Test bench for measuring torque, rpm, volt and current.

[jadujadu]

Hi everybody!
First of all I have to excuse my english, but I hope everybody can understand me.

I'm building a test bench for AC motors, where the AC motor will be connected to a DC motor(will be running as an generator) with an adjustable load connected.

I've been trying to find a formula for calculating the torque produced by the AC motor, with just using the readings from the DC motor,but I can't find anything... I know that the torque is proportional to current, but I just can't figure out how to do this.

This is a new type of problem for me, since I have only been working with oil and gas installations, with very little motor installations, and non DC-motor installations.

The only parameters for the DC motor I have at the moment is:

Generator: 2,2kW @ 1500rpm
Motor: 2,0kW @ 1400rpm
Rotor: DC 220V 12A
Excitation DC 220V 0,8A

I would really appreciate if anyone could help me out here.
Great forum btw!

 

[MikeHalloran]

It would be simpler to just use the DC motor as a brake and measure its reaction torque. I.e., remove the base, mount it on bearings so it's free to rotate about its own shaft axis, and restrain it from rotating with a load cell on a radial beam attached where the base used to be. If you don't care about transients, you could use a mechanical force scale instead of a load cell.



Mike Halloran
Pembroke Pines, FL, USA

 

[jadujadu]

I forgot to say that I'm modifying an old test bench, which is built just like you decribed, with a loadcell to measure torque.

The old bench is analog, and I want to convert it to digital, so that I can get the parameters on a touchscreen. The touchscreen will also be connected to a VFD so that I can controll the speed of motor that is being tested.

Thanks for the reply anyways!

 

[jadujadu]

I could use the old loadcell setup, and input the measurment to a DAQ, but I really would like to calculate the torqe without it.

 

[jraef]

Do you have something to measure the kW being used by the motor and an accurate speed it is running? If you also know the motor efficiency, you can calculate the torque backwards from that.

Tq (N-m) = kW x 9550 / RPM

But to be accurate, you would need to adjust the measured kW by the strated motor efficiency, because you want the real shaft kW (absorbed power) if you are trying to determine mechanical torque from that. So take your measured kW and multiply it by the motor rated efficiency at the percent of load that you have applied to it. Most motor mfrs will be able to provide that for you now, but if you are using an old motor that may be a challenge.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[LionelHutz]

You could use also the power output for the DC motor plus the DC motor losses (efficiency) to calculate the shaft power and then use the speed to calculate torque. However, the efficiency will vary at every speed and load point and will require you to build a table of values to use in the calculations.

So, it's really the simplest to calibrate the load cell with some known weights and then use it to measure the torque.

 

[btrueblood]

Any measurement you make will have errors associated with it. So why compound your problem (of measuring torque) by measuring something else, and then using other measured parameters to calculate torque, thus compounding the errors?

 

[waross]

Mike had it the first post. Use a load cell. Then play with the current figures at different loadings, calculate the torque, compare the calculated torque to the measured torque and see how much error you have avoided.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jadujadu]

Thanks for all the great tips! I think I'll just stick with the existing load cell, and use the 4-16mA signal.

Another question, what is the actually the difference between PLS and DAQ? I know that the DAQ is much faster, but there must be other differences?

 

[jadujadu]

Sorry, forgot that PLS is called PLC in english.

 

[waross]

Does DAQ become DCS in English?
There was a recent discussion here.
Difference between Scada-PLCs and DCS
thread830-157405

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jadujadu]

You were right, thanks!

Another question:

I'm going to measure te voltage going from te VFD to the motor. Since te voltage doesn't have a perfect sinus-shape, I have to use true RMS to get the right measurment. I'm familiar with RMS, but how do I measure true RMS? I've read something about using differentialprobs.

 

[electricpete]

The term "true rms" emphasizes that the algorithm will work for any waveform, (not just a sin waveform).

An example of an algorithm that is NOT true rms would be finding the peak and dividing by sqrt(2). That works for sinusoid but not for many other waveforms.

An algorithm that gives true rms for any waveform will ususally be based on sqrt of sum of squares of time samples over a representative interval, divided by sqrt(N) where N is number of samples.

Older analog measuring circuits often did not measure true rms (for non-sinusoidal signal). All or most modern digital measuring devices compute true rms (for any signal).

=====================================
(2B)+(2B)' ?

 

[danw2]

>Does DAQ become DCS in English?

I don't think so. I hear the term DAQ (data acquisition) used in the test and measurement world; using stuff like PC cards with fast I/O and a software package, like National Instruments.

The typical process DCS has a sampling rate of what, 5 to 10Hz? Is that fast enough for this type of test stand?

 

[waross]

Thanks danw2. I suffered a seniors moment. Grin

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jadujadu]

Thanks for the answers!

I'm quite new in the world of power electronics, so I was wondering what the difference between a "normal" voltmeter and a differential probe is?

 

[MikeHalloran]

Jadujadu, please start a new discussion of your very own, with your very own topic title, separately.

Then please Red Flag your own message above, and in the dialog box that appears, ask that the message be deleted.

Thank you.


Mike Halloran
Pembroke Pines, FL, USA

 

[jadujadu]

Well, this is my very own discussion? I've started the thread because I have some questions about building a testbench. The question about differential probes is related to my testbench.

 

[MikeHalloran]

Sorry, I didn't notice.
However, consider that it's really a separate question that deserves its own title. Consider also that a new discussion may draw in some different viewpoints.


Mike Halloran
Pembroke Pines, FL, USA

 

[LionelHutz]

What's a "normal" voltmeter?

A handheld voltmeter is differential already. Many panel voltmeters are differential as well.

A differential probe is used to isolate the ground in a scope.