Any Options Other Than Limit Switches? 3

[Powercat98]

Hello!!

I have been asked to use a small (1/30 hp) worm drive gearmotor (12V DC) to rotate a discharge chute in response to operator input.

What is the best way to limit the maximum travel of the chute to 180 degrees? I know I could use limit switches/relays, but I'm hoping for a "cleaner" solution.

Would a mechanical stop in conjunction with an "amp" switch to protect the gearmotor be a good idea? I'm thinking that an older automobile power window system would be similar.

Since I rarely delve into electrical projects at work, any ideas or suppliers for components would be greatly appreciated.

Thanks, ~Eric

 

[VEBill]

If you can find a safe place to install an absolute position sensor (one that knows where it is even on power up), then you could add an intelligent control system for the ideal solution (maximum flexibility, self-monitoring, safety). But even then, it might be wise to add safety limit switches near the extreme positions if bad things happen past the ends.

Another trick is to design the mechanical drive system so that it simply reverses direction at the extreme limits. In other words, the chute moves back and forth while the motor just keeps going. **Like windshield wipers.** This way there is no end stop to worry about. All you have to deal with then is getting it to stop in the correct locations, but that may be much less critical and perhaps can better tolerate failures.

 

[zekeman]

I've been through this and can say unless you want to spend big money, the limit switch is as good and reliable a solution as any and is probably the most reasonable.In addition you can protect the motor, by using the proper fuse to cut power (increased current caused by hitting a spring over-travel stop) due to switch failure.

 

[CharlieGill]

Limit "switches" are the best way to go, although I would try to use inductive proximity switches if possible. Even if the chute is not ferrous metal, you can attach a small metal target to it.
An "amp switch" would be expensive and tricky to set up to eliminate false trips. You would also be subjecting the mechanism to stress each time you went to a limit.
I would use (2) prox switches and a micro PLC. I tend to favor Allen Bradley MicroLogix but others are available. They cost well under $200 and considering the wiring you save and the flexibility I think they are worth it.
AB has a free version of the programming software for the small MicroLogix's.
Automation Direct (and others) have small PLCs with inexpensive software.
If you do it with relays, it requires (2) 3 pole relays to reverse a DC motor and detect the limits.
Charlie

Charlie Gill
--

http://www.calibrator.com--

 

[CharlieGill]

I did a quick check...
Automation Direct DL05
part# D0-05DR, 8 DC IN, 6 Relay OUT, $99.
Software is free for a 100 word program which will be plenty for your application.
This gives you all the IO you need for (2) limit switches, Operator input switch(es), motor leads, alarm light, etc.
Charlie

Charlie Gill
--

http://www.calibrator.com--

 

[OperaHouse]

Since it is operator directed, all you need are two limit switches in series and two diodes. When the limit switch opens, the diode in parallel with the switch contact will only allow current in the reverse direction. I can think of other ways but this seems simple and cheap.

 

[CharlieGill]

I hadn't thought of using diodes, but you are right. I always try to do everything in software
A DPDT switch, 2 diodes and 2 limit switches will do quite nicely.
Simple is better (but software is more fun).
Charlie

Charlie Gill
--

http://www.calibrator.com--

 

[lumenharold]

Why not just stop the gear teeth at 180 and 0 and then undercut the rest of the tooth area. That way the worm can run without a limit switch or proximity sensor or anything else unless the operator needs feedback that the limit has been reached. The trick is getting the gear to re-start when the direction is reversed. I have worked with this type of system before with the exception being that the gears were a rod and nut. If you monitor the motor back emf then you should be able to tell when the gear is running free (that would be the fun software part).

Harold
SW2009 SP4.0 OPW2009 SP2 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[OperaHouse]

"stop the gear teeth at 180 and 0 and then undercut the rest of the tooth area"

"you should be able to tell when the gear is running free"

You mean by the grinding sound of the last tooth?

 

[lumenharold]

Why would it grind OperaHouse? If the gear teeth are undercut to the minor diameter then it should just sit there and go "click, click, click..." until the direction is reversed. Why wouldn't it work on a worm gear as well as it works for a nut and screw? Perhaps I'm missing something but if the gears are aligned right and the pre-load is correct then the gear should just rethread onto the worm. As for being able to tell that the worm is running free that should be detectable by watching the current draw.

Personally I do not like limit switches. Parts wear and the limit switch settings don't remain constant over time. There's always some hysteresis is a limit switch as well. I've had too many of them fail to want them in a system. If you have no choice (software options are not available) then that's fine. When a limit switch fails it doesn't make grinding sounds. The motor dogs down until the gear head strips and then there's a bigger problem.

Just my opinion based on my experiences so feel free to ignore it!

Harold
SW2009 SP4.0 OPW2009 SP2 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[BobM3]

Interesting approach with the gear. The torque acting on the gear would have to reverse at some point during engagement for it to work though, correct?

What kind of motor is it? If it's an AC motor, maybe you could count line cycles. Of course if you missed one every so often you could have a cumulative error build up.

 

[fangas]

I'm thinking geneva gear

Ed

 

[Powercat98]

Thanks for all of your helpful replies!!

I guess I was being a little naïve to think I could effectively eliminate limit switches from this project. I would still like to know how older (1980’s - 90’s) automobile power window systems work. I’ve never seen one use limit switches, and I don’t think the motor is monitored by a control unit. Anybody know what the concept is?

VE1BLL/CharlieGill:
I think you’re both right on as far as the ideal solution (max flexibility, safety, etc), but for now I would like to limit the travel of the chute without utilizing a control system/PLC. Although that may change by the time this project is complete.

OperaHouse:
A DPDT switch, 2 diodes and 2 limit/proximity switches would be fairly simple.

The 12V DC gearmotor has a “starting” current of 12 A. Are there any recommendations for good limit switch suppliers/components that will handle this current? Like CharlieGill suggested, I think an inductive proximity switch (or any sealed magnetic type switch) suitable for agricultural applications would be best.

 

[VEBill]

One concept I've never had the chance to implement is that limit switches (or functional equivalent) should be installed in a location so that if they fail, they're destroyed. This may inspire them not to fail (joke).

More seriously, this sort of approach may provide a simple and inherent fail-safe if you can arrange things so that the resultant mechanical overshoot either opens the circuit, or shorts it to ground (and thus opens a breaker).

For example - imagine a glass fuse mounted at the ultimate mechanical limit of motion.

 

[ScottyUK]

Some of the mid-90's window controllers used current sensing to detect when the motor had either reached the open or closed position, or had hit an obstruction like an arm or a head in the aperture. The company I worked for at the time was making them.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[BobM3]

How do current sensing methods get pass the initial high starting current?

 

[OperaHouse]

A simple current limiting method is to use a lamp in series. This does not work if the motor is loaded close to its maximum. Resistance of the lamp greatly increases as current rises. I have used in place of fuses when diagnosing automotive problems. Windshield wipers easily work with a headlight lamp. Recommended that you have very soft stops to allow time for the lamp to heat.

A variation of this is to use resettable PTTC thermal fuses. Specs are like nailing jelly to a tree. Best to greatly undersize the fuse to actual current drawn compared to fuse rating.

The trouble with methods like this are the human operator. Typically bored and easy to amuse, I think they would just hang on a chute and let the teeth make music with the chopped off teeth method. I expect slamming into a stop would be a typical mode of operation for an operator.

I had a bus manufacturer that wanted to eliminate switches on a wheel chair lift with a current sensor. I made him buy a graphing DVM to monitor current. He was thrilled to get an excuise to buy one. Once he got the data, it was doubtful that even a micro could figure out the wild current swings from an end of travel stop.

 

[VEBill]

I still think that the ultimate solution is the one I proposed at the very top (1st reply, 2nd para).

Make the drive mechanism inherently back-and-forth with no end stops (just like windshield wipers). Then the only remaining problem is to have the chute auto-park in the correct locations.

With this approach there is no destructive failure mode. If it overshoots the mark, the mechanism just turns around anyway and comes back from the other direction.

And furthermore, an operator over-ride system can be provided in case the auto park sensors fail.

 

[imcjoek]

The pop-up headlights on my Vette use the "overcurrent on smashing into something" method.

I have to index the worm wheel often to find fresh teeth...

(do not recommend!)

 

[StephenVB]

I made a simple controller a few years ago to move a rack and pinion driven by an inexpensive 12VDC wiper motor (about $25 then) made by AM equipment. The positioner controlled doors and moved until it stalled. The motor had much more torque than necessary so a current limit was advisable to prevent damage and injuries. Since the travel and mounting varied, limit switches were not a good option anyway.

I used IR3312 high side drivers (max 100A) and nmosfet low side drivers. The high side drivers have current limiting and the current is proportional to the voltage on the current resistor so I used a 2-ch I2C adc to measure the current. The current limit was adjusted with rocker switches, 8 steps from 4 to 32A. The adjustment wasn't critical. Once the motor stalled the current rose so quickly that all currents above the normal load had about the same stopping time. This was controlled with a $1 AT89C2051 micro processor. I charged $100 each for these in single quantities, less than a simple PLC without an analog input.

This had some significant advantages over limit switch designs.
1)No wiring or mounting for limit switches and no adjustment.
2)This would stop if something or someone was stuck in the actuator or the actuator bound up, without damage to the motor, assembly or person.

This worked great. I highly recommend simplifying the design as much as you can. I worked on another design that used limit switches to lower a plasma tv over the windshield in a motor home. If the limit switches failed the electric actuator burned up and if the little old lady got between the TV and windshield it pushed her out onto the ground, apparently a bad thing.