Wheel movement sensor / alarm

[Halfwit]

Hello all,

I work with a couple of friends at a mill. I run the hammer mill area: 150hp motor driving the mill, auger motors and feeders and the like... very low tech.

Recently, thanks to a fire, I lost my single phase 3 hp feed auger motor. I had to get some work done, so i wired up a 5hp 3 phase motor in its place. I though this would be great, more power and more reliability (I was always repairing the 1 phase motor due to dust intrusion - yes it was sealed... lot of dust).

Well, I have plenty of power - so much so i blow through belts like no tomorrow.By the time I see the system is jammed the motor has pretty well chewed through my double, B52 belts.

I tried to limit the current using a thermal over load relay (TK-E02-900) but even though I have t set to well below the motor's rating, it never trips - I believe its due to the fact the system never gets to static friction and stays kinetic...

So, I wanted to design a wheel sensor motor. One that would light and buzz when the wheel stops.

I was going to use:
1) a magnet/coil to generate a signal (magnet glued to wheel, stationary coil).
2) signal from coil to drive a transistor or opamp to amplify that signal
3) Send amplified signal to a timer re-set.
4) allow timer to trigger the lights and noise makers if not re-set in time.

But alas, I cant get the coil / magnet to trip the transistor. (my op amp just sat at +rail, so i moved to a simple transistor). If I open the coil leg and touch the leg a signal is amplified and my LED (stand-in for down stream circuitry to be designed later) it lights up just fine.

Am I asking too much of the coil? Is there really that little v/a in a coil excited by a magnet? I must be doing something stupid... I can measure only 1.6mv when I pass the magnet past the coil.

I attached a simple drawing, but its a simple LED driving circuit.

I am open to ideas / circuits I could build. I'm not married to my idea, I just want to get some work done without burning up belts.

 

[duke37]

I can think of two alternative methods.

1. Use a reed switch.
2. Use a Hall effect device.

Both of these can be tested statically unlike a coil which needs a varying magnetic field.

If you turn off the motor when the shaft stops, you will have to provide some circuitry to enable start up.

 

[Halfwit]

Hey thanks for that. I considered each of those after I couldn't get the coil to work on the breadboard. I like the hall effect sensor over the reed but was wondering if the 150hp or even the 5hp motor that are 6' and 1' away respectively would interfere with the hall effect sensor. Do you think the motors will? I guess it's not that strong a field, cause my tools don't get stuck to my motors.... Lol.

Plus.... Now I'm curious why the coil and magnet won't energize the transistor but me grabbing the leg does! It's really bugging me now - - you know?

Thanks so much for taking some time.

 

[CDRIVE]

Plus.... Now I'm curious why the coil and magnet won't energize the transistor but me grabbing the leg does! It's really bugging me now - - you know?

Thanks so much for taking some time.

Did you wind it on an iron core? What AWG is the wire? What's the orientation of the coil to the magnet?

Chris

 

[Electrobrains]

That the transistor reacts on touch on the open base has to do with that the base is "floating" at the threshold of just being switched on. In fact it is placed (biased) in a high impedance position by the air and creepage resistances and inner leakage (compared to the low impedance state, when your coil's ultra-low resistance draws the base towards ground).

Your body works a bit like an antenna and capacitor. If you touch the open base pin, a small current will flow from your body. Because of the high impedance, most of that current will flow into the transistor and activate it.

You can make an experiment: Take off your shoes (the current will go through your socks!) and touch the base!
Does the LED still lit up? Or make the test outside or in a place where you don't have electric cables or machines near you (power the circuit with a battery).

To detect a voltage from your magnetic circuit, you would probably have to bias up the base position with high ohm resistors and decouple the coil with a capacitor.

 

[Halfwit]

No, I didn't wrap it on an iron core... I tried to use my needle nose pliers but got no change.

Orientation: To get the most number of turns, I just left it in the spool. I swiped the magnet across the coil, not through, because that's how I can mount the magnet on the wheel. As I was not getting any results I did run the magnet through the center of the coil, but still nothing. It only responds when I touch the terminal end with my hand.

Gauge: 22 & 30

 

[Halfwit]

Thanks EB, I will try it outside in the morning. I kind of have a hard time conceptualizing that the charge on my body is more than that which I can generate with a coil and magnet.

Chris, do I need an iron core to make it work?

Thanks again

 

[Halfwit]

EB,
Also, to bias up and decouple the coil, is that in series: gnd->coil->cap->resistor->bias?

 

[Electrobrains]

Thanks EB, I will try it outside in the morning. I kind of have a hard time conceptualizing that the charge on my body is more than that which I can generate with a coil and magnet.

I don't know what your coil can generate, but for sure you can store a lot of charge on your body (suppose you have experienced the sparkles after walking on certain carpets).
See: http://en.wikipedia.org/wiki/Human-body_model
The main problem here though, is the "impedance mismatch".

EB, Also, to bias up and decouple the coil, is that in series: gnd->coil->cap->resistor->bias?

You would put 2 resistors at the base of the transistor, one connected to +12V and the other one connected to 0V. That voltage splitting would have to place the base potential near the switch-on voltage (approx. 0.6V). The resistors should be high value (the upper one in the Mega-Ohm region).
The decoupling capacitor is inserted between the transistor base and the wire from the coil.

 

[CDRIVE]

A couple of notes here.
(1) For your transistor to turn on by simply touching a floating Base junction indicates that your transistor is a very high beta model.
(2) Unless you are located in an area devoid of power lines your body is constantly subjected to their radiation, especially within your home. You are a walking talking antenna.
(3) The current that your body can impart to other objects like the base of your transistor is very low and will be in the microamp region.
(4) Up and until your finger actually touches the base of the transistor the voltage between ground and your finger can easily exceed a few volts. This can be seen by simply touching an oscilloscope probe. The moment you touch the base this voltage will drop <= 700mV during positive excursions of the sine wave. Negative voltage excursions will probably be very low too. This is because there is no DC path from ground, through your body, through the BE junction and back to ground. This means that the small current that's flowing is AC and at the mains frequency. IE: Capacitive.

A Hall Effect Transistor would be a superior design but I've included a pictorial of your scheme with notes just for theory's sake. You can wind the coil on a large nail as it's soft steel. It's best to cover the nail with one thin wrap of masking tape prior to winding directly on the metal. 100 turns should be sufficient. Cut the nail to 1" before winding the coil but leave the head. On the other hand, it's easier to just cannibalize an old relay coil.

Chris

 

[duke37]

The relay coil is a good idea, it should be used with its original core.

The relay should be an AC type to get a laminated core and the copper ring should be removed. The magnet will be placed where the armature was.

You may be able to use an AC solenoid also.

 

[CDRIVE]

Duke, yes, when I said relay coil I meant core included less the armature. I'm not really sure that an AC relay core is required but I'm certainly open to all opinions. What is your reasoning for it? On the other hand it certainly won't hurt.

Chris

 

[duke37]

The reason for using an AC relay is that the core is laminated so the Eddy current losses will be minimised. This is also the reason for removing the copper ring which is there to provide some phase change on half the pole piece to reduce vibration.

I sorted out a shaft monitor on a combine harvester some time ago, this used a reed switch.

 

[CDRIVE]

Duke, that makes sense. Thanks for the clarification.

Chris

 

[Halfwit]

OK! both circuits work! Thanks so much. I tried to make my own coil and neither circuit worked... I was bummed. So I went to work and grabbed a coil from a junk timer (Omron H3CR-F8). That did the trick.

The coil I made:
120 turns (just to be sure)
26 gauge
wrapped on a 1/4' bolt.(It's all I had)

I assume it didn't work because of the metal the bolt was made from. Nails bend, bolts don't bend. I will re-wrap my coil on a nail later.

I found that by changing the capacitance in EB's circuit changes the triggering distance, which I thought was pretty cool. At first I used a .1uf and I basically needed to get right on the coil to trigger (less than 1/4"). Then I went to a 47uf and it moved out to bout an inch. I'll mess with the different components to optimize later - if I need it. Cdrive's circuit works between 1" - 1 1/4" (performance difference between the two circuits part of the reason I started messing with the components).

Thanks so much everyone!

 

[CDRIVE]

Yes, what EB described is called Quiescent or Static Biasing. It will greatly increase sensitivity. Great care must be used when setting forward bias near threshold voltage of the BE junction though. If temperature rises Q1's beta will increase and cause the LED to be on constantly. Adding a capacitively bypassed resistor in the Emitter circuit helps greatly reduce this effect. A properly designed magnetic pickup should produce more than enough signal bias on its own and shouldn't require any quiescent bias.

Chris

 

[CDRIVE]

So, I wanted to design a wheel sensor motor. One that would light and buzz when the wheel stops.

After re-reading this you may want to consider having the circuit kill the 3PH motor with a 3 pole contactor. After all, by the time you respond to the alarm the belt may be toast. If the motor already uses a latching magnetic contactor we can probably tie into the control side of it.

Chris

 

[Halfwit]

That's interesting. With what you are describing sounds like there is an option where I could turn off perhaps more than one piece of equipment - which, really is my ultimate goal.

I have a 3 ton feeder that feeds the auger with the 5hp motor we have been trying to alarm. If I don't turn the feeder off in time, it makes this huge pile of sawdust that I have to dig out by hand.. Kind-of stinks. So... Turning both off would be most excellent. I'm going to do some reading about what you mentioned. An look at what hardware I have on the first motor. It's a little different than the 5hp/3ph.

If I can turn off one motor, I can turn off two.. So I'm pretty psyched.

Process:
Pile'O dust -> Me and my Bobcat -> 3 Ton Feeder -> Feed Auger (5HP/3ph) -> Hammer mill -> auger -> transfer auger -> elevator -> Silos.

 

[Halfwit]

OK, I see what you are saying. That makes a lot of sense. For the small 3phase motors we are using Fugi:
3 Pole Contactor SC-E04,
Thermal Overload relay TK-E02
Auxiliary Contact block - driven by 120v

If I understand you correctly we can interrupt the 120v going through the auxiliary block upon trigger from the wheel sensing circuit. Would you use a relay or a transistor to interrupt the 120v / de-energize the coil?

Excellent idea, thanks.

 

[CDRIVE]

Yes, it's more than feasible to kill all the machines if you like. Is this a pulp mill?

Chris