Does reducing throw increase servo load?

[astroboy]

In planning retracts for my current project, I thought I had figured out correctly how much servo throw was needed; I made aluminum extensions for the servo arms. However, when I tried the linkage with my Futaba 9C, I had to set the endpoints to around 40% in order to get the proper gear leg motion.

My question is: Do the servos have to work harder because they're using less of their available throw? I don't know how servos work(these are Cirrus CS26BBs), but if they move in steps, does setting the endpoints so low reduce the size of each step, or just reduce the number of steps used? If it's the latter, it seems that the load on the servo would be increased, as it's having to move the gear leg farther with each step than if the endpoints were set higher. I had done a test of the force required to move the retracts, and they're probably still all right as they are, but I'm wondering if the margin of safety has been greatly reduced. As you might surmise, it would be something of a hassle to change the servo arm length, so I'd like to avoid it if possible.

Thanks,
Jeff

[pmpjohn]

Jeff,
It is probably a small point as it is retracts that only get used twice a flight and not a control surface that is in constant use. As a general rule it is better to use a shorter output arm and more of the available servo travel. This takes load off the sevro gear train and off the power required to run the servo motor. Less power to run the servo = more power to run the drive motor. In the case of a control surface it also give a greater amount of precision in the positioning of the surface. ( More steps being used. )

John

[Bill Glover]

Another general point (again, not so relevant on a retract installation) ... if you use a longer arm than necessary you are amplifying the effect of any free play in the servo gearbox, giving a little extra slop in the setup.

As John said, ideally you should aim to use 100% servo throw and adjust the linkages to get the movement you require.

[green66]

If you aren't changing anything else, then yes - servo load (torque) will increase in proportion to the added servo arm length, thus current drain will increase by the same percentage. This is simply due to the increased lever ratio (nothing to do with the particular servo technology, electronically reducing travel, etc).

But like the others said, since it's for retracts it shouldn't be a problem since the application is intermittent duty. Just ensure req'd torque plus some margin is within servo's torque spec.

[astroboy]

Hmm... I was just about to go in and change the linkage so that it used more of the travel, but now I'm not so sure, since all three of you are saying it's less of an issue with retracts. It is still a concern, since if the margin of safety is, well, marginal, the risk of stripping gears is there. I guess I need to figure just how much the load has increased. I'm not at all sure that it's as simple as 100% divided by the 40% the endpoints are currently set at. As it is now, the servo travels through about 40 degrees of arc to operate the gear. I always thought that the nominal arc for a servo is 60 degrees, in which case I'm at more like 67% of the nominal throw, and not 40%, which I'd feel much better about. I checked the enpoints of the setup with my Flash 5(instead of the 9C), and found that the enpoints had to be set quite a bit higher. I don't recall exactly, but I believe they were in the range of 60+ per cent. So I wonder if the difference is in the 9C expecting the servo to have more of a range--or something like that?

Thanks,
Jeff

[green66]

Try to measure the force (in ounces) required to cycle the retracts in both directions, using a postal scale, or known weight on the pushrod, etc. Required servo torque will be the measured force multiplied by the distance (in inches) between the pushrod attachment point to the servo's axis of rotation, measured perpendicularly to the pushrod. Then add about 20% safety margin - Your servo needs to have at least this amount of torque capability.

When considering servo torque reqmt's, forget about everything related to transmitter adjustments or programming, e.g. limit points, travels, etc - These have no effect whatsoever on the servo's torque capability. The only relevant variables are the two above (force and moment arm).

If you have large travels, steep servo arm angles, or pushrod force that varies during the stroke, the max torque could occur anywhere along the servo's travel, therefore should calculate torque at several points along the servo's travel.

Hope that clarifies/helps.

[astroboy]

Green66:

In that case, I've got it made. I basically did what you've suggested before I started the installation. In an admittedly crude test, I mounted the retract vertically in a vise, with the arm extending straight up, and gently set weights on it until it moved. I kept the weights as directly over the arm as I could. I don't recall exactly what it took, but as long as I kept the arm vertical and didn't let it tilt to one side(meaning misaligned with its axis of travel, I think it was somewhere around 12 ounces. Since the CS26BBs are rated at 25 oz/in, and the pushrod is about 13/16" from the servo's axis, it should be good for nearly 31 ounces of force--which I guess is kind of backwards to the way you've stated it, but still the same thing.

Thanks,
Jeff

[green66]

That's right. Servo should handle it easily.
If the 12 oz force is fairly constant through the whole pushrod travel, the worst case (highest torque) will be when the servo arm and pushrod are 90 degrees to each other, at which point the max req'd servo torque = 12 X 13/16 = less than 10 oz-in ==> fat margin :)

And keep in mind that 25 oz-in is the servo's torque spec, not a damage limit; gear strippage probably won't occur until, I'm guessing, up around 40-50 oz-in.

[astroboy]

Originally posted by green66

And keep in mind that 25 oz-in is the servo's torque spec, not a damage limit; gear strippage probably won't occur until, I'm guessing, up around 40-50 oz-in. [/B]

That's the kind of thing I LIKE to hear!

Thanks,
Jeff