Hi,
I am looking for a neat and cheap way of protecting motors and electronics from high stall current. I have some very basic electronic knowledge, but it and I are fuzzy and dated!

The set up in brief is 3s lipo (around 12v) into ESC (electronic speed controller) into brushless outrunner motor. I am very familiar with RC cars / robots etc, so this side of things isn't a problem.

What I want to achieve is a way of limiting current to around 5 amps so that if the motor is stalled for any length of time, it won't overheat. There is likely to be some thermal protection in the ESC. But it would probably 'recover' too slowly. I need a fairly fast reset. Some soft-starting effect would be acceptable.

I have looked at PTC thermistors as a possibility, but slow recovery time and max current is I think a problem.
I have heard of inrush and current limiting diodes, but know nothing about them.
I have seen a simple circuit employing an LM317 and one or more 2n3055s with an adjustable threshold here http://www.circuitsgallery.com/2013/07/high-current-adjustable-power-supply.html

...but I am not clear on whether it does the job! Building the circuit wouldn't be a problem.

I am open to any suggestions / ideas / solutions. Please go easily on me, Volts x Amps = Watts is about my limit at the moment. Many thanks.
P.

 

Most DC motors draw the same current stalled as do during initial start-up, so how is the supply supposed to know which is which?

There are lab supplies that have automatic switchover, separately adjustable voltage and current-limits. They dont do so well when trying to start a motor.

 

The best protection is to get the motor moving. If it doesn't move after some period of time then a shutdown might be required. This suggests a microprocessor solution. Are you up for that? NO. Well how about fuses or breakers. After a certain period of time at some multiple of their rated current they will open. Of course then a $55/hr plant electrician has to go and replace the fuse, but it's only money. If you have to replace the fuse it may or may not be convenient.
Motors are normally designed to take a great deal of abuse. What they are contained in may not be so tolerant. What is your experience in this regard?

 

Hi
Mike ML: Point taken regarding start up and stall. I am trying to avoid high current draw when motor is inadvertently held at stall for more than a couple of seconds.
Papabravo: I am not frightened of a "microprocessor solution" that can detect and respond to periods of high load. But cost might become an issue. I know that reset-able breakers exist, but then as Mike ML suggests, normal start up might well trip the breaker. The project is along the lines of RC cars, so manual reset is not ideal.

The solution I am hoping for also needs to be shock proof, so no incandescent light bulbs in line

Thanks both.

 

I have looked at auto reset circuit breakers and perhaps this one will do what I need:
http://www.maplin.co.uk/p/100a-auto-reset-circuit-breaker-ak13p

This one is rated 10 amp, which will certainly trip out on long stall periods, but would allow 120w at start up and be tolerant of short surges. What do you think? It's so cheap, it's a no-brainer to try out anyway!

 

Embed a thermistor in the motor, and trip if the temperature rises above a certain level..

 

Hi, I could do that, but I want to avoid thermal runaway and would prefer a faster recovery. If I attach to the motor (bearing also in mind that an out-runner has a spinning body), it will be slow to reach 'break' temperature and slow to reset.

 

Hi
Mike ML: Point taken regarding start up and stall. I am trying to avoid high current draw when motor is inadvertently held at stall for more than a couple of seconds.
Papabravo: I am not frightened of a "microprocessor solution" that can detect and respond to periods of high load. But cost might become an issue. I know that reset-able breakers exist, but then as Mike ML suggests, normal start up might well trip the breaker. The project is along the lines of RC cars, so manual reset is not ideal.

The solution I am hoping for also needs to be shock proof, so no incandescent light bulbs in line

Thanks both.

I wan't thinking that 'fear' was a factor, more like development cost, as you mentioned, and time. You did say 'neat and cheap'. Microprocessor solutions are never 'cheap' when you build one or a handful, only when you build thousands to millions. Good luck in your search.

 

Do you know the stall and loaded run current?

 

I wan't thinking that 'fear' was a factor, more like development cost, as you mentioned, and time. You did say 'neat and cheap'. Microprocessor solutions are never 'cheap' when you build one or a handful, only when you build thousands to millions. Good luck in your search.

Maybe a simple missing pulse detector with a 555 would be enough to sense if the motor is either not rotating or too overloaded for it to maintain its rated RPM's... that would be cheap and simple, I guess...

 

Hi Cmartinez, I like the idea. A sensor could be in the form of an led reflecting onto a mark rotating on the motor body into er! Are they still called ldrs? - light dependent resistors? Or there may be a better sensing method?

I think the logic would be:

If RPM > 1Hz, do nothing,
If RPM<1Hz for 5 seconds, then send ESC PWM input to 0 volts for 5 secs. Then, system resets

That's the easy bit! How I would configure a 555 or similar to manage this, I haven't a clue, so I'd best go and do some reading.

 

Do you know the stall and loaded run current?

Not exactly Ronv, we haven't settled on Lipo, ESC and motor configuration for this task, but ESC is likely to be 30a continuous. Motor in normal loaded run, typically between 4 and 8 amps. At stall, could well be in the order of 30-40 amps, maybe more.

 

Since it is heat that will destroy the motor (rather than any arbitrary value of current) you should act on this and embed a thermistor, as per MikeML's suggestion. If you use a microcontroller, you could avoid some of the issues you mentioned in post#7 by looking at the rate of heating as well as the absolute value to determine what action to take.

 

Hi Cmartinez, I like the idea. A sensor could be in the form of an led reflecting onto a mark rotating on the motor body into er! Are they still called ldrs? - light dependent resistors? Or there may be a better sensing method?

I think the logic would be:

If RPM > 1Hz, do nothing,
If RPM<1Hz for 5 seconds, then send ESC PWM input to 0 volts for 5 secs. Then, system resets

That's the easy bit! How I would configure a 555 or similar to manage this, I haven't a clue, so I'd best go and do some reading.

For a sensor, you could mount a magnet on the motor's shaft, and use a digital output hall effect sensor, or you could use a cheap optical or magnetic encoder, or you could even use a simple photosensor detecting a white streak painted on the shaft. Any of these options would output a pulse or series of pulses per revolution. All you'd have to do is configure a 555 circuit that would detect a missing pulse and then deactivate a relay that would shut down the motor... You wouldn't even need a microprocessor.... I can't think of anything simpler, to tell you the truth.

 

Here's a circuit that I designed many years ago (about 15, if you must ask...) that does just what I described in my previous post.



EDIT: The circuit should be run with a 5V source, instead of the 6V shown in the diagram, otherwise the hall sensor would probably burn out after a while. Also, R3, R5 and R6 should be 470Ω instead of 1K otherwise the LEDs would burn too dim.

 

How about an analog approach? Low side current sense resistance, could be as simple as a loop or two of 12 gauge wire, a series resistor charging an integrating capacitor for the delay time and a resistor to ground for the recovery time. Run this integral of amps x time to a comparator and have that switch the safety-out circuit.

 

Cmartinez and #12: I love these ideas, they get to the problem before the motor overheats significantly. Obviously that's good for the motor, it's also good for prolonging the charge in the battery. The battery eliminator circuit for the RC receiver (BEC) is 6 volts, so I have a two potential supplies for your 555 circuit. Noting that it should be 5v, I have bought and used, DC stepdown boards for just 99p in the last year which will work well. Special thanks for letting have the 555 circuit diagram.

Thanks to everyone that has responded so far, you have put me onto several different and interesting strategies. I'll report back in due course. Any more ideas would be much appreciated

 

I'm an analog person. Microcontrollers make me look like a dinosaur, but when an analog solution will do the job in less than 10 parts, it usually embarrasses an MCU for simplicity.

 

You could use the ACS712 hall effect current sensor:

https://www.sparkfun.com/products/8882

This IC falls within your current range, and allows for logic level voltage outputs for micros or digital components.

 

You could use the ACS712 hall effect current sensor:

https://www.sparkfun.com/products/8882

This IC falls within your current range, and allows for logic level voltage outputs for micros or digital components.

Yes, a current sensor such as the one you mention, or another by these guys should do the trick.

BUT! these sensors deliver current information in real time. That is, it will deliver a waveform of the current being consumed at a particular moment. If your motor is a DC motor then that shouldn't be a problem, but if it's an AC motor, then you MUST feed their outputs to an RMS converter chip (they ain't cheap!), that will deliver a DC voltage depending on the load. After that, you only have to feed that DC voltage output from the chip into a comparator and, voila! you're all set up. In fact, even if your motor is DC, I would still suggest using the RMS converter chip, since a brushed DC motor will show "spikes" in current due to its brushes commutation and that could probably confuse the comparator.
Trust me, I've used the Ampsense current sensors and that particular chip in a previous project... it was a drilling machine... the sensor delivers an overload signal to the controller that then adjusts the feed rate of the drill, and also proceeds to "peck" the material being drilled... and the system is working marvelously to this day.