Need a proximitiy sensor only detect liquid not foam! Any suggestions? Recommendations?

Thanks - neat! :-) See my reply to Kieth.

Ed

Yeah, you beat me to the metered pump.

I once got some left over blood pumps that used a partial circle ""cage"" and 3 rollers squeezing the outside of silicone rubber hose.

I never checked but I suspected those pumps could have been easily metered.

If you have a known pump on a regulated power supply with a fluid of known viscosity pumping no air, couldn't you just fine tine the TIMING?

You'd STILL be double checking fill with an industrial pass/fail scale on the conveyor line, right?

I used to pump NaK that way ;-) ...Jim Thompson

"Tanks"!

Peristaltic - yes, easily metered. Just count turns (and calibration check occasionally in case the tube develops a bit of set). I have a somewhat elderly one I use for pumping wort and beer (got tired of the whole siphoning business - I enjoy it with water, I hate it with beer that's got plenty of dissolved CO2 in it, which happily un-dissolves in the high part and breaks the siphon. Even with wort, there's the issue of starting it cleanly.) Someday when my copious spare time opens up I'm going to deal with the serial interface to the microprocessor on the pump so I can have it dispense (fill this size bottle) rather than standing there pushing buttons (without a computer to talk to, the only option with this model). The manufacturer has not nicely thrown the outdated software for that out to the world (no doubt afraid someone would hold off on buying a new pump), and it's too much trouble to track down a copy - but I do have a manual with the command language specified.

With those constraints, i suggest a well smoothed pressure sensor at a known level in the tank. There are techniques that allow for recurrent auto zero of the pressure sensor over the long term.

Hmm, interesting. All the ones I can think of involve a couple of solenoid valves: one connecting the two sides of the pressure sensor and one from the outside of the sensor to ambient. That would involve either pumping the overflow back into the tank, or dumping at least a few CCs onto the floor each time.

Cheers

Phil Hobbs

for

sump

=20

I think that there are a few valves, but the cal sequence involves blowing some small bubbles instead.

Given that fluid pressure is hydrostatic, blowing bubbles doesn't change the pressure. I can imagine calibrating the scale factor by blowing a bubble, shutting off the valves and then adding more air, but I don't see how you can do an autozero that way.

Cheers

Phil Hobbs

even put it in the food.

Don't most aquaria have a bubbler? The air pressure in that little tube will be rho*g*h, and you can put a tee fitting and pipe the pressure almost anywhere. A U-tube manometer, with capacitive or other readout, is easy to build.

Think BIG! Think SUMP and where the pumps are located. Aquarium level itself is determined by a "spillway". ...Jim Thompson

for

that

sump

and=20

=20

=20

Can't argue that. But the system is use in long life water level detection and does work. I will have to see if i can get copy of a manual. That should explain how it works and manages calabration/autozero type functions.

I'd be interested, if they have a non-obvious method.

Cheers

Phil Hobbs

Alas, the simplest way is likely to be (1) a PVC pipe section with some holes drilled in it (to make a no-turbulence-here zone) with (2) a toroidal float that contains a magnet, trapped on (3) a tubular shaft with reed switch(es) inside. A bit of vinyl tubing can protect the switches and their wiring from any corrosive effluvia.

More elaborate, a capacitive dipstick with CD4046 to oscillate it and generate a voltage output (phase-lock the 4046 to a fixed oscillator and look at the control voltage) is a candidate only if you trust the surface of the capacitive dipstick not to grow a conducting film (like, PTFE or silicone-oiled-glass).