Here we come to some cool stuff.
To get down to 1 Pascal it's enough to use a mechanical pump.
A good one possibly.
A cheap one will do, but (they say) you should use some good oil if you really want to meet the specs, or the pump will
stop pumping down much before the nominal operating point.
Anyway, you won't go much below 1 Pascal (7.5 mtorr) or so.
To do that you need to add some sort of other device.
The most common choices are diffusion pumps (cheaper) or turbomolecular pumps.
So... currently the larger stainless steel vessel is under construction
(see here).
In the meanwhile, I got a turbomolecular pump off ebay, together with its controller for some 200-300 EUR.
A real bargain.
It's an Edwards ETX255, speed 60000 RPM, a bit more than 200 l/min pumping speed (for N2, from the datasheet).
The controller is an EDXC80, extremely small super simple to use since it allows for a manual operation of the pump (on/off).
An analogous output from 0 to 10 V reports the output returns the actual RPMs (0 to full speed of the pump).
Check out the datasheet for the controller, that's really simple stuff.
Here's the turbomolecular.
It comes with a nice grid that should prevent possible fragments to get into the blades.
Needless to say, these babies are pretty touchy and you should prevent anything from getting into the blades while they are spinning.
Actually, you also MUST avoid sudden increases of pressure in the chamber when they are rotating at full speed.
Their blades are too thin to sustain serious aerodynamic forces!
So, if a window from your vacuum chamber breaks and pressure suddently gets in, you may end up with a mess of a turbomolecular.
Check out google images.
Power supply
First things first, the power supply.
The EDXC80 controller requires from 70 to 85 V, 93 W.
I connected in series 3 LED generators of 25 volts each.
These can actually be slightly tuned, so I can stay a bit below 85 V (say, around 75 or 80 V just to stay safe).
Here's the little guy.
It's so simple that you don't need a schematic for that, anyway there's one below.
Just, a switch gives power to the three generators, whose input is connected in parallel (every input has the network
voltage, 230 V here in Italy).
Their output is connected in series as to reach the 80 volts and goes directly to the pump controller.
Watch out! 80V are not nuts. And be careful with the 230 V...
In the image, you can also see a cheap voltmeter, that will read the pump speed (at least, roughly).
The voltmeter is powered by one of the 25 V generators.
At first, I had connected it randomly to the very last one, but this biased it at +80 V (with respect to the cathode at least)
and that's not good for measuring stuff from the pump!
Indeed, measuring the 80 V would only result in big sparks.
The problem was solved by powering the voltmeter with the first power supply, the one whose negative pin is the "0" for
the "0-80 V" cascade.
The second switch on the right in the image starts the pump by setting the PIN 1 to zero.
Remember to include the grounding! And notice that the pump body should also have a place for a grounding wire.
Connection to the vessel
Turbomoleculars should be connected to the vacuum chamber by use of a flexible joint.
I will use a short KF40 bellow, which although not ideal (connection should be as large as possible) will do the job for now.
Also, one should incorporate a butterfly valve for pressure regulation.
Test
Here's a small video of a switch-on test.
Switched it off immediately not to stess it.
It works! :-)
Next steps: air or water cooling and full vacuum operation!