Need help w/level sensor

The photo thing should work if you have clean open/dark gaps. Maybe you have hum or something fooling the comparator. Check it with an oscilloscope. Post a schematic if you can.

People often weigh hoppers with load cells.

Capacitive sensing also comes to mind.

There are some commercial vibrating-rod point sensors that work with granular solids or really yukky liquids.

John

No hum, clean supply (I even tried it with a battery).

Schematic is at URL mentioned in post.

Yes they do. This would have added about another $5000 to the cost (gravimetric vs volumetric feeding). Would have also added problems since we have a rigid hookup between hopper/feeder and the rest of the processing equipment; it's pressurized and at high temps so we use a stainless steel tube.

Yup, though of those. Don't know how well cap sensing would work though. This is very dry material. If I can find an easy-to-apply circuit, I might give this a shot. I don't think the "tuning fork" solution will work either because of the nature of the material (density, ability to pack and bridge, etc.). I was thinking of trying this out with a vibrating piezo disk and a PLL circuit. Any material around the oscillating piezo would de-tune its resonance and "trip" the PLL. I've consulted with Omega and other level sensing "experts" and they've all thrown up their hands.

Thanks for your ideas though. This stuff is a real booger to work with.

Have you checked the comparator outputs with a scope? The only way a

339 will make the output led go dim is if it's oscillating.

I'd prefer lower impedances and a Schmitt trigger instead of the comparators.

It's a little weird that you are offering the public permission to use a circuit that doesn't work.

John

Now that last comment just begs for a cynical metaphor about the political arena, but I'm gonna keep my mouth shut (to a point).

--
www.wescottdesign.com

Hum can happen from other sources than the power supply. Are you sure it isn't ambient light?

Messing around with the transistor pull up is probably a good idea. I'd be wary of a circuit that tries to work at DC -- it's just asking for ambient light trouble. If it works, great. If it doesn't, then you may want to consider modulating the source and AC coupling the receiver -- but only after you've verified that what you have is simple enough, but not too simple.

Do something to make sure the powder doesn't clog up your emitter and/or sensor windows...

--
www.wescottdesign.com

So invert the polarity of the return. When there's no material, the acoustic impedance match will return a stronger signal; when there's material, the return will be damped, and the signal will decrease.

Good Luck! Rich

What about a float? If the machine vibrates enough to keep the powder kinda liquid-y, and if you can get something light enough (and maybe if you don't feed right on top of the float), a toilet-style float should rise to the surface of the powder.

Dunno if all those 'ifs' are satisfied, of course.

--
www.wescottdesign.com

There are also level sensors that have vibrating bit wot sticks into the material being measured, and gets damped.

Eg.

And similarly, gently rotating paddle wheels wot get stalled by the material.

I take it you just need a level switch rather than a level measurement?

Somehow this reminds me of "McLintock," where John Wayne says "Somebody should punch you in the face. But I won't ... no I won't ... the HELL I won't!" ;-)

Cheers! Rich

This looks interesting. We need some way of measuring what's in the hopper. It can be a continuous indication, or a series of switches. Something simple would be best, but *reliable* is critical (it's a pressurized hopper and hard to work on) and accurate is important.

OK, I'll check out the output.

Thia also makes sense, but this looked like a simple circuit, and the comparator *should* switch ON/OFF. The photoxtr and LED are mounted on opposite sides of a 2" square piece of aluminum C channel, and I covered the other sides with cardboard just in case the transistor is picking up the ambient light from the overhead fluorescents.

I'm not offering anything. Except a request for help (is that an offer?). I don't know if I used the same exact parts the person that designed the circuit in the original URL used. But the one I built

*should* work if I can find the magic words and hold my mouth just right while standing on one leg in the light of a new moon.

d

may

Yeah, I originally thought of modulating the IR LED and using one of those remote control pick up modules. I went for cheap and easy (I thought) first.

The stuff I'm working with is terrible from a dispensing point. It has static (it's very low moisture) so it clings, it bridges, it has a high slump angle, it packs, and probably a bunch of other bad habits we haven't seen yet. The only *good* thing is that we process the material we get and remove the fines. *That* stuff is even WORSE. Dust explosion waiting to happen!

The idea is that we don't want to run out. We won't get a return until the hopper is empty and parts that reflect are exposed.

We thought of that. Thought of using something like ping pong balls that would ride on the surface. I don't know if they would migrate back to the surface, or how long it would take if they do, when we refill the hopper. Something to try if I can't get this optical sensor to work.

No, no, tell us all about it.

John

A picture would surely help (if it's not violating NDAs or the like). Sounds like you're filling from the bottom?

| | | | |,.,.,.|

"[O]ffering the public [option] + "doesn't work"

You do the math. ;-)

Cheers! Rich

Most likely it is the open/dark gap is not clean because the powder coats the surfaces - and maybe if the gap is too small, it clumps and thus totally blocks the gap. Check the clumping first; if true, a larger gap is necessary (say 0.1 to 0.2 inches). If it is a "mere" case of powder coating the surfaces, then a more intense drive (pulsing may be needed to limit power and heating) and perhaps a bridge-like voltage level sensing is needed.

How about being crazy and try an optical fiber where light in the fiber "leaks"? Use it as a leaky transmission line / antenna; a light pulse travels along the fiber and where there is product, some of the emitted light bounces back into the fiber and is seen in a reflected wave at the appropriate time, similar to radar. One word of caution: most optical fibers used for this purpose slowly darken due to hydrogen absorption; there is supposedly a new fiber (and/or system) more resistance to that problem. Then again...why not use old-fashioned twinlead (terminated in 300 ohms) and a pulse of RF? TDR here should be usable...