Vacuum pump

[Twospoons]

One could probably  employ the same technique as used in wet-bulb humidity sensors, and soak the liquid into some kind of wick. That would give very high surface area, remove the hydrostatic pressure issue, and probably break bubbles too.

[davekni]

At micron pressure levels, I don't think any bubble boiling is feasible.  I think it will be all surface evaporation/sublimation.  Mercury vapor pressure a bit above room temperature could be used, though with the hazard of mercury.  Ice at -50C or so would be reasonable, but not easy to make or know temperature accurately.

[Alberto]

Thanks for your answers

[Alberto]

Hello and happy new year! (sorry my english)

One final question. I have found a good one (15 microns advertised, probably 25 at real), but it is only 1,8 CFM (51 liters per minute). I usually see in the youtube experiment videos 4 or 5 CFM, so I don´t know if 1,8 CFM it is not enought. I mean, for small experiments with plasma and small volumes.

If i understand well at atmosferic presure that means that move 1,8 CFM, but at the moment that it starts making vacuum the rate goes down and less CFM means I´ll need more toime to get the final vacuum right?

so 1,8 CFM is OK?

thank you for all your help and advice

[davekni]

One final question. I have found a good one (15 microns advertised, probably 25 at real), but it is only 1,8 CFM (51 liters per minute). I usually see in the youtube experiment videos 4 or 5 CFM, so I don´t know if 1,8 CFM it is not enought. I mean, for small experiments with plasma and small volumes.

I'd guess 1,8 CFM would be fine for your experiments.  I think my pump is 2,5 CFM.  Might take a bit longer to wait for outgassing of experiment containers to drop low enough.

If i understand well at atmosferic presure that means that move 1,8 CFM, but at the moment that it starts making vacuum the rate goes down and less CFM means I´ll need more toime to get the final vacuum right?

Yes, AFAIK that is correct.  Gasses (ie tanks of compressed gas) are usually measured in the volume they would expand to at atmospheric pressure.  Same for vacuum pumps, I think.   At 1/2 atmospheric pressure, I think the pump will remove roughly 1,8 CFM of that pressure, which is 0,9 CFM at atmospheric pressure.  In other words, a 1,8 CFM vacuum pump connected to a 1,8 CF vacuum container would reduce pressure exponentially, to a factor of 1/e every minute.  Of course, this ignores outgassing or any other source of gas being added to the container.  And also of course, as pressure gets lower, approaching ultimate pump capability (ie. 25 microns), gas removal rate will decrease.

[Alberto]

One final question. I have found a good one (15 microns advertised, probably 25 at real), but it is only 1,8 CFM (51 liters per minute). I usually see in the youtube experiment videos 4 or 5 CFM, so I don´t know if 1,8 CFM it is not enought. I mean, for small experiments with plasma and small volumes.

I'd guess 1,8 CFM would be fine for your experiments.  I think my pump is 2,5 CFM.  Might take a bit longer to wait for outgassing of experiment containers to drop low enough.

If i understand well at atmosferic presure that means that move 1,8 CFM, but at the moment that it starts making vacuum the rate goes down and less CFM means I´ll need more toime to get the final vacuum right?

Yes, AFAIK that is correct.  Gasses (ie tanks of compressed gas) are usually measured in the volume they would expand to at atmospheric pressure.  Same for vacuum pumps, I think.   At 1/2 atmospheric pressure, I think the pump will remove roughly 1,8 CFM of that pressure, which is 0,9 CFM at atmospheric pressure.  In other words, a 1,8 CFM vacuum pump connected to a 1,8 CF vacuum container would reduce pressure exponentially, to a factor of 1/e every minute.  Of course, this ignores outgassing or any other source of gas being added to the container.  And also of course, as pressure gets lower, approaching ultimate pump capability (ie. 25 microns), gas removal rate will decrease.

Thank you!

For example, how much it last for your pump get the final vacuum? For example 0,25 liters

[klugesmith]

Don't overlook the vacuum conductance of hose between chamber and pump.
It's measured in same units as pump speed (CFM, l/s, l/minute, etc.).
A 1.8 CFM (0.85 l/s) hose going to perfect vacuum would reduce chamber pressure just as fast as 1.8 CFM pump with no hose.
With hose and pump in series, you have a 0.9 CFM system.
Hose conductance varies with pressure, but is constant at the low end of our range of interest here.
I think 1.8 CFM is about 1 meter of 3/4 inch hose, and less than 2 inches of 1/4 inch hose.

There are simple formulas online.  Here's a picture from one reference
 https://www.vacuumscienceworld.com/blog/conductance-influence-in-vacuum-systems

[davekni]

There are simple formulas online.  Here's a picture from one reference
 https://www.vacuumscienceworld.com/blog/conductance-influence-in-vacuum-systems

Great point about tubing, and good reference too!  I tend to underestimate effect of tubing when doing vacuum experiments.

For example, how much it last for your pump get the final vacuum? For example 0,25 liters

Unfortunately, there is a simple calculation, and then there is reality.  Simple exponential calculation may be good to 1 mbar or so (760 microns).
A 1.8 CFM pump is 0.85 liters/second.  To make math simple, presume that tubing reduces this to 0.25 liters/second.  Then pressure would reduce by 1/e every second.  Pressure = initial_pressure * exp(-time_in_seconds).  After 7 seconds, pressure would drop by exp(-7), from 1 bar to 1 mbar.  If pump and system were ideal, pressure would reduce by another factor of 1000 over the next 7 seconds to 1 ubar (0.76 microns).
In reality, as you drop below 1 mbar, outgassing (from vacuum chamber walls, tubing, even pump internals) can add gas almost as fast as pump is removing gas.  Speed becomes dependent on rate of outgassing.  There may be ways to estimate that too, but depends strongly on material properties.  (I've also found out the hard way that many types of tubing are not good for vacuum use.  Especially typical silicone rubber tubes.  Air diffuses rapidly through tube walls.)  Getting from 1 bar to 1 mbar is much faster than getting from 1 mbar to 32 ubar (25 microns).

[jpsmith123]

Before I bought my Adixen/Pfeiffer vacuum pump (2005SD), I experimented with three or four different varieties of cheap air conditioner service type two stage rotary vane pumps, including the 3 cfm model from Harbor Freight.

I ended up returning or selling them all and saving up some more money until a good deal eventually came along on a high end pump on ebay.

Nowadays I see people selling some cheap pumps on ebay with an advertised ultimate vacuum spec. of 0.3 Pa (which would be about 2.25 millitorr) such as the following linked one.

https://www.ebay.com/itm/293778324267

I don't know if this particular pump will actually achieve the advertised ultimate vacuum, but one of the cheap pumps I bought many years ago actually buried my Varian TC gauge just like the Adixen/Pfeiffer pump does (whereas the Harbor Freight pump bottomed out at around 25 millitorr IIRC). The particular pump I had which did that was made in Argentina and it seemed well made except for the horrible mechanical design of the inlet and exhaust ports, which is why I sold it.

What I would look for today in a cheap pump would be: (1) a good ultimate vacuum spec. (I would consider 0.3 Pa to be good if it's real); (2) a pumping speed of at least a few CFM; (3) inlet and exhaust ports which have enough open area around them so that they could be adapted to connect to standard vacuum hardware e.g. KF25 flanged stuff; and (4) (if possible) gas ballast.

(In my experience, one good thing about ebay is that if you buy that vacuum pump and you find the ultimate vacuum isn't anywhere near the advertised spec. of 0.3 Pa you can return it, usually free of charge).

BTW there are a few places selling relatively inexpensive Chinese made vacuum hardware such as stainless steel tees, crosses, hoses, nipples, valves, foreline traps, etc., with KF flanges, so if you can find a pump whose inlet port is 1/4" NPT for example (and there is enough room around it), you can get an adapter like the following linked one and this will generally make it easier to use the pump.

https://www.ebay.com/itm/224801525604

[Alberto]

Thanks for all your help!!!

So, 0,3 pascals are achievable with a 2 stage pump?

Thank you

[jpsmith123]

So, 0,3 pascals are achievable with a 2 stage pump?

With high end pumps that are in good shape, no doubt. And some of the cheap AC service pumps seem to be capable of it as well. In my experience I had one cheap pump that was able to get down to 1 millitorr or better (measured right at the pump's inlet port with nothing else attached) and one that didn't meet the advertised spec. which I returned.

I have a Varian #801 thermocouple gauge controller with a #531 thermocouple tube.

The #801 looks like this:

https://www.recycledgoods.com/varian-801-thermocouple-vacuum-pressure-gauge-tube-control/

You can often find them for sale (used) on ebay and elsewhere fairly inexpensively. The type 531 TC tubes can also often be found on ebay, new and used, often at a good price. Here is a spec. sheet on the tube:

https://www.agilent.com/cs/library/datasheets/public/Copy%20of%20531%20and%20536%20Thermocouple%20Gauge%20Tubes%20Data%20Sheet.pdf

And here's a publication you may find useful: "Vacuum and Leak Detection Reference Formulas, Properties, and Glossary Training Guide":

https://www.agilent.com/cs/library/brochures/brochure-vacuum-leak-detection-reference-formulas-properties-glossary-5994-5366en-agilent.pdf

Once I got the Varian #801, I wanted to put it in some kind of enclosure. I ended up buying one from Duniway:

https://www.duniway.com/part/tcg-box

If you acquire/build a TC gauge (or some other type of vacuum measurement device) first, then you can use it to test the performance (ultimate vacuum) of any pump you buy and then return the pump if it's not at least somewhere close to the advertised spec.

BTW there might be some info here that may be useful:

http://www.belljar.net/articles.html

[Alberto]

Thank you!

[the_anomaly]

I realize this is an old thread but wanted to throw in my 2c.  I've done a lot of vacuum work mostly in 10-100mTorr but also in <10^-5 Torr using turbo molecular pumps. 

So, 0,3 pascals are achievable with a 2 stage pump?

Same as above, a good quality name brand (pfieffer, leybold, savant, alcatel, trivac, etc.)  All the used pumps I've bought off ebay vary from 2Pa to 13Pa even though their ultimate pressure is listed as less.  I have not had good success replacing the vanes either.  Leybold tech support told me its difficult to rebuild the pumps, usually need to grind the housings etc.  Still 2-13Pa is plenty for experimenting with electrical plasma.

I've found that using push to connect tube fittings can work well for electrodes/tubes when you drill the stop out.  Easy job if you have a lathe or even a mill.  If not, careful use of a drill can get the job done.  An example of a setup is shown here:



At higher pressure:


At lower pressure:


Honestly my cellphone can barely capture the beauty of the striations.

I'm using pure CO2 in a flow through regime.

Also, you can use teflon tape for moderate pressures but I've also found that superglue and various versions of Loctite work well for thread sealants.   

Here is another useful document on level of vacuum:

Vacuum Conversion Table.pdf

You see at the millitorr range how close you are to ultimate vacuum.

[Alberto]

Thank you!

I ended buying a vacuum pumt in a hardware store near my home and the result was great. 15 microns and 2,5CFM for 220€.

The problem is I want to make triodes and CRT and I´ll nedd a higher vacuum that I expect to reach using getters