Digital Tool for Measuring Radio Control Servo Torque

[sunworksco]

What kind of digital tool can be used for measuring torque on a radio control servo?

 

[mu38&Bg#]

Digital force gauge or torque gauge. Both use a load cell.

 

[sunworksco]

The tool needs to hook up to the servo arm. The servo on the right is the one.

http://www.theamya.org/hints/sailservo.php

Maybe this meter.

http://www.huatest.com/en/product_view_13_138.html

 

[ksouers]

You could use a cheap digital scale from Harbor Freight. They display in metric or Imperial units plus a couple others.

Of course you'll have to do the math to convert to Newtons or ounce/inches but a pretty cheap way to get started.
I think I paid around $10 USD for mine, maybe less.

 

[Tin Falcon]

The tool needs to hook up to the servo arm. The servo on the right is the one.

Many balances what we call in the industry top loaders have a place to put a hook on the bottom of the balance this is normally used for a pan suspended in water and used for specific gravity testing but could work for your testing as well.
Tin

 

[petertha]

If you are just testing for look-see values, a weigh scale rig or even suspended weight rig will suffice. If you are wanting more definative values or comparing different servos to manuafcturers specs in a closer bracketed range, other considerations start to enter the picture, primarily 'dynamic' torque vs 'holding' or 'stall' torque. This topic comes up a lot in RC. Also, for meaningful comparison, you need to standardize on the input voltage(ideally regulated) and max current draw under load as this can have an effect on torque values. Most typical RC servos will perform better at elevated voltages... to a point. Analogs behave different than digitals & gear sometimes cheapy gear trains fail before the motor cooks.

Some Google links
http://media.hyperion.hk/dn/at-ser/atservo.htm
http://www.rcuniverse.com/forum/m_1402077/anchors_1402077/mpage_1/key_/anchor/tm.htm#1402077
http://www.trailofdestruction.com/servoguide.htm

 

[sunworksco]

I am building a radio control winch for sailboats. It uses a 6-12 volt dc motor that drives a geared transmission with an output shaft that turns at 300 degrees. The output shaft is a 5K potentiometer that is connected to a digital motor controller. I'm going to build 12 of these for my fellow sailboat club members. I would like to accurately test the winch for torque. I have all of the electronic data devices to test the amp, current and voltage of the motor.
The torque reading is very useful for calculating how many square inches of sail can be winched in.

 

[Swede]

As suggested, why not hang weights from a known moment arm and see if the winch can lift and/or hold the mass? Why go digital? It seems like overkill.

The only problem with this approach is that you can potentially damage the drive train, although I suspect the motor will stall before gears are stripped.

If you are making a dozen, perhaps one can be sacrificed in destructive testing to see what it can do. The same problem exists with a digital device... the winch must be 100% loaded to get a reliable reading, and loading the winch to capacity might cause damage.

 

[sunworksco]

I need to test each winch to document the winch model and also before each one is sold.
The test needs to be for speed, newtons and ounces/inches.
I need it for other designs, as well.
Would also like to video the test and show it on YouTube and link it to the future website.
The winches will retail for around $200.00@, so not sure if I want to destroy any.
The closest similar winch in ounce/inches is an Australian drum/pulley winch that retails for around $250.00. The geared arm winches are more reliable and have more power for pulling the sheets in.

 

[mu38&Bg#]

How fast are these? Measuring speed could get complicated if they are fast. If you're developing a product measuring is certainly worthwhile. But, I would certainly include a destructive test myself. Having been around RC servos for quite a few years, the ones that are rated near their ultimate limits are quickly weeded out by hobbyists, either by thorough testing or misapplication.

 

[sunworksco]

The speed to pull the 1,000sq.in. sail completely in is less than 3 seconds.
The sail winch that i'm building is like a super-sized RC aircraft servo but with an exoskeleton made in aluminum. The motor is an MFA DC 360 http://www.mfacomodrills.com/motors/360.html
The geared transmission is self-braking.
The gears are made out of aluminum anodized stain-less steel.
The motor controller is PC programable for position and speed and can use up to 12 volts.
Believe me, I have used the winch on over 1000sq.in. of sails and the winch is bullet-proof.
I will market the winch for up to 1,100sq.in. of sails and anything over that can use a winch for each sail. I'm planning to use 2 winches in my 1/6th scale J-Boat.

http://www.ludwigrcyachts.com/j_class.htm

 

[petertha]

.. like a super-sized RC aircraft servo but with an exoskeleton made in aluminum. The motor is an MFA DC 360 http://www.mfacomodrills.com/motors/360.html

Wow, those are very nice boats.

Dumb question, but that motor you referenced with 360 gram-cm stall torque (= 5 in-oz), how does that translate to commercial servos like this one below at 200 in-oz? Is it a 'multiplier' type thing through your propsed transmission/gears?
http://www.ehirobo.com/shop/product_info.php?products_id=6666

Back to testing, I really think the 'pull a weight on a line' may offer the simplest & best solution from everything you described.
- a temporary pully disk wheel (vs an arm) would take care of the constant radius through full rotation & translate into a numerical torque value, in-oz, ft-lbs whatever
- you could monitor voltage through your power supply & current draw with a DCA clamp meter
- you would also determine lineal line draw simultaneously which seems to be important
- you could start with moderate weight, see how things go, then ramp it up incrementally to where the magic smoke comes out of the motor, your choice

Anything 'digital' I can think of like a weigh scale is more suited to measuring force directly, translatable to torque with som egeometry yes, But teeny displacements, not really your application.

 

[sunworksco]

The combination of the larger motor and gear ratios of the transmission makes it very powerful.
I already have a motor data recorder that downloads to my PC. It records everything, including temperature of the motor.
I will consider the counterweight method that you are recommended.
That is a good starting point.

 

[doubletop]

If you are going to make so many and sell them for that price don't you owe it to your customers to do some destructive testing and then de-rate the spec by a suitable margin. If they want to over drive them, then its their problem.

Doing the destructive testing and instrumenting the electrical load you could use the data to design and build a current limiter so they can't be overloaded mechanically.

Up to you really

Pete

EDIT - I've just read the motor spec, it gives you the torque, apply whatever gear ratio you are using to that and you have the torque for your winch. Prove it with the methods suggested. Am I missing something?

 

[Swede]

EDIT - I've just read the motor spec, it gives you the torque, apply whatever gear ratio you are using to that and you have the torque for your winch. Prove it with the methods suggested. Am I missing something?

I was thinking the same. With a known motor torque, a known gear train, the final torque should be easy to calculate, subtracting a few percent for frictional losses in the gear train. Given the motor's rated torque recalculated for the gears, a safe way to go would be to derate that value by some percentage.

 

[mu38&Bg#]

With enough data on the motor you could calculate the torque, current, and heat lost in the motor. I the the spec sheet provides just enough if you can manipulate the figures.

I like Peter's idea with a pulley and weight. You don't need a load cell at all. If you watch the current or servo signal PWM on your data recorder you'll have time assuming it collects the data into a file against time somewhere.

 

[sunworksco]

Thanks for all of the intelligent advice.
It will be very interesting to do the test with weight on the end of the winch line.
Kindregards,
Giovanni