Linked Control Columns & Loading – Part 2 – The Pitch Axis

Last Updated on April 4, 2026

It has been a long while since I wrote Part 1. Hard to believe. It took time to get things finished up and generally working, and then the rest of the sim had to be built back up on top of it. Throw a bit of “life happens” and dips in motivation and time just flies by.

In Part 1, I talked about my thought process and general design idea. I had gotten most of the mechanical engineering figured out, installed, and had begun testing the control loading software with my design.

Now that I have a mostly complete working design, we can talk more about the actual build, so in this post I’ll cover the supporting structure and the Pitch Axis.

When I get to it, Part 3 will go into the Roll Axis build, and finally Part 4 will talk about the software side.

Where I’m at now

Before I talk through what I’ve done, I want to give you a taste of where things stand currently.

The following picture is a few months old now (as of end of 2025) and I’ve since made further progress. But, it gives you an idea of how the working columns ended up with the new floor and the shell back on top. I’m quite happy with the final result, which looks reasonably close the real thing.

Building the Support Structure

Due to ceiling height limitations in my garage, my sim floor height can’t be more than around 12″ above the garage floor. Taking into account the thickness of the sim floor material (3/4″ plywood), the underfloor box bottom (another 1/2″), and having to leave some room underneath for clearance to be able to move the base around (around 3/4″), I had around 10″ of usable space for the mechanism. That’s a pretty tight space and that drove some of the design choices you see.

For reference, a typical 737 mechanism under floor ideally needs at least 18-24″, so obviously a full OEM solution wasn’t practical and a different design was needed.

I settled on a box structure using 2″ x 10″ lumber sidewalls and a 1/2″ plywood floor. It’s VERY heavy, but also very solid and not easy to tear apart or shift around. I wanted to make sure it was built robustly since this thing was going to be under the floor in a difficult to maintain location later.

Later, this box would be raised off the floor slightly and screwed into a space in floor structure (which would prove to be an interesting challenge later due to its weight and size). I decided to build a separate box structure versus completely integrating into my frame build. This was so that later on I could move the entire column structure forward or back within the floor frame until I found the proper final location relative to the instrument panel and cockpit shell.

Parts Used

• 2″ x 10″ Treated Lumber for Sidewalls
• 1/2″ Sanded Plywood for bottom
• Kregg Pocket Hole Jig for creating solid wood corner joins. Fantastic product if you haven’t used one.

Building the Pitch Axis

The pitch axis was probably the easiest to build. I was fortunate to acquire an OEM 737 NG cross-tube with its supporting bearings, weighted cable arms, and “jam breaker” springs, along with the two OEM columns and associated hardware.

Side note:  Each cable arm provides control of separate pitch mechanism.  Should one side jam, the other side can still operate the redundant system, albeit with more push/pull force required to break through the spring pressure.  Surprisingly, though, not a lot of force is required to break free.  

For a sim,  you don't need a break-free spring mechanism, but removing it isn't a viable option (older NG cross-tubes don't have the springs).  

Obviously, the cable arms had to go due to their size and my limited space. I carefully hack-sawed them off using a hand saw, which then allowed the assembly to fit within the confines of the supporting box structure, and freely rotate on its bearings without binding or catching.

To support the cross-tube, I designed a pair of 1/4″ aluminum brackets in Fusion360, then sent it off to SendCutSend to precision cut and bend for me at a very reasonable price.

After some fine-tuning of alignment, you can see the final mounted cross-tube below:

Now, I needed to attach a motor to provide pitch control of the columns and torque for the feedback.

Thanks to the significant and instrumental work Fabian of Orange County 737 Sim had already done on his Control Loading solution, he guided me toward a 220V NEMA34 90ST servo motor and driver (aka control unit) combo. I found them on Alibaba at around $250 USD per axis. The 90ST is quite powerful, to the point you can apply enough torque to where you can’t move the columns at all.

A heavy-duty planetary gearbox was also recommended to provide the torque and smoothness needed for the long moment-arm on the control columns.

Make sure you note the input and output shaft diameters so you buy the right fitted gearbox for your motor. The input and output shaft diameters of the gearboxes are not always identical.

Note:  The 90ST suffers from something called "torque ripple" due to the way its internal armature is designed.  It becomes noticeable at higher torque values where you'll feel slight "bumps" in the pitch movement, like the teeth of a gear but subtle.  Normally you don't need to go that high in torque, but it's something to be mindful of.

With the motor and controller found, I needed to mount it and link it up to the cross-tube.

I designed a motor “shelf” in Fusion360, then sent it to SendCutSend for fabrication and bending using some 3/16″ aluminum. The shelf is bolted into the structure bottom and to the sidewall to prevent any chance of bending or movement.

The motor shaft itself is keyed, so I found a keyed Taper Bushing on Grainger that matched the gearbox output shaft diameter. Once these bushings are tightened, and by virtue of the key, there is no chance of slippage. Which is important since it’s easy to put quite a load on the shaft due to the long moment-arm of the columns.

I then designed a custom A-Arm in Fusion360 that bolted onto the hub, and had SendCutSend fabricate it for me out of 1/4″ powder-coated aluminum.

Fortunately, the OEM cross-tube had built-in A-arms, one of which I was able to leverage. I created an adjustable linkage using a hexagonal connector with threaded ends, and a couple of threaded shaft ball-ends I found on McMaster-Carr.

I added some bump stops to prevent the columns from traveling beyond their OEM spec limits. Essentially, I created some simple adjustable stops out of L-brackets and wood block that the linkage ball ends would bump against at max travel. I used a digital angle gauge on the mounted columns to get the column deflection.

Finally, I connected a CALT Draw Wire potentiometer to a 3D printed bracket mounted to one of the flanges on the cross-tube to measure cross-tube rotation which gets output to the simulator.

CONTROL COLUMN TRAVEL REFERENCE (measured from the column shaft center): 
Neutral Position: 6.9 degrees (+- 0.8 degrees) forward from vertical
Full Up Elevator:  7.0 degrees rearward from vertical
Full Down Elevator:  19.8 degrees forward from vertical

All said and done, the final Pitch Motor configuration came out pretty nicely and I’m happy with the result.

Parts Used

• Custom 1/4″ aluminum supports (cut by SendCutSend)
• Custom 3/16″ aluminum motor supports (cut by SendCutSend)
• OEM 737NG cross-tube and bearings (sourced privately), cut to fit
• NEMA 90ST Servo Motor and Driver combo (Alibaba)
• 1:50 Planetary gearbox reducer for 90ST nema34 AC servo motor input shaft 16mm (eBay)
• POWER DRIVE Split Taper Bushing: G, 7/8 in Bore Dia, 1.56 in Bolt Circle Dia, 2 in Overall Dia, Inch (Grainger)
• 1/4″ thick custom-made aluminum pitch motor A-Arm (cut and powder coated by SendCutSend)
• Carbon Steel Turnbuckle (McMaster-Carr)
• 2x Ball Joint Rod ends (McMaster-Carr)
• CALT Draw Wire Potentiometer (Amazon)

Final Obervations

Overall the setup has worked out quite well. At high torque, the columns cannot be moved at all. At low or no torque, the columns feel good when moving by hand, and on auto-pilot, the columns also move smoothly on their own.

Note that the planetary gearbox ratio has a lot to do with how hard it is to move the columns by hand. The lower the gear ratio, the easier the columns will move , but movement will be less precise when driven by the motor. The higher the ratio, the opposite effect.

I tried 1:10, 1:50 and 1:100 gearboxes. I found 1:50 was a nice balance of feel and smoothness in my setup. 1:100 required too much force to move the columns by hand, and 1:10 was resulting in shaky movement when on auto pilot. Others have found 1:10/20/25 work for them. So it really depends on your setup. I would say 1:25 to 1:50 are good starting choices for the pitch axis.

I mentioned shaky movement. A side-effect of having such a compact setup is that the control arms are short, which means it’s harder for the motor to move the columns in small, precise increments without a large lever effect at the end of the columns. Therefore, all tolerances need to be very tight to prevent any slop from magnifying into exaggerated movements, resulting in shaking as the software tries to over compensate.

To visualize, imagine on auto pilot the software says move to position X, but the columns are shaking between X+1 and X-1. The software sends a counter command to get it back to center, and what you end up with is a constant oscillating battle to stabilize the position, often getting so severe that you have to put your hands on it to stop it. Think Pilot Induced Oscillation (PIO).

So in addition to minimal slop, it is imperative the software also starts and stops movement smoothly. If it is too abrupt, it can cause the columns to enter an oscillation. Increasing the gearbox ratio with short control arms helps with this because it takes more turns of the motor to move the same distance, resulting in smoother, more precise movements overall.

Conclusion

Ok, so what did you think? Not too terribly bad, but not a walk in the park either. It took a fair bit of research and paper sketches to first come up with a plan, and then a few adjustments along the way. But it was a fun project nonetheless, and wait until you see the final results in action!

As mentioned, in the next blog post in this series, I’ll talk about building the Roll Axis. My pulley solution probably isn’t the simplest, nor the easiest, but it’s certainly one way to accomplish the mission. Stay tuned!

Back to Linked Control Columns & Loading – Part 1 – May the Force be with you!