How to know when a sawdust collector is full

wapo54001

Senior Member
A friend has asked for help with a problem he has at his commercial woodworking shop. All of his power tools have vacuum hoses attached with appropriate gates that open when a particular tool is turned on. Sawdust created by the tool is vacuumed through a hose system and deposited in a large bin in the basement of the building.

If the sawdust is allowed to accumulate above the top of the container it clogs hoses and requires several hours of work to clear. The friend wants some form of sensor to illuminate a light in his workshop when the bin is perhaps 3/4 full so that someone can go down and empty it into a separate container.

Previously, a system was installed that apparently involved an electric motor with a long shaft with some sort of fan blades at the end of the shaft. When the sawdust reached the level of the blades the motor would be stopped and that somehow illuminated a warning light. This system failed at some point and was removed and discarded.

Is there another method that would work equally well or should I look at the motor solution again? I would program a picaxe to work with whatever sensor is appropriate to illuminate the warning light. I'm not sure if the motor solution used the increased current of a stalled motor, or whether some sort of pulse measuring sytem was used, and I assume current to the motor was removed at the same time the light was turned on. Because of the dust, I don't think that any kind of IR or light sensor would work but perhaps it would.

Any thoughts appreciated.
 

premelec

Senior Member
Some years ago I built units that detected flow of material with op amps and LDR LED combinations; just 'looked' to see if there was motion in the solid matter or not... worked quite well so think a LED shining on an LDR at top level would work well. In my case there was some static electricity caused by the material flow; I used brass screen to protect the electronics from sparks.
 

depeet

New Member
Some years ago I built units that detected flow of material with op amps and LDR LED combinations; just 'looked' to see if there was motion in the solid matter or not... worked quite well so think a LED shining on an LDR at top level would work well. In my case there was some static electricity caused by the material flow; I used brass screen to protect the electronics from sparks.
Sawdust generates dustclouds which could make LED-LDR detection unreliable with false alerts as result.

Another approach is to use a scale that measures the weight of the sawdust-collector and alerts when a critical weight is reached.

Second proposal is to put the bin on a vertical moving platform that is supported by springs that compresses when the weight increases. A microswitch under that platform closes when the platform has descended enough through the weight.

Both proposal need to adjusted by trial en fail until the end results is satisfied. Buth I guess that the one with spring and microswitch is the most simple and cheapest
 

hippy

Technical Support
Staff member
Previously, a system was installed that apparently involved an electric motor with a long shaft with some sort of fan blades at the end of the shaft. ... I'm not sure if the motor solution used the increased current of a stalled motor,
That would be my guess. A 'weak motor' with low torque would likely stall with increased current flow when it was surrounded by sawdust but not when in free air or a dust cloud. Rather than having traditional fan blades which 'cut through air' it would likely have had paddles to make it harder to rotate when the sawdust reached its level.

It could be that the system did not measure stall current directly but detected rotation, which would cease when stalled, reduce as stalling approached. There are probably mechanical means which can determine 'rotational tension', drag, or whatever the technical term is, which might have been used to activate a switch and turn on the light.

If the motor was fitted externally, as I would expect, that should be easy enough to replicate.

Whatever they are I would expect they are common in industry, for example, "paddle level detector" ...

https://www.indiamart.com/proddetail/rotating-paddle-level-switch-18037119173.html
 
Last edited:

erco

Senior Member
Differential pressure sensor compares vacuum at pump to top of bin, calibrate alarm for pressure drop from full bin.
 

rq3

Senior Member
A friend has asked for help with a problem he has at his commercial woodworking shop. All of his power tools have vacuum hoses attached with appropriate gates that open when a particular tool is turned on. Sawdust created by the tool is vacuumed through a hose system and deposited in a large bin in the basement of the building.

If the sawdust is allowed to accumulate above the top of the container it clogs hoses and requires several hours of work to clear. The friend wants some form of sensor to illuminate a light in his workshop when the bin is perhaps 3/4 full so that someone can go down and empty it into a separate container.

Previously, a system was installed that apparently involved an electric motor with a long shaft with some sort of fan blades at the end of the shaft. When the sawdust reached the level of the blades the motor would be stopped and that somehow illuminated a warning light. This system failed at some point and was removed and discarded.

Is there another method that would work equally well or should I look at the motor solution again? I would program a picaxe to work with whatever sensor is appropriate to illuminate the warning light. I'm not sure if the motor solution used the increased current of a stalled motor, or whether some sort of pulse measuring sytem was used, and I assume current to the motor was removed at the same time the light was turned on. Because of the dust, I don't think that any kind of IR or light sensor would work but perhaps it would.

Any thoughts appreciated.
All of the "wet/dry" shop vacs I have ever seen have a large rubber ball below the vac motor. When the wet vac is full of water, the rubber ball floats up to plug the vac intake, preventing water from going through the motor.

I would think a similar idea would work just as well for sawdust, with the position of the ball detected with a microswitch.
 

geezer88

Senior Member
I've worked with several industrial dust collecting systems. The stirring paddle with a stall detecting switch is the method of choice for several reasons:

Floats simply get buried as the dust rises.

Delta pressure switches tell you when the bags filters are getting clogged. Useful for filter changes, but not bin level.

Light beams get blinded by the dust particles, but may work on granular stuff that isn't too dusty.

The ones I've used measured torque on the motor when the motor twisted on it's mounts and tripped a micro switch. These days a motor current sensor would probably be simpler and more reliable.

Other thoughts:

I don't know how well ultrasonic systems would work. Distance measured from above may work when dust collector is off.

Another method that might work is some kind of a vibrating cantilever beam driven at at it's resonant frequency. It would then change amplitude as it gets buried.

The Picaxe capacitive inputs might work, but it would take some experimenting to find out.

good luck, and keep us posted,

tom
 

wapo54001

Senior Member
Thanks all for the suggestions, and I've passed them all on. I think the motor & paddle would be the most dependable, The dust levels are pretty severe and I don't expect optical solutions would work. I'm startled to find that commercial paddle motors are available and a bit dismayed by how much they cost! Again, thanks for all the suggestions, they were very helpful in focusing our ideas.
 

erco

Senior Member
tom wins! Great to hear input from someone with real experience. That's worth a lot more than speculation.
 

premelec

Senior Member
I agree and it would be helpful to see picture of actual item to be monitored when it is almost full.

@wapo54001 should be not hard to run a small permanent magnet DC motor at fairly low speed and
monitor it's current... just protect it from dust with appropriate shaft and seal ;-0
 

wapo54001

Senior Member
The "barrel" to be monitored is made of heavy duty paper product, has a latching top with wrap-around locking ring. The barrel is perhaps 3~4 feet tall, 30~36 inches in diameter. Actually, there are two such units that fill pretty evenly so only one needs to be monitored. I can easily monitor voltage across a resistor in-line with the motor, with a stopped motor drawing much more current than a freely rotating one.
 

wapo54001

Senior Member
Hmm, read the spikes to check for rpm? How would I do that? Let's assume an 08m2 -- I don't see the need for anything more than that . . .
 

Pongo

Senior Member
Hmm, read the spikes to check for rpm? How would I do that? Let's assume an 08m2 -- I don't see the need for anything more than that . . .
Here's one suggested circuit to generate speed dependent pulses. Those pulses might be too short for the 08M2 to COUNT in which case add a pulse stretching monostable (LMC555) or bistable.

When I have done this previously I have used a simpler circuit with an inductor/choke on the ground side of the motor and some diodes/resistors, but I have no brushed motors at the moment so I can't give you a proven circuit :(
 
Last edited:

hippy

Technical Support
Staff member
Detecting commutator disconnection is an interesting approach.

Not sure what voltage conditioning would be required but it might be possible to use just a PICAXE on-chip comparator or ADC to detect those though the additional circuitry did not seem that onerous.

Just latching the cleaned-up pulses on a digital input using a POKESFR configuration would avoid needing a monostable and might be enough to allow determining if rotation were occurring above a particular speed. Alternatively using an on-chip timer as a counter could work and even give a rough indication of actual speed.
 

AllyCat

Senior Member
Hi,

The rise in current when a motor stalls is so large that: (1) it should be easy to detect and (2) it's usually necessary to take some precautions that the high current won't cause damage. For the latter I would use a "hardware" method (maybe just a series resistor) and not rely on the software in a PIC(axe). ;)

The operation of the commutator in a motor is quite complex. The (carbon) brushes are designed to more than just span the "gap" between segments, so current may flow into (or out of) two adjacent segments for a significant proportion of the time. As the commutator rotates, the current should gradually transfer from one coil/segment to the next. Therefore "spikes" don't necessarily occur, and the amount of ripple may depend greatly on the motor characteristics and its operating conditions. In principle, a three-segment commutator should "ripple" through six cycles per revolution (three sectors for each of two brushes) but I wouldn't want to rely on it.
Personally, I dislike the use of trimpots, particularly in circuits where they're not needed and/or are included to hide a deficiency in the design. ;) That pot affects the operation of the circuit in at least three different ways! : As a "gain" control it attenuates the "ripple" from the motor from 0 to 100%, but its source resistance also changes over a range of 0 to 1.25 kohms (or more) which alters the time constant formed with C2. However, its main purpose is to apply a dc bias to the transistor, which needs its base to be at "about" 0.6 volts to give a switching output. Thus with around 1 volt at the top of the pot (as shown on the schematic), it needs to be set at about 60% towards the top.

However, the comment that "about 50mV to 100mV peak to peak ripple waveform is ideal" is worrying. 50 mV pk-pk is of course +/- 25 mV so the peaks could be only 25 mV above (or below) the bias voltage. But the Vbe of a transistor has a temperature coefficient of about 2 mV/degC, so if the "room temperature" changes by more than about 12 degrees then the transistor may be biased permanently on or off. Therefore 100 mV pk-pk might be better and acceptable for a simple "indoors" household environment.
But personally I'd throw out the pot and use a simple (one resistor) voltage feedback from the collector to the base of the transistor, to ensure that the transistor is always biased adequately.

Cheers, Alan.
 

Pongo

Senior Member
@Alan. I agree with you that circuit is not ideal. As posted above I have made this work more simply.
 
Last edited:

wapo54001

Senior Member
I'd like to report the final outcome for my friend, to close the loop: he has decided to buy a purpose-specific device, the Bin Master mini rotary unit for $193.00. The price is very reasonable, and I recommended he consider those because of the complete dust-proofing, etc. Since it's a business for him, it's more about reliability and ease-of-use than it is about technical ingenuity. Further, even if I did it with an assembly of parts it wouldn't have been much cheaper to accomplish and would have had a number of reliability weaknesses.

I was quite amazed by the variety of ideas that surfaced in the process, and the idea of reading pulses across a DC motor was entirely new to me. Thanks all for the very helpful discussion.
 
Good to hear that the problem is solved. I was going to note that sawdust and air can be a very flammable combination and any electrics/electronics within such an atmosphere need to be extremely carefully designed (the sort of precautions they take in mines for instance) so going with a purpose built commercial application is much the best way to go.
 
Top