diffusion pump chamber .. and i have not heard it before so this likely related to the increase to ~ 5*10-2 torr seen when pump is hot.

any experience with a similar situ?

i will share what happens with this.

Thanks again
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Chris Bradley »

Richard,

Just to, perhaps, update this FAQ given recent availability of some parts...

It was once the case that turbo-molecular pumps were as rare as hen's teeth on ebay, but they are becoming more readily available. Several folks here have now acquire one, or even a few, off ebay.

Today we might see the more modern turbo-molecular pumps in combination with molecular-drag pumps coming up as 2nd hand deals with the frequency that turbo-molecular pumps once used to (i.e. very infrequently!).

The difference is that these newer pumps have an 'additional' pump, Holweck or Gaede types, which are types of scroll pump, that sit on the same axis as the turbine blades of a turbo pump and act to pump gases at pressures just at the point between 'molecular' and 'viscous' flow which reduces the necessary size of the backing pump. These pumps will start 'pumping' at higher pressures than 'straight' turbo-molecular pumps (that require lower than 100 microns before they 'bite'). There is also a consequent increase in the permissible backing pump pressure - sufficiently so that the best diaphragm pumps can now back a good turbo-drag pump.

What this means for the amateur is that in future it may not necessarily be the case that the only backing pumps are clunking great two-stager rotary types. Some suppliers offer desk-top 'mini-vacuum stations' that consist of exactly such parts - a small diaphragm pump that backs a turbo-drag pump - and can pull high vacuum straight out to atm, all there right on your desk!

So this might be expected to be a good 2nd hand buy in the future for the amateur, but what about now? Well, you *can* get these things if you keep your eye out. I have acquired a KNF Neuberger 4 stage diaphragm pump that can pull 500 microns that does a reasonable job at backing my Varian V70. It is not brilliant, you loose out on outright compression ratio performance, but is just fine for pulling into the sub-micron range.

I will posit a speculation on what could be done with such a set-up: Low power and quieter operation is not the biggest advantage that I think might be realised by the amateur, with such pumps. These diaphragm pumps have a specific purpose in life – all the internal parts that the process fluids flow by are usually all-viton, or some other choice of vacuum and chemical compatibility specific to the application. The pumps are designed to pump chemicals without contaminating them, and recompress them. I think it is not inconceivable that a fusor gas circuit could be constructed consisting of a diaphragm pump backing a turbo-drag pump that recompresses the evacuated gasses back into a reservoir. That resovoir could then be used to feed the fusor. So, once the reservoir has been charged with deuterium, the system can be sealed and operated without any further admit of deuterium. Of course, over time the reservoir of deuterium will become contaminated with the outgassings from the chamber and system, and some will no doubt leak. But in principle I think it could work as a fully sealed system for as long as one's skill at achieving joint sealing and chamber cleanliness will permit, and it would surely be better than just pumping the D straight to outside!
ggombert
Posts: 17
Joined: Sun Apr 04, 2004 2:15 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by ggombert »

Hi,
Attached is a picture of a small air cooled diffusion pump system that I put together over the last year and a half with most of the ‘big ticket’ items from EBay, the rest of the vacuum fittings are from Lesker’s. This is an illustration of what Richard discusses in his FAQ and what can be done with a modest investment of time and a little net ‘bargain hunting’.

The two stage 6CFM pump will pump down the entire system to ~18mTorr and the diffusion pump once turned on will bottom out the high vacuum gage at 10-5 in about three minutes. I have a Varian Cold Cathode gauge as well but do not have all the fittings (yet) to hook it into the system

Glenn
Attachments
Diff Pump System.jpg
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Richard Hester »

Note - Air-cooled diffusion pumps generally need a fan blowing right on them to work properly. The nice little Edwards pump shown in the previous post (I have one too, love the KF input flange) requires a minimum air flow of 2.7 CFM.
User avatar
Alexi Hammond
Posts: 25
Joined: Sat Dec 04, 2004 1:26 pm
Real name: Alexi Hammond

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Alexi Hammond »

Don,

I had encountered similar problem with a little Edwards diff pump. There was a popping sound and no pumping. It happened after mechanical pump oil got sucked into the diff pump. An oil change fixed the problem.

Alex
ggombert
Posts: 17
Joined: Sun Apr 04, 2004 2:15 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by ggombert »

I use a Lesker foreline (metal wool) trap to keep mechanical pump oil out of the Edwards diff pump.

http://www.lesker.com/newweb/traps/fore ... pgid=replc

Glenn
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by John Futter »

Glenn

Richard I hope was not referring to Alexis post
The air cooled pump will still work without the forced air cooling albeit with vastly reduced pumping speed, and by not cooling it, it allows the diff pump oil to rise up into your vacuum system --very difficult to clean off.
ggombert
Posts: 17
Joined: Sun Apr 04, 2004 2:15 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by ggombert »

John,
If you look carefully at the back of the picture that I posted you will see a 2.75 in fan on the back of the diffusion pump.
Glenn
daghike
Posts: 5
Joined: Tue Mar 01, 2011 5:43 pm
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by daghike »

Thanks Alex,

I have gotten down to 2*10-6 since I added 100ml more of oil.. There is definitely a small leak inside that is hindering the pump down. its not a big one thats why its hard to find it.

for ex. when roughing the chamber the pressure was lower than when just backing the diff pump. somehow when the diff pump oil got in motion it overrides this leak somewhat and does the pumping out.. just hitting this ceiling.

will do one more round of leak testing and replace it with another diff pump.

cheers,
Don
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: FAQ - Mechanical Pump - Diffusion pump?

Post by Richard Hester »

Noted - folks that have an air-cooled pump with no fan - fix that...
Post Reply

Return to “FAQs: Vacuum Technology”