How Do I Build a Torque Measuring Device?

isn't it possible to get your information from the ac-drive? Most drives I know have an analog output you can configure as you like, so in your case give a readout of the torque

Oh, maybe, I will look but I dont remember seeing any.

I should have also mentioned the motor goes to a gear box...that is then connected to the transformer..sorry. So really the sensor goes from the output shaft of the gear box to the input shaft of the transformer core drive.

no, there is no analog output,,,and it is a dc motor.

It depends on the uncertainty you can tolerate for your measurements. If you can afford possible errors in the range of a couple of percents then you can use the input current as a value proportional to the resistant torque. You can also measure the current by using a special but available current measuring equipment which can be mounted on the supply cable. It is the lowest price solution. If you want to measure at the box output it will cost you much more and you will have to place the sensor which means to displace the box with respect to the machine introduce couplings, be sure every thing is well centered and so on. If you load does not vary much the losses in the box will be also quite proportional to the load. The current sensors work with a Hall sensor sensitive to the magnetic field generated by the current around the cable. It is simple to put in place and low cost. Try this possibility first.

I agree with nick name, this would be the easiest way to do. If you'd like to know the torque after the reduction gearbox, just do the conversation by multiplying it per the reduction factor, and the gearbox efficiency of course. Your motor eficiency may not be so hard to approximate, if, as told before, some percents of error is alowed. Just remember that P=Ui times efficiency, and also that P=torque*angular speed (all those stuff in coherent units of course). Oh, don't know about your motor drive, but if it does not come out with constant or some control for speed, you'd also need to measure it (easily done with an ordinary hobbist optical sensor). I'm affraid a data acquisition equipment should be used if datalogging is required. If you need real time (or almost) display of the curve, go for full data acquisition, signal conditioners, and all that good things that money can pay for. If you would like just to acquire data and make the analysis after the event, than you should be comfortable with a simple data logger or some current and tension measurement devices that record it for you from time to time.

Yes, thank you for the help. could you recommend an actual product for a sensor, signal conditioner, data acquisition, and a simple logger? Or maybe a company that would send me all the things I would need to make it complete?

I think it is much cheaper to change the dc motor with an ac motor and use a frequency drive to control the speed. This frequency drive should have an analog output which can give you the motor torque (Nm) or even the motor absorbd power (kW).

Eventually a load cell unter the motor feets to measure the torque.

Some ideas expressed before, like deriving the electrical power, or having the motor stator react against a load cell are simpler.

The solution you researched is the most precise, but more expensive. Some rotary load cells use radio signal instead of a slip ring.

Other means, such as encoders at both ends of a deformable shaft would give you angle deformation data from witch a torque value could be calibrated. All you need is 1 impulsion per turn and measure the phase angle difference. Gears and magnetic pick-ups work too.

As for the data acquisition part, some USB electronics are really quite inexpensive, and come with basic software.

Is this to be a permanent installation, or do you need to test once?

Do you have a budget for this?

What kind of resolution do you expect?

As hardware may be between $1-3K, or $20K for some solutions, this may be directly what drives the technology to (not) use.

If a qualitative value is ok, some polarimetric means can be used.

"The solution you researched is the most precise, but more expensive."

I think, may be I am wrong, that the solution was not "searched" but the only one known. It is of course THE best solution if the uncertainty imposes it. If not the slightest value analysis -and engineers are today more and more obliged to think economically- indicates that the current measuring approach is the best value related solution. A current measuring device costs in the range of 50 to 150 $. It is even possible to build one DIY with an iron ring and a Hall sensor from Allegro. The reading can be done if records are only point oriented with a simple micro-voltmeter (20000 points) from Fluke or others. The supply for the Hall sensor can be made with a 9V battery + a voltage regulator to 5V.

Why make things complex when it works simple as well?

Here is a test motor I use for the initial build...at the bottom is the completed transformer coupled to its motor.

Thanks to everyone so far for all the input.

to answer Gigaconcept's questions..

Not a permanent setup, want to reuse this on occasion though on different units.

I was thinking to keep it cheap, but I am prepared for a couple thousand $ if need be.

resolution of maybe 5-10 ft- lbs would be nice. as I do know one part will fail at 254 ft-lbs.

Yes, a value could be used but it the deviation is large enough versus gap position to notice the difference,,i suppose.

Very interested to see what might be the suggestion for $1-3 K

thanks.

I think the best is for you to measure the torsion of this shaft under load, either using electrical pulses or a good scientific stroboscope.

Using electrical power alone will give you power in, but not power out of the mechanical system, witch may have varying efficiency.

The first method may enable you to calibrate the electrical power vs torque method, and evaluate if is is precise enough.

Simple, cheap, and can be made precise. Derived from above:

Place two targets on a torsion shaft and use two photo detectors to get pulses. Feed the pulses into a differential edges timer. The time difference is function of the speed of rotation and the torque.

You could even take the measurement manually using a stroboscope or fast camera and a parallel guide near the shaft. The position error between the two gradings at the shaft's ends is function of the torque and the spring constant.

There are company that rent out load cell and data logging equipments.

would you happen to know the name?