I recently remembered that I had picked up a variety pack of cheap nozzles for a 3D printer a long while ago.  The quality was so low I did not end up using them, but they do have holes that range from 0.2 mm to 1.0 mm in size and so I could compare them to my 3D printed nozzles.

I'll keep this short, they did not work.  I tested them on every compressor I have access to and in every case they were less effective than my 2 mm test nozzle.  The air flow was lower, and the velocity at the tip was also lower.  I think there is so much friction at the outlet that they constricted the air flow beyond what Bernoulli suggests.  However the 1.0 mm nozzle was not too far below from my 2.0 mm nozzle, and my 2.0 mm nozzle was a solid improvement over the 5 mm tube with no nozzle.  This just reinforces what I found earlier with my test nozzles, there is some sort of an optimal nozzle size near 2 mm that maximizes air velocity at the tip.

I suspect we could measure CFM for various nozzle diameters and would find that there is a non linear relationship where CFM falls off rapidly below a particular nozzle diameter.  I will save that experiment for another day.  I did find that the 1.0 mm nozzle produces a more steady flow than the 2.0 mm nozzle when using one of my smaller air pumps.  However I don't think a smooth flow of air is nearly as important as a strong flow of air.

In doing all this testing I noticed that the water separator on my regulator leaks at very low inlet pressures.  It needs 10 or so psi for the auto drain valve at the bottom to fully engage.  Since we don't have any humidity in the air I never end up collecting any water anyway, so I took the valve apart and moved the spring from the top to the bottom so it holds the valve closed by default and you now have to manually push up on the valve to drain.  This seems to work well and allows me to hook my low pressure air pumps to the inlet side without causing an air leak.

Since my operating range for air pressure is between 5-30 psi for cutting the 160 psi gauge on my air regulator was not very accurate.  So I decided to pick up a Winters 0-60 psi 1/8" NTP gauge to replace it with.  This is working great, the gauge starts to register at 3 psi, and appears to be quite accurate.  It is much easier to dial in an exact value.  I don't know that I need this level of control, but it is miles ahead of the mess on my compressor and will let me test more precisely going forward.