Gallery, Projects and General > How do I??

How do I measure the torque on a rotating shaft ?

(1/2) > >>

picclock:
Hi
I need to measure the torque applied to a varying load. The torque levels are low 1-20 gm cms. Would like to resolve 0.1 gm cms. It needs to be settable. A continuous measure would be best but I could at a pinch settle for a limit detection.

The rotation is supplied by a small geared electric motor, running at around 500 rpm, which is speed controlled by a microprocessor.

The best I can come up with is to mount the motor body in some bearings with stops to limit the rotation to a few degrees, balance it, and use a spring with an adjuster to exert torque on the motor. I could then use an opto switch to determine when the motor has rotated, overcoming the spring force.

The other thought I had was to use a low speed motor in bearings with a pot attached to measure the motor rotation against a spring with a gearbox to supply the load. this would increase the amount of torque to be measured at the motor end though the accuracy of the final output would be affected by the gearbox losses. 

Any ideas most appreciated.

Best Regards

picclock

PekkaNF:
When I checked some electric motor static propellor thrust and torque (separately) vs. setpoint. I build a simple balance which had an arm to electronic kitchen scale (electronic is easy to tare). That scale was surpricingly accurate, but I was not primary concerned about the absolute accuracy, just relative values (improvement or not...).

I used normal ball bearings without any seals, they were very good and simplified the MDF/aluminium profile contraption construction a whole lot. It was actually detecting change of the power supply cable movement (had to resort very supple ones....max 50-70 amps 15 sec and so).

Dunno if this helps, because propellor simplifies the mountig whole lot, did not need to worry about the missaligmentof the bearings and such....I don't know how your system is but I probably would think of letting the motor turn freely few degrees in relative to gearbox (it sounds like a small motori and bearings will take up the motor weight etc.) and measure the force motor is exterting. That is if you can mount the motor rigid on the gear box shaft (if you have a coupling there you need to device external support to motor----ballbearings?).

Pekka

RussellT:
I wonder whether it's possible to do this electronically?

In general terms (no units) power = torque * rpm.

If I understand your system correctly the rpm are constant so measuring the applied power should be a measure of torque.  The difficult bit is in knowing what the losses are. Motor efficiency is usually plotted against rpm so might not change very much but I have no idea whether speed control by PWM would affect efficiency.

Perhaps someone with more knowledge of electric motors would like to comment.

Russell

picclock:
Thanks for the responses, PekkaF and RussellT.

I think I may have fund a workable solution thanks to the kitchen scales idea.

As before motor mounted in ball race free to radially rotate but mechanically limited (stops). Load cell (100gm) attached to short peg off of motor (in lieu of kitchen scale). Load cell mounted on spring set to deflect and limit maximum force on cell to prevent overload damage. 24 bit A-D load cell amplifier and cheapo micro to read info and display on OLED screen. Micro will control motor speed and limit max torque from user setting - two buttons up/down with torque value displayed on screen, and dynamic value displayed as %age of max with an on screen bar.

Sounds like a plan to me anyway.

Once again, thanks for the help.

Best Regards

picclock

John Hill:
I do not think the ball bearing mountings are really necessary.

Mount the motor by the armature shaft and fit a spring to resist rotation of the motor body and measure the stretch of the spring to determine the torque.  An alternative to the spring would be to fix a string to the motor body so that increasing torque would life a weight sitting on an electronic scale.

However as Russell suggested I think it could all be done electronically.  Plot rpm and current under no load then measure current at the same rpm to determine power increase and hence torque.


 

Navigation

[0] Message Index

[#] Next page