High Torque Motor Solution

[ct4242]

Howdy,

I would like to preface this post with the fact that I am not extremely well versed in the electrical engineering department but should be enough to hang with the lingo.

I am working on building a fast acting shut-off valve system for use in HVAC engineering research. The setup consists of 2 ball valves to be actuated by some sort of servo motors. These will be connected to an Arduino board which I have experience using.

The requirements of the motors are as following:

Motor 1: Around 3NM of running torque and needs to be able to travel 90 degrees in .15 seconds (100rpm)
Motor 2: Around 15NM of running torque and same speed requirement

I am having trouble finding motors, specifically for motor two, that can satisfy this criteria. Any recommendations of brands, specfic motors, or sales reps that could help would be much appreciated. Ideally, these motors would be stepper motors as we are hoping to be able to control flow with this system too.

Cheers,

CT

 

[SwinnyGG]

That's a very fast move for that level of torque.

If I were you I'd be investigating pneumatic actuators for this application.

 

[ashtree]

There are lots of standard actuators out on the market. However your speed of operation is pretty rapid. Are you sure that a quarter turn valve is actually the right type for the operation?

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[MikeHalloran]

Let me tell you a story.

A couple of decades ago, I was able to make a size 23 stepping motor reliably turn a Hamilton chromatography valve by 1/4 turn in 50 ms.

A Hamilton chromatography valve is a teflon plug valve. The clear bore is 1.5mm in the larger size. The plug is maybe 4 or 5 mm in diameter. The operating torque is pretty damn small relative to your requirements.

The whole thing was run by an Intel 8741, which is probably about as fast as your Arduino is at a much higher clock speed, and was programmed by me in assembler.

The real trick to getting a stepper to move fast is to kick it, hard. There are other ways, but I came to favor a dual-voltage driver. I.e. the UC changed four outputs (and transistors) at a time in order to make the motor step, starting at a fixed rate determined by experiment, and ramping up and down along the way by about a 3 pct change in step timing at each step, then seeking back and forth at the fixed speed to find the center of a position signal provided by an optointerrupter looking through a disk with four slots. The disk must be metal.

At the start of each step, the UC turned on two additional transistors that shorted the resistors in series with the stepper windings. Your sparky friends can probably explain R/2R, R/4R, R/nR setups for steppers. In my case, the resistors were set up so that the steady running voltage applied to the windings was 5V, the nominal motor rating. While the series resistors were shorted, the voltage applied to the motor was around 28V.

That high voltage pulse applied to the stepper at the start of a cycle does not result in an overcurrent in the motor. The pulse is there to fight the inductance of the motor coils, and build up the current in the motor as quickly as possible. Steppers are weird; a size23 will move in 1ms >if it's going to move at all<, then ring a little, then remain stationary until the next step command comes along.

There are bigger steppers. They will require a slightly longer pulse or a higher kick voltage or both, and your speed requirements are a little less demanding. You will require a good oscilloscope with a good current probe, and a good sparky, and a very heavy power supply, if you are going to homebrew a driver like mine.


If you don't have any sparkys available, you might be better off to use fairly large poppet style air valves and an air cylinder with a bellcrank to turn your valves. It will likely be cheaper and quicker than using stepper motors, at least to see how fast you can make the ball valves turn.

Good luck and have fun.




Mike Halloran
Pembroke Pines, FL, USA

 

[3DDave]

I would set up a stiff spring and use a reduction drive to load the spring and then a solenoid to release the spring.

This way the energy to operate the valve would be stored ahead of time and as much force as required would be available.

I aided in developing a mechanism (the energy storage and trigger) like this to drive a soccer ball a few hundred feet for one of the FIRST competitions.