Windshield wiper motors have attached gearboxes, and their ratios are
seldom known. Motor current alone will not provide the torque at the
output of the gearbox, since the gearing will increase the torque. Even
if you do know the gear ratio, there are unknown friction losses in the
gearbox, and the torque is always less than what is implicitly derived.
Explicitly measuring the torque is the best method.
For the previous respondents: a "Prony brake" is a dynamometer, like the
kind used to measure the torque of the driven wheels of a car. Many
physics classes spend a unit using a desktop Prony brake to measure
output of smaller motors. You can make one using parts from a hardware
Explicityly measuring torque is "a" method, and by no means the best all the
If you are testing the characteristics of a motor, it may be the best way,
but you can't do that during operations. The original poster explicity said
"Is there a way to meaure torque, other than using a pony brake."
I have suggested one, is it the "best?" That can not be known until other
facts are also known, for instance:
(1) Will torque measurement need to be made under active load?
(2) What is the accuracy needed?
And what would that factor be? 30%? 50%? 70%?
I have to wonder why you didn't mention that to determine torque at the
output of a gearbox you'd also need to know the ratio of the gearbox,
and have at least some guess at the gearbox efficiency.
He has a motor and wants to know the torque. How exactly are you proposing
to find out the "characteristics of the motor" other than by measuring the
torque? Some kind of model relating the motor windings and magnetic fields?
If he knew the motor charateristics (speed/torque curve, time constants) he
wouldn't need to measure the torque other than to determine the gearbox
ratio and efficiency. But he has a junk wiper motor so you know he doesn't
know the motor characteristics.
The assumption is you need to know the torque. The question is how and in
what application. He may need a static measurement, but when he said
"without a pony brake," I read that to mean a more dynamic -- less
intrusive methodology. That was my take, maybe I'm wrong.
Easiest way to calibrate torque measurement:
Varible output power supply
Attach the measured rod to the motor.
Attach the fish scale to the other end of the rod.
Measure the exact distance between the fish scale and the center of the
motor shaft. (Lets say it is exactly 12 inches)
Apply power to the motor, read the fish scale. If it says 1 pound, then you
have one foot/pound.
Increase the current to the motor, read the fish scale.
Make several measurements along the current range of the power supply to
satisfy yourself that it is linear.
You will see the current/torque behavior of the motor.
During run time, all one needs to do is read the current being used my the
motor to determine the torque.
If you don't have a current meter in your robot, you can measure the RPM of
the motor, against the PWM of the supply, and assuming you've figured out
the power transfer characteritics of your amplifier, you can have a pretty
good representation of the torque.
I don't understand how that would be done either. If he measures the stall
torque and the no load speed he can draw a line and have the curve he needs
for any situation - dynamic or static.
The rest of your description of how to measure the stall torque is the
detailed version of what Gordon said to do.
Right and if he needs to know the current torque being applied to a load, he
would be able to calculate it. Many times an application needs to know the
amount of torque being applied for safety reasons.
As I said, without more information about the exact application, only he did
not want to use a pony brake.
It's called a Prony brake (not pony brake), and it's much simpler than
this setup. The result is a torque taken at operating speeds and loads,
and is not static. You don't need to attach a rod to the shaft, you
don't need a meter, you don't need a variable power supply. You put a
leather or heavy fabric belt between two fish scales, and run the motor
at 12V -- low or high speed winding, your choice.
He can run the strap through the pulley he intends for the winch, or set
the brake on the output pulley of the winch, which will then also
include any further gear reduction provided by his winch setup. And
compensate for any further frictional losses inherent in all winches.
We can assume the OP is trying to simplify things and avoid a mechanical
measurement. You're still talking about a mechanical measurement, and
one that is more involved than what high school physics class students
do all the time.
The other option is to look up the name on the motor (Trico? Bendix?)
then go to a car parts store to see if they have a catalog with specs.
The specs might even be on the Internet.
LOL, a consistent typo. I know it was Gaspard de Pony, I'm sure of it!!
That's funny. My fingers don't always type what I tell them too, and my
eyes don't always read letters right. Dyslexia is not without humor.
Maybe, I read it as perhaps needing to do it during an active load. There
are application where monitoring the torque applied to the load during
operation is a requirement.
See if you'd agree with me:
1. Estimate the weight of the object(s) to lift.
2. Get some kegs of equivalent weight, filled with brew, of course.
3. Invite some crazy friends over.
4. Forget scientific measurements and just use the rig to lift the kegs.
If the rig fails, the kegs crash to the ground, and everybody has a good
time as beer goes all over everyone. Example of turning a negative into
If the rig doesn't fail, the beer's still in the kegs, waiting for you
and your friends.
IOW, success either way.
: (1) Will torque measurement need to be made under active load?
--Doesn't have to be, but that would seem to be best, yes? I could
use a weight suspended from a cable to simulate the load it needs to lift.
: (2) What is the accuracy needed?
--Not sure I know; i.e. I'm lifting something that weighs, say, 20
lbs; all I'd like to know is whether or not I have a reasonable (50%?)
safety margin and that I'm not about to burn out the motor during repeated
--Getting ideas myself here, but I'm purely mechanical in my
thinking; i.e. I can see putting a 30 lb weight on the cable, then
listening for groaning and feeling for temperature increase in motor
housing but that's not "accurate" enough, yes?
"Steamboat Ed" Haas : For some reason hung up on
Hacking the Trailing Edge! : Mexican Oompah bands...
Since you mentioned a winch in your first post, you can always rig up a
block & tackle arrangement, where even a puny motor would lift 20-30
pounds, without breaking a sweat. It all depends on how fast you want
You can use some simple math to roughly calculate the requirements, and
since these motors are $15 or so on surplus market, it makes sense just
to try it out in a real-world test when you've done the initial
Anyway...Most windshield wiper motors run about 80-100 RPM no load on
their fast windings. If you fasten a 6" cable drum onto the motor shaft,
you'll be pulling about 19 linear inches for every revolution of the
motor. Even at a loaded speed of 60 RPM that's 19 inches per second. Is
that fast enough? Too fast? You can run the steel cable through some
pulley blocks to increase torque (and decrease linear speed) if needed.
Mikey (Kap'n Salty) posted a link to someone's page for a Fomoco
windshield wiper motor. Looks pretty standard to me -- fast and slow
windings, worm gear drive, etc. I think you'll find the torque they
tested the motor at is well within what you're looking for, perhaps
as-is. Your motor is likely pretty close, because windshield wipers have
to provide the same function from car to car.
: Anyway...Most windshield wiper motors run about 80-100 RPM no load on
: their fast windings. If you fasten a 6" cable drum onto the motor shaft,
: you'll be pulling about 19 linear inches for every revolution of the
: motor. Even at a loaded speed of 60 RPM that's 19 inches per second. Is
: that fast enough? Too fast? You can run the steel cable through some
: pulley blocks to increase torque (and decrease linear speed) if needed.
--I'll be using a smaller drum; probably 3" dia, but as you say
fast enough. Length of pull will be small, probably 4ft at the most, but
speed can be varied, which bring up another little conundrum.
--At the moment I'm controlling motor speed with a little program
I wrote for the Basic Stamp; basically ramping up to a speed, holding it
for a while, then ramping down (not strictly necessary for the
application; just testing my ability to write the program). I'm outputting
the signal to a Victor motor controller; i.e. one that's intended to be
used with an R/C reciever, so it accepts the same inputs as a servo. What
would be ideal would be a way to control speed of motor (equivalent to
position of a servo output arm) and direction with a pot. I'm looking for
links to a tutorial on how to do this; got any pointers?
: Mikey (Kap'n Salty) posted a link to someone's page for a Fomoco
: windshield wiper motor. Looks pretty standard to me -- fast and slow
: windings, worm gear drive, etc. I think you'll find the torque they
: tested the motor at is well within what you're looking for, perhaps
: as-is. Your motor is likely pretty close, because windshield wipers have
: to provide the same function from car to car.
--Yeah, mine is either a clone or identical I suspect.
"Steamboat Ed" Haas : For some reason hung up on
Hacking the Trailing Edge! : Mexican Oompah bands...
Are you wanting to provide R/C-type pulses to the Victor controller? You
can do that with a 555, for example. Or you can have the Stamp provide
the pulses, using PULSOUT. (Am I understanding your question?)
I'd be surprised that Victor doesn't provide some insight on this, as
their controllers were heavily used during the combat robot craze, and
people were adapting them for all sorts of applications.
: Are you wanting to provide R/C-type pulses to the Victor controller? You
: can do that with a 555, for example. Or you can have the Stamp provide
: the pulses, using PULSOUT. (Am I understanding your question?)
--Correctamundo. Figured out how to do what I want it to do a
couple of days ago; chapter 4 of the Parallax book "What's a
Microcontroller?" covers it. Took a *lot* less programming than I had
feared. Now I can control rotation speed and direction with a *tiny*
pot. Next step will be to do something a little more robust that can
survive a "real world" situation, then build it into the art car.
: I'd be surprised that Victor doesn't provide some insight on this, as
: their controllers were heavily used during the combat robot craze, and
: people were adapting them for all sorts of applications.
--Yeah, that's where I got mine; salvaged from a middleweight
that never saw combat, thanks to the big-ass lawsuit between RW and BB.
Now all I got to do is figure out a few more projects, to use my three
remaining Victors! :-)
"Steamboat Ed" Haas : The other night I
Hacking the Trailing Edge! : dreamed about wasabi...
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.