Centering the Sub-Trim Debate: An In-Depth Analysis

This article is written by valued CK Aero contributor Earl Haury

Various opinions exist as to how subtrims affect servo travel, especially when used with dual servos on elevators. This discussion shows data that can help understand subtrims. While the measuring equipment was setup I made a few more measurement to satisfy my curiosity. A comparison of the performance of old with new servos, trim effect, and expo data are shown.

Correct geometry of linkages connecting the servo to the control surface is paramount. Subtrims should not be used to adjust surface neutrals at the expense of linkage alignment! Particularly elevator linkage with the typical dual servo system. Exactly matched elevator surface movement is a necessity and can only be achieved with exact linkage matching. Subtrim is a tool that allows us to adjust the position of the output arm of the servo and should only be used for that purpose.

Most servos have splined outputs and usually arms are supplied with the splines offset a bit from the arms so that they can be rotated to achieve the best alignment. Some prefer to use aluminum servo arms that don’t provide that option and we need some means to align the arm at neutral. This is where subtrims come in.

To explore this question I rigged a pair of new Futaba BLS173SV servos with a wheel to which is attached an angle bracket onto which Xicoy angle sensors are mounted. The Xicoy meters have 0.1° resolution, are a bit sensitive, work fine for a bench mounted rig, and can be set to zero at any angle.

To drive the servos I connected them to a Futaba 7008 RX via Sbus. An oscilloscope connected to one of the normal elevator channels provides a digital readout of the elevator channel pulse widths.

A Futaba 18MZ TX completes the system. For Subtrim data AFR was set linear with a 100% range, Ailevator elevator settings = 100%. Elevator function was set to the LST slider, using the detents to provide repeatable TX position output. ATV affect was explored at 100/155 and 140/155%. Trim settings were centered. Subtrim setting effects were measure at various offsets. For expediency servo travel in one direction only was measured.

Some things to consider beyond subtrims when setting servo travels. There’s a limit to the amount of travel available. Extending any combination of settings, ATV / AFR / Subtrim / Trim will be truncated to that maximum. A normal setup should never approach this limit. Historically we’ve been taught to use the maximum range of servo travel and TX pulse width range. Good advice, but less important than the “good ole days” when servos were clunky. Consider that travel time is also important, especially for snaps, less travel is quicker.

On to the data of this exercise. The first chart shows the relationship between subtrim settings in percent to servo rotation in degrees for a pair of new servos. Note that the rotation is approximately 0.065° per 1% of subtrim, significantly finer than can be obtained by using the servo splines. Also note that there are slight differences between servos, so a different subtrim to “square up” a pair of servos is likely.

The next chart shows the servo travels with subtrims set at 0, 50, and 100%. Note that the response is linear and the travel is reduced at 100% vs. 140% ATV. Also of note is that the servos match really well to around 45° of travel, then tend to diverge beyond that.

Another way to look at the subtrim effect is to reset the angle meter to zero after each subtrim change. This also shows a linear response with the subtrim offset.

What about different subtrims on a pair of elevator servos? This chart shows how the servos track when one is at subtrim=0 and the other 100%. Note the 6.5° change at zero when the angle meter was reset. Also note that these are used servos with 1000+ flights.

The conclusion from these data is yes, the subtrim will not affect the control surface travel when used to set the servo center. The process should be to install the servo arm onto the spline as near perpendicular to the pushrod line as possible, then use subtrim to fine tune it to exactly perpendicular. Don’t be tempted to adjust each elevator neutral with subtrims, fix the linkage instead! That being said, a small amount of subtrim to zero a surface is OK if one half turn on a clevis is too much or little.

OK, on to other observations. The next chart shows a comparison of performance of a pair of used BLS173SV servos at 1000+ flights (all E power) to that of a new set. Pretty remarkable!

Then a look at the effect of trim settings on the tracking of a pair of servos. This chart shows the trim authority when set to minimum increments. Measurements were made in 5% increments. Note that each trim % is about 3X that of one subtrim percent.

Finally a look at Expo affect at AFR settings of 50 and 100%. The upper set of curves are at 100% AFR, while the lower are 50%. From top to bottom of each set Expo is 0, -25, -50, & -75% respectively. Interesting that -25 expo delivers a relatively small change from linear.

You’re free to draw your own conclusions from the data. My take is that the Futaba 18MZ and BLS173SV servos provide performance beyond what most of us can appreciate. Subtrims won’t adversely affect servo travel, but if used to patch linkage alignment errors there will be an effect on surface travels due to those linkage errors.

When setting up an airplane with dual elevator servos I suggest this process:

Happy programming!