| Author |
Message |
    
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 34
Registered: 12-2005
| | Posted on Sunday, October 22, 2006 - 06:16 pm: | 
|
I started this thread in hopes of finding simple setups to test motor performance. Torque curves are non existant for some motors (Chinese especially) and the ones available rarely are at the power levels we would run, or may not be the optimum for that motor. The only scientific test I've come up with is winch type setup that you would increase the weight and speed to get different stall points and chart where you would loose steps. Any one have any other ideas?
Dirk |
Gerald_D
Registered
Username: Gerald_d
Post Number: 317
Registered: 11-2005
| | Posted on Sunday, October 22, 2006 - 07:30 pm: |
|
Dirk, you are talking of a "Prony Brake" or a "dynamometer" - googling those terms turns up a lot of stuff. Here is a nice one.  |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 35
Registered: 12-2005
| | Posted on Sunday, October 22, 2006 - 11:45 pm: |
|
I'm sure a machine like that is out of my price range. When I said simple I also meant inexpensive. I was looking more for ideas with a MacGyver touch , basically a poorman's dynamometer. |
Mike Richards
Registered
Username: Richards
Post Number: 45
Registered: 05-2006
| | Posted on Monday, October 23, 2006 - 12:26 am: |
|
Dirk,
I use a hanging 'fish scale' that reads up to 25 lbs. To use it, I tie one end of 1/8th inch nylon cord to the scale, loop the cord around a V-belt pulley that is fastened to the shaft of the stepper, ramp the motor to a set speed, and then gradually pull on the cord until the motor stalls. The 25-lb scale is a little light for the larger steppers. A 50-lb scale would have been better.
If you use my method, be careful. There is often no warning that the stepper is ready to stall. Also, there can be significant heat developed as the cord tightens around the pulley. Nylon melts easily, so repeated tests would probably require sisal or other non-synthetic cord.
Remember that most motors are measured in oz*inches. Unless my math is messed up, that means the radius of the pulley is the inches part of the equation and the scale * 16 is the oz part of the equation. |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 36
Registered: 12-2005
| | Posted on Monday, October 23, 2006 - 01:57 am: |
|
Hey Mike
I remembered your posting about your cord test. My winch idea was a variation of your test. Increasing speed on a certain amount of weight until stall. I've done quite a bit of web searching since my first post. I'm thinking about hooking up a generator (brushed dc motor) to the steppers shaft (direct couple the shafts). The step frequency going to the driver would give Shaft RPM. The load could be a bank of light bulbs or resistors hooked to the generator. This seems like a rather easy solution as the current produced by the generator can be calculated to HP or torque. It could also be interfaced with Data Aqusition software on a computer to graph the torque curves.
Here's a pic showing one I ran across on the web
What do you think?
Dirk
 |
Gerald_D
Registered
Username: Gerald_d
Post Number: 319
Registered: 11-2005
| | Posted on Monday, October 23, 2006 - 05:28 am: |
|
Using a motor as a generator as a dynamometer is fine as long you don't want to measure very low speed (near or at stall or startup). The pic above doesn't show a torque measuring system.
If you looked at some Pronies, you will have seen that the wrapped cord/belt type system (as used on exercise bikes) should officially have a fish scale at each end of the cord, and the difference in readings is the one used for the torque calc.
Modern dyno's have a torque arm resting on a strain-gauge loadcell. |
Gerald_D
Registered
Username: Gerald_d
Post Number: 320
Registered: 11-2005
| | Posted on Monday, October 23, 2006 - 09:23 am: |
|
After thinking about it in the traffic this morning, you need 3 devices:
a. A braking unit: The device that applies load or resistance
b. A torque measuring system.
c. A speed measuring system.
Starting from the bottom, the speed measuring on a stepper is from the input signal, but maybe you want to test servo motors one day.....?
For torque measuring, I reckon a torque arm pressing on a digital kitchen scale would do the trick nicely....
But the braking unit needs the most thought. You probably want to measure steady readings only, so you can have a fairly high inertia. In fact, to smooth the pulses of the stepper, you probably want a fairly high inertia, or "flywheel" effect. If you don't smooth the pulses, your torque reading could be difficult.
A DC brake motor brake with resistors is good for the higher speed readings. If that motor is big enough to stay cool at a locked rotor torque equal to where your stepper might "fight it", then you simply need a driver (Gecko 320?) to power the DC motor adequately. In other words, dispense with the passive light bulbs and actively drive the brake motor instead.
To mount the brake motor, get one with shafts out both ends. Put a U-cradle under the motor with its shaft going through (frictionless) bearings in the arms of the U. The torque arm attaches to the motor case and stops it from turning.
(For my sins, I've run diesel engines against dyno's like this - Arthur Schenk AG was the dyno producer) |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 38
Registered: 12-2005
| | Posted on Monday, October 23, 2006 - 09:41 am: |
|
Gerald
You may be right as there may be a rpm that this method becomes ineffective. The unit pictured is a teaching aide but still runs aroud 4000.00. I ran across this link
http://www.innovatia.com/dyno_design.htm
and I haven't plunked down the 50 bucks but it seems a good way to start. It mentions limitations so that may be low rpm measurements. The fish scale cord wrap might be a good idea for low speed.
Dirk |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 39
Registered: 12-2005
| | Posted on Monday, October 23, 2006 - 09:55 am: |
|
I missed your post as I was writing mine. I like your idea of a braking motor for the load. An encoder would probably be better for rpm as when you lost steps it would show up but the pulse signal would keep on going.
Dirk |
Gerald_D
Registered
Username: Gerald_d
Post Number: 321
Registered: 11-2005
| | Posted on Monday, October 23, 2006 - 02:03 pm: |
|
Some classic U-cradle/torque arm constructions:
http://en.wikipedia.org/wiki/Dynamometer
http://www.cs.wright.edu/people/faculty/sthomas/stand2.html
http://www.magtrol.com/motortesting/pb_dynamometers.htm
http://www.piperdynamometers.com/dcdynamometer/dcdynamometer.htm
http://www.cpengineering.com/products/dynamometers.html
http://virtual.vtt.fi/virtual/mobile/rakewww/rake_www_hdvehicledyno.htm <---link for a laugh - brake not visible
http://www.dynesystems.com/Midwest/Midwest.htm
Prony (friction) brake ideas:
http://www.physics.lsa.umich.edu/demolab/demo.asp?id=483
http://www.me.utexas.edu/~lotario/me244L/labs/pmdc/pronybrake.html |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 40
Registered: 12-2005
| | Posted on Monday, October 23, 2006 - 03:39 pm: |
|
After giving this more thought. A standard motor or engine will loose RPM as a increased load is put on it. A stepper will not loose rpm because the driver dictates the speed of the motor, it will instead loose steps. I'm thinking at the moment that the only test that can be made is stall torque at different rpms. Am I missing something
Dirk |
Gerald_D
Registered
Username: Gerald_d
Post Number: 322
Registered: 11-2005
| | Posted on Monday, October 23, 2006 - 04:04 pm: |
|
As you said, the stepper will lose steps if overloaded, and that does mean a loss of speed - rather severely
I guess you would also keep an eye on current/voltage at the point just before it loses steps/speed to make a judgement of whether the motor can produce those speeds/torques continuously? |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 41
Registered: 12-2005
| | Posted on Monday, October 23, 2006 - 04:32 pm: |
|
My thinking is to apply the load at a given rpm until you loose steps and then back off the load until the motor will run for a given period with no loss of steps. I'm not sure you can reliably measure the current, but it's worth a try.
Dirk |
Gerald_D
Registered
Username: Gerald_d
Post Number: 323
Registered: 11-2005
| | Posted on Monday, October 23, 2006 - 05:09 pm: |
|
Dirk, a multimeter's "DC current" is apparently the true current in a stepper motor. When I tried to verify this snippet in my hairy memory, I found this thread that should interest you - a lot!.....
http://finance.groups.yahoo.com/group/geckodrive/message/567 |
Dirk Hazeleger
Registered
Username: Dirk
Post Number: 42
Registered: 12-2005
| | Posted on Monday, October 23, 2006 - 05:35 pm: |
|
Wow
That is copied and stored for safe keeping. Very good info indeed.
Dirk |
Dick van Randen
Registered
Username: Dvanr
Post Number: 26
Registered: 03-2006
| | Posted on Tuesday, October 24, 2006 - 05:13 am: |
|
Re the brake for the Prony.
How about using the disc brake from a mountain bike? Mount the metal disc on the motor shaft and the brake caliper on the arm to the scale? You'll need some kind of alignment jig to keep it all from going in orbit but thats details , details , details... |
Motors