Yes, very practical.

Ken

 

Did some one say 555: Delay variable, 18V a little high for 555, so add a white LED between +18V & pins 4 &8.

 

Instead of relying on the iso output to produce timing current.

What if R3,C1 timer were replaced with a one shot circuit, that could have its timing hardwired and variable through a pot?

 

But, that's not the problem that I now see. If you were to try my common-emitter arrangement (just the 10K load resistor...ignore the delay part) for the opto-coupler, what collector voltages do you see when the cylinder is present at minimum and maximum distances and absent?

Ken

Ken,

What is the appropriate way to test for collector voltages? I've got the opto sensor isolated (power to the LED) with just and open connection from battery positive to the collector and then measuring the voltage potential from the emitter to ground.

All it reads is battery voltage no matter how close an object is to the sensor

Thanks,
-Jacob

 

Put a 10K resistor from the collector to battery positive. Like my circuit, to the left of the diode. Measure the voltage from collector to emitter. Do this under the various positions you expect the cylinders have in relation to the detector. I noticed that your detector's focus point is 5mm from the lenses. Is that were your cylinder's surface will be? Ideally you will see the voltage change from near the battery voltage level to near zero as the cylinder pass the detector.

Ken

 

Ken,

Sorry it took me so long to get back to you. I hooked up a 10k resistor to the collector (per the diagram) and the voltage goes from battery voltage and drops down to about 12V at the focus point. (If I actually touch the opto sensor the voltage will drop all the way to to 7-8V).

If it would make it easier at point for you to get your hands on one of the actual sensors I am using, I wouldn't mind paying a deposit or something like that.

Thanks again for your help so far,
-Jacob

 

Actually, I rigged up a Radio Shack IR LED (T1-1/2 @20mA) and phototransistor (T1-1/2) mounted parallel in a reflector configuration. It worked, with the delay, with a change from 18V to ~14V.

A lot is going to depend on the diameter, the surface, the speed, and distance of the cylinder. Can you spell those out?

If I were designing your circuit I would:
1. Pulse the LED with a fast, high-current, low-duty cycle oscillator to increase the range.
2. Put the output of the optocoupler into an adjustable comparator (LM311) to get better defined switching.
3. Put the output of the comparator into a monostable like mine (because of the 18V). Or a 7555 like Bill and Bernard suggested.

Or....maybe just an 8-pin microcontroller, a voltage regulator and a couple of MOSFETs. Oh yes, and some resistors and capacitors.

So many ways to skin a cat.

Ken

 

A lot is going to depend on the diameter, the surface, the speed, and distance of the cylinder. Can you spell those out?

Ken,

The cylinder is about 2-3" in diameter (it's got a taper in it) The object moving inside the cylinder is about 1.5" in diameter with a shiny white surface. At the trigger point for the opto sensor the object could fall anywhere between 0-0.5" away from the sensor. The motor free spins the auger mechanism at about 18,000 rpm. Loaded I would guesstimate (from the motor curve) that is drops to about 12,000-15,000 rpm.

Also, I was thinking that as long as we could control the switching distance for the opto sensor, I could adjust it so that when the auger mechanism comes to rest at the back half of the cylinder it will not reactivate the sensor.

Thanks,
-Jacob

 

OKKKKKKKKKKKK. I can't visualize what you just described. Maybe a sketch would help me a little. Or, maybe I just need a little more coffee this morning.

Ken

 

Ken,

I attached some rough sketchs for this part of my prototype. Hopefully it makes things a littler clearer.

Essentially the opto sensor triggers the motor when an obect falls into the "funnel". The motor has a thin strip attached to it. When the motor spins this strip around, it contacts the object and spins it back out of the top of the funnel.

So as long as I could adjust the switching point of the opt sensor so that is not activated by the "auger" in a stationary postion or the far wall of the funnel, it should work okay. I had requested delay power for the motor as I'm not sure if the opto sensor would continue to register the object in front of it with the motor spinning it around at such high rpms. On my bread board setup there was a little delay in the sensor picking up objects passing in front of it so I wasn't sure if it would be a problem.

The only issue I can see if if the auger itself just happened to come to rest directlying in front of the opto sensor and switched the motor back on. However, the auger is a 2-D strip that is only 0.030" or so in the profile dimension that the sensor would be looking at. I'm hoping it won't be a prohibitive issue.

Let me know if you have any more questions Ken,

Thanks again,
-Jacob

 

Always more questions. Have you tried the actual mechanics of the funnel/ball/auger/motor operation? How often does the ball enter the funnel? What's the spin-down time, assuming a spin-up/run time of a second or two? What's the ambient light in the environment? Have you thought of an optical beam-break configuration just above the top of the funnel/auger-tip. Or, do you need to sense that the ball has settled into the bottom of the funnel?

And...what's the goal of this whole project?

Ken

 

Questions are no problem

I have tested the setup with a push button switch to control the motor and it does work for what I need it to do. The ball could enter the funnel every 30-40 seconds during operation. Spin down is under a second.

I don't have to sense that the ball has settle per say but for the fit/finish of this application a self contained reflective sensor would be the most convenient. I also don't have much room for adding additional components outside of the current OD of the prototype.

This project could be used indoors or outdoors. The opto sensor could be exposed to sunlight (although not pointed directly at it). It would be exposed to regular ambient light in a typically lit room if used indoors or exposed to natural ambient light if used outdoors. These parts are not covered by anything. The top of the funnel is open to the "atmosphere".

I'm just working on a little gimmicky toy and automating the control for returning the ball from the funnel is where I'm stuck.

Thanks!

 

The opto sensor could be exposed to sunlight (although not pointed directly at it). It would be exposed to regular ambient light in a typically lit room if used indoors or exposed to natural ambient light if used outdoors

This will be a problem. The circuit ideas so far do not account for a variety of ambient light levels.

Ken

 

A beam break would help to aleviate stray light & distance sensitivity, also vane stopping in beam path probably no problem. I like K Moffett #27 suggestion of using comparator to allow easy sensitivity adj & drive to delay cap.or input to 555. Nice schetches. now I know how to return ping-pong balls to automatic server.

 

This will be a problem. The circuit ideas so far do not account for a variety of ambient light levels.

Ken

I have seen some of the reflective sensors that have an ambient light filter on them, would this help the situation?

All in all as it sits, I wouldn't be opposed to trying a circuit setup as we have been discussing to see just how much the changing light conditions could affect things.

-Jacob

 

It might not be a problem if the auger stops in front of the sensor. The motor would just start up and the rotor would (hopefully) stop again somewhere else. Or you could start the motor slowly, and if that clears the sensor (i.e. it was the auger being sensed) then stop it. Otherwise, spin up to toss the ball out.

You can always use an LED with a chopped drive, and a sensor that looks for that frequency, or even that frequency in phase with the LED excitation. That rejects ambient light totally.

 

John P,

I agree, a chopped light source would be best...high ambient-light rejection and higher drive current. I'm still not comfortable with the vane not blocking the LED or the detector. I think I would put two LED/detector pairs are right angles and AND the outputs. Then only the ball could block both. But then, maybe the stopping position may random enough to work.

Ken