Need to apply a variable load to a DC motor

[jmarkwolf]

I'm fabricating a dynamometer for D.C gear motors (automotive type seat control gear motors).

I've implemented a go-kart disc brake to apply a variable load on the motor, but it is not linear, nor constant.

I'm wondering if I can replace this mechanical brake with another D.C. motor wired as a generator, and vary the load electronically to affect a controllable linear braking force.

Can anyone on the forum advise and/or offer a link to some useful data for doing this?

[Skimask]

What rpm range?

[IanB]

To improve linearity you probably want to include feedback. For instance measure the torque on the motor and then vary the breaking force in a control loop until the measured torque equals the set point.

[jmarkwolf]

The motors are gear motors so speed is very slow, maybe 10-30rpm, with high torque.

May need to characterize higher speed motors as well.

Sorry I can't be more precise, this project is a "moving target".

[mikerj]

You'd need to gear that up quite a bit to use another motor as a load.  Other options are a water brake, air brake or eddy current brake.  Something from an old rowing machine might be usable?

[rstofer]

It seems like you are trying to build a Prony Brake - Google around, there are quite a few hits.
Basically, the mechanism holding the brake calipers needs to go to a weigh scale so torque can be measured directly.
Horsepower is a function of torque and RPM

Yes, you can use another DC motor as a generator and dump the power into a resistor load bank.  But, really, that seems like the long way around although it has the advantage that MOSFETs can be used to control the resistive load.  Given some kind of microcontroller with PWM output, this could make a very nice dynamometer.  I'm pretty sure the Smock Check dynos use something like this.

If you build up a mechanical Prony Brake, you can add a control loop to increase/decrease braking while collecting RPM and torque.  You get linearity by using the torque feedback to control the brake force.

[dmills]

Stepper motor, loads of poles so it generates well even at very low speed, and the two 90 degree windings mean that of you load each equally with a resistive load you should get constant torque Vs angle (sin^2 + cos^2 = 1).

Regards, Dan.

[max_torque]

Work out what Torque and Power you need to absorb, across what speed range first!

Plenty of options for the controllable brake, for large truck alternators (geared up via polyV belt) to hydraulic pumps etc, but you need to know the speed and load range before selecting any of them!

[calexanian]

Automotive alternators work very well for this purpose. I recall a scheme where they used a fixed resistor load and varied the excitation field current to vary the net load.

[_Wim_]

Why not use the same type of motor as a brake as the one you want to test? Put a variable high power resistor in between, and you have an adjustable brake. Put a multimeter in series (current measurement) with the resistor to monitor the braking torque (torque= motor contant x current).

[Berni]

DC Motors make pretty nice brakes. I would also suggest using another one of the same motors.

They can also be used for braking at low RPM if you actually push current backwards trough the motor using a power supply. The torque of a DC motor is proportional to the current trough it so theoretically using a current source would provide constant torque. That way you don't need a external mechanical control loop, but you would want to have some form of weight scale torque sensor mounted to the motor just so you can see what its doing.

[Richard Head]

For a very low RPM load you could opt to simply lift a reasonably heavy weight a few feet off the ground. You could attach a pulley to the end of the shaft and wind twine around it and then connect it to the weight. It's not dissipative but will still load the motor by a very precise and reproducible amount. You can also accurately calculate the torque output and power output efficiency etc. using basic physics.

[joeqsmith]

I'm fabricating a dynamometer for D.C gear motors (automotive type seat control gear motors).

I've implemented a go-kart disc brake to apply a variable load on the motor, but it is not linear, nor constant.

I'm wondering if I can replace this mechanical brake with another D.C. motor wired as a generator, and vary the load electronically to affect a controllable linear braking force.

Can anyone on the forum advise and/or offer a link to some useful data for doing this?

I have played with a few motor loads. 

My seats move very slow and I can imagine that trying to measure at the gear drive's output shaft will be a tough problem.  Very high torque and low rotational speeds seems like it will really limit what you can use for a load.   I assume you want to measure shaft speed and torque.  Any idea on the accuracy/precision you need?    Personally, I think what you have now sounds like the right direction.  You just need a way to control and measure it.   Post a few pictures of your rig. 

I wonder if you post in some of the mechanical engineering forums if they may be able to help.

[_Wim_]

I really think a dc motor as a breake would be the way to go. DC motors have a reasonable stable "torque constant", that defines the amount to torque generated/absorbed vs the current through the armature.

I do not have an electronic load (like a BK precision 8600), but it should be possble to connect it to the motor in constant current mode (for a fixed torque independant of the RPM) and in constant power mode (for a fixed braking power). By using the same motor with reduction, in theory it should always have enough braking power, as some energy is lost in both motor reductions, so the driven motor can never produce more torque than can be absorbed by the braking motor...

[Dave]

Automotive alternators work very well for this purpose. I recall a scheme where they used a fixed resistor load and varied the excitation field current to vary the net load.

Fantastic suggestion. Add a torque sensor on the shaft and make a PID control loop for controlling the excitation current based on the torque. It should work perfectly.

[Smokey]

What you need is a current mode servo amplifier and a servo motor with feedback.

The real question is how much are you willing to spend to get the best solution?

For like $500 you can get new stuff.  I can think of a few types of systems you could piece together from ebay for around $150 with the added risk of untested stuff.