I have to study the behaviour of a PMSM drive with variable load. The problem is I have to find a way to apply this variable load being able to measure it. Thanks for any suggestion.
The current drawn by the motor is a measure for the torque of the load. If you want to be exact, measure the current drawn by the motor at no load and substract it from the value at the load you want to measure.
For the most accuracy, if you are using a "clever enough" drive, this torque can be read from the drive.
Sorry, but you've got me wrong. What I need is
some mechanical system that would allow me to
apply a variable load to the motor. A device controlled by a relay isn't good enough, because
it can help me only to apply a load formed by
impulses. Anyway, I don't need a very complex
device, just something simple and efficient.
What you're looking for is a dynamometer. If you have or someone you know has mechanical skills, you can make this with an old disk brake from a car and some other junk. Here's what I'd do:
1. Figure out a way to couple the shaft of the motor so it drives the brake disk.
2. Figure out how to connect a scale to a lever attached to the brake caliper (which has to float) so that you can get a readout of the force (example: a 1-foot lever connected to a pound spring scale will let the scale read out in foot-lbs).
3. Figure out a way to pressurize the caliper so that you can vary the load by varying the pressure. You may be able to do this just with your foot on a brake pedal, depending on how accurate your measurement needs to be!
How you do all this depends on where you live; I'm guessing from your names that it's Romania, and that Romania is probably pretty close to where I live (Costa Rica) as far as what you can easily buy. So you will have to be a bit creative, maybe make some friends in auto junkyards or small machine shops to find what you need and/or get it built.
Alternately, you may be able to find an old hydraulic pump and a pressure relief valve. You'd connect the inlet to an oil reservoir, and run the output back to the tank through a small orifice. You could then adjust the load by changing the relief valve pressure setpoint. For that matter, you could probably even use a centrifugal water pump, pumping water through an
orifice if the RPM band you want to test is fairly narrow. I don't know offhand how you'd calibrate it, but I suppose you could do it empirically.
Realize that all the energy that the motor is putting out will be immediately converted into heat by these systems, so be careful to not burn
yourself or other things up!
I'm happy to help more if this isn't enough info or if you need other ideas.
On July 5, 2002 02:41 pm, Willy Smith wrote:
<clip>
> What you're looking for is a dynamometer. If you have or someone you know has mechanical skills, you can make this with an old disk brake from a car and some other junk.
<clip>
> Alternately, you may be able to find an old hydraulic pump and a pressure relief valve. You'd connect the inlet to an oil reservoir, and run the output back to the tank through a small orifice.
<clip>
I believe that small commercial dynamometers typically use hysterisys brakes, and larger ones use hydraulics. There is of course also the prony brake, which consists of a metal cylinder with a leather belt around it acting as a friction band. As the belt is tightened, the friction and therefore the applied torque, increases. If the metal cylinder is hollow with a lip about the open end, a certain amount of water can be placed in it to act as a coolant. This method is often used in student labs as it is fairly inexpensive and simple for small loads. Another means is to use a DC motor and regenerative drive operating in current control (torque) mode.
The problem in all cases is to measure the torque accurately. There are in-line torque transducers which can be placed between the brake and the motor under test. Commercial dynamometers normally have a very carefully
designed mechanical system to avoid imposing unmeasured loads. Measuring the torque accurately will likely be the most expensive part of the exercise. If you don't need to measure torque but just apply a variable load, then of course this problem goes away.
> I don't know offhand how you'd calibrate it, but I suppose you could do it empirically.
If torque is measured directly, then the dynamometer can be calibrated by locking the brake and attaching a balance beam to the end of the shaft. Attaching a calibrated weight at a known radius will accurately apply a torque which can be calculated. When using this method, it is very important that there is no significant unmeasured load (e.g. from bearings) in the system.
> Realize that all the energy that the motor is putting out will be immediately converted into heat by these systems, so be careful to not burn yourself or other things up!
This point is very important. The amount of heat which can be generated from even a small load can be quite significant. If for example, a prony brake is allowed to run dry (the water will get warm and evaporate), it will get quite hot very quickly.
--
************************
Michael Griffin
London, Ont. Canada
************************
Responding to H. Ozevin's Fri, Jul 5, 9:50am advice:
Caution! The no-load current will have a relatively large reactive component because of the magnetizing current requirement. Therefore, the proposed method will introduce large errors.
A common way of doing this is with a hydraulic pump and a restrictor valve. Varying the valve opening smoothly changes the load on the motor.
The fluid is simply recirculated. This may require cooling for long term tests as the power has to go someplace. A 5 gallon pail works well for small motors. The pump folks can recommend the best pump and fluid for this application. Most other mechanical means suffer from heat concentration, the pump idea makes it fairly easy to get rid of the heat. I have used an automotive power steering pump and hardware store items to test small motors very inexpensively. This
method is used extensively to test automotive starters with a larger pump.
Regards
cww
--
Free Tools!
Machine Automation Tools (LinuxPLC) Free, Truly Open & Publicly Owned Industrial Automation Software For Linux. mat.sourceforge.net.
Day Job: Heartland Engineering, Automation & ATE for Automotive Rebuilders.
Consultancy: Wide Open Technologies: Moving Business & Automation to Linux.
Facebook
Google
GitHub
Linkedin
