initial version 08/26/2005
From the Shop: Vacuum Pump
|The pump not assembled.
Unfortunatly only B&W
|The assembled pump|
This pumps yields a vacuum of 0.89 bar (12.9 PSI) at 400m over sea level at 5 bar (72 PSI) input.
The pump has mainly two parts:
It is a modified quick release shop air fitting. The body of the jet is silver soldered to the fitting. The jet has a 2,1mm bore and a cone opening to the vacuum side.
A rotary part (a 22mm hex stock of aluminium did the job) has a vacuum chamber on the inlet side a connection for the vacuum hose and a conical bore (diffusor).
You can download a plan (PDF).
The parts are easy to make on a lathe. Just for the conical bores, you need to make two D-bits out of drill rod. To bore brass or free machining aluminium with them, they don't need to be the very best.
The jet's cone has a slope of 1:10, the diffusor's slope is 1:32. These values aren't to critical.
The connectur for vacuum is a brass tube simply pressed in. You may change this without problems. The exact location of connection is not critical.
I started with some slight modifications with Don's scetch and initialy got 0.85 bar (12.3 PSI). Continuing from there, I tried to optimize and made some observations where it is worth to further tune the pump.
But first, some values I messured with different input pressures (2 bar: 29 PSI; 4 bar: 58 PSI; 6 bar: 87 PSI; 7 bar: 101.5 PSI) and different modifications. Manometer was a class 1.6, location 400m (1312 feet or 215 fathom :-) ) above sea level.
|Unmodified||Mod. 1||Mod. 2||Mod. 3||Mod. 4|
|29 PSI||4.4 PSI||5 PSI||5.2 PSI||5.2 PSI||5.2 PSI|
|58 PSI||11.6 PSI||12.3 PSI||12.3 PSI||12.3 PSI||12.5 PSI|
|87 PSI||12.3 PSI||12.6 PSI||12.8 PSI||12.8 PSI||12.9 PSI|
|101.5 PSI||12.3 PSI||- *)1||- *)1||- *)1||- *)1|
*)1: The pressure regulater had not enough throughput to keep the pressure.
You can see, that the vacuum is more or less constant above 87 PSI. The jet dosn't allow more throughput.
I smothed the intake side (quick release fitting side).
I fine tuned the distance between jet and diffusor inlet (ø 5mm bore). The distance is not not too sensible if it is in the range of 1 mm. If you select it to be to small, no air will be pulled out of the vacuum chamber. If you make it to big, the conical air stream of the jet will hit the corner of the intake (ø 5mm) and will be deflected into the vacuum chamber.
I widened the end of the diffusor with a steeper cone. No change..
I shortened the cylindrical part (ø 5mm) of the diffusor. Initialy it was 40mm long and shortened it to be 20mm. A slight improvment, but adjusting the distance got a bit more sensible. I didn't shorten it more, because I would have had to make a longer D-bit.
You can improve the pump with a smooth, non-turbulent stream into the vacuum chamber. Don't make the cylindrical part of the diffusor to long. I guess that 5 10mm are at about the best.
I made no trials with different cone angles.
Here is a good article about different vacuum pumps, their consstruction and theory, on the net. You have to look around in the "Vacuum Pumps and Chucks" section.