One of the jobs I have "working" for my son, is making torque meters for his rubber band model airplane motors (yeah, they call 'em motors). I've been making them to his specs, specifically the length of .039 wire (18" long ) to wind around the inside to be compressed to turn the needle so you can see how much torque is applied to the "motor." To make a long story short, he claims that the meter is off, that it doesn't read as high as it actually is. It seems that 3 inch ounces will probably break the motor and he's winding to 2 1/2 inch ounces and they break. There seems to be no commercial unit that will accomplish this. I set myself a goal, build one. I have a tiny scale that reads in grams, and every 28 1/2 (or so) grams is an ounce measured one inch from the center of an axis. I've been working on this for a couple of weeks mentally and three days physically. and here it is.
The torque meter, the cylinder on the left, (with a hex handle for turning (this is replaced in practice by the winder) turns the body clock wise, the spring inside gets twisted and in its turn turns the output shaft/hood (riding in a pair of Oilite bearings) in the middle of the contraption. The little lever in between the two big bearings presses straight down on the scale, reading in inch grams. There is a pointer on the body of the scale which is stationary relative to the body, allowing the user to mark where ounces are on the body. Ultimately a printed tape is applied to the body so the user can read directly on the scale.
Having achieved my goal of the making a gauge, I now know that the 18 inch wire should be 19 5/16 long to read 8 inch ounces on one revolution of the body., It currently reads 8.6 in oz.
The back side (toward us) of the body is resting on a pair of roller bearings, captive inside the block, concentric with the other shaft. This is my calibration device.