sign up .

May 25, 2020, 07:58:57 am

Login with username, password and session length
35 Guests, 3 Users
  • Total Posts: 55081
  • Total Topics: 7341
  • Online Today: 33
  • Online Ever: 582
  • (January 22, 2020, 08:44:01 am)
Users Online
Users: 3
Guests: 35
Total: 38



Will Depart in...



Mike's 737 TQ Project

Started by sagrada737, March 20, 2020, 06:22:53 am

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.


March 20, 2020, 06:22:53 am Last Edit: March 20, 2020, 07:51:42 am by sagrada737
Hello Folks,

Although my 737 Throttle Quadrant (TQ) project has been completed for many months, I thought I would share a bit about how I converted a real 737 TQ so it could automate with my Sim.

I found a salvaged 737 TQ in England.  With shipping, it cost me around $1500.  I think it was from a 737-400 but I'm not sure.  In any case, when it arrived, I began to consider how I would "transpose" the real deal into a TQ that I could interface with my Sim.

The first thing I noticed was that there was the world's supply of switches, brackets, wiring, etc. in this heavy duty TQ system.  The 737 is built like a Tank...   So I began to remove all the wiring and clean up the TQ the best I could.   I decided not to take the TQ fully apart, as I was concerned about not having the right tools to put it back together.  Anyway...  I managed to remove all the unnecessary items within the TQ and made ready for the automation mods I wanted to make.

My strategy was to use the OpenCockpits DC MotorPlus Card for control, using SIOC code to manage the TQ automation.  Fortunately, this has already been done by a few cockpitbuilders, Fred K. being one of them.    Fred was a great help to me in learning the basics of SIOC.   Thanks again Fred.   I used his SIOC layout for my setup, with the only requirement being I had to change the code to match my use of potentiometers and switches, etc.  I should point out that I have done some programming in the past, but I am by no means good at it.  It took me a while playing with SIOC to learn some basics of this quite capable "Event Driven" software.

As I removed the existing TQ wiring, I was careful to retain the wiring and switches that I needed to interface to the Controller card.  But most of the wiring was simply removed, which made way for extra space in the TQ for my own internal modules/hardware and related wiring.

The biggest challenge on the TQ was to figure out how to control the TQ Levers.  Most folks that have converted a real 737 TQ have mounted the motors and pots external and rear of the TQ.  I wanted to have all control modules within the TQ body, so I designed a special drive setup that was installed at the bottom of the TQ.  This system needed to be fairly compact and include right angle geared motors, clutches, drive sprockets, and feedback potentiometers.  I decided to use a 6 RPM gear 12vdc gearmotor, which had minimal backlash, with a 6mm output shaft.  I brass-shimmed the output shaft to use a 1/4" clutch, which worked quite well.  All the connecting shafts are bearing supported to relieve the assembly from additional friction, making for fairly smooth operation.  I matched the drive shaft rotation to a high quality 10K potentiometer via a timing belt reduction that used 340 degrees of Pot rotation.  The Pot drive shaft was bearing supported to relieve the Pot from any rotational stress. 

For the TQ Lever drive cable, I machined a cable-to-chain connecting piece that joined the TQ cable to #35 chain at one link, and an adjuster link on the other cable/chain connection.  There is some slight backlash in this mechanical linkage, but it doesn't seem noticeable when flying.  However, the mechanical design of the drive motor to Pot interface is zero backlash, so information is fed accurately into the Controller card for processing by the SIOC code.  Here is a photo of the TQ Lever drive module within the body of the TQ.


As you might appreciate, the SIOC code allows reading and writing FSUIPC values to/from the Sim software, which is P3d in my case.

For the Parking Brake Lever, I left the linkage alone and used its geometry to guide the connecting shaft to close a micro switch to sense Parking Brake On/Off.  I used a small "super magnet" to hold the parking Brake Lever in its UP position, which also allows it to be easily released to its Down position.  Here's a pic of that assembly...


For the Speed Brakes, I used a HyTech Servo (easily interfaced with the OpenCockpits DC Motor Plus card).  The difficult part here was to establish and sync the Speed Brake linkage with the existing linkage as coupled with the Servo.  This required special consideration to the ServoMotor placement and its lever as connected to the Speed Brake linkage.  Here is a photo of that configuration...

20180810_163250 - Copy.jpg

For the Flaps position sensing, I made a bracket to mount a 10K Slide Pot that is moved by a Lever attached to the existing Flap cable.  This was mounted in such a way as to allow for 90% of Pot movement, so that the Pot value is sync'd with the Flap Handle movement (Flaps Up to 40 degrees).  Here is a pic of that setup...


For control of the Trim Wheel, I used a 30 RPM DC gear motor with a #35 chain sprocket to drive the Trim Wheels.  This "mimics" the movement of the real 737, but by no means is as fast or responsive as the real deal.  It is just to have some trim wheel movement that syncs with what the aircraft is doing.  This was done to keep things simple, by not requiring a special high-speed drive system for this function.  I'm happy with it thus far...   Also, the SOIC code has parameters that are set to show properly the Trim Indicator on the TQ.  This was a bit tricky, due to the fact that the linkage is not linear, but through the SIOC code I managed to make it respond fairly accurate.  Here is a pic of the DC motor placement, tucked within the body of the TQ.  You can see the sprocket behind the motor.


I think I covered most of the systems.  The rest of the control/sensing involved uses the existing switches in the TQ to interface via FSUIPC.  I ended up mounting the USB OpenCockpits DC Motor Plus card at the front of the TQ, where there was space available.  Overall, this conversion of the TQ was fairly easy, but did require some head scratching as to synchronizing some of the modules.  I was pleased that the TQ Lever system worked out well.   Thus far, the TQ has been problem free, and the SIOC code has been very reliable.   I do have an issue with the TQ Lever algorithm causing the Lever to move on its own sometimes, but it quickly corrects on its own and behaves then behaves normally.   When I have more time, I will try to track that down.   Fred has the same issue with his SIOC controlled TQ...

This was a fun project within the range of any Sim Builder.


P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Such nice setup!
I also had the wires go out the back but i rebuild mine the same way as you did. There was several problems with positioning the chains through the lever pulleys.

I also found it hard to create the mou ting brackets of the shafts/clutches and engines mounts on.

How do you move the speedbrake manually? Can the servo take the beating of move manually?

Br Peter


March 24, 2020, 07:37:50 am #2 Last Edit: March 24, 2020, 08:34:42 am by sagrada737
Hi Peter,

Glad to hear that you too tackled automating a TQ project.   It was a lot of fun and helps a great deal to boost the sense of Sim immersion.

Regarding mounting the internal TQ Lever drive system...   I used the existing brackets at the bottom of the TQ to mount the TQ Lever Drive Assembly.   This was a bit tricky, as the "chain tension" is critical to achieve as little backlash in the Lever movements.  Since the lower internal mount brackets flex a little, I had to install a perpendicular brace to help keep the chain tension stable.

I made the TQ Lever Drive Assembly with a CNC milling machine, so all the holes, etc. were lined up perfectly.  This allowed for the shaft support bearings to do their job, making for a very smooth working system.

Regarding the Speed Brake movement...   I use a HiTech servo for its automated movement.  At the end of each software controlled movement, the software routine turns OFF the power to the Servo, so it can turn freely via the linkage connected to the Speed Brake Lever.  This is necessary when you need to deploy Speed Brakes manually.   If you don't power down the power to the Servo, you risk damaging the internal plastic gear system.  In a little bit, I'll post some photos of the TQ, as I have not installed it yet in the Nose Section.

P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Thanks for the info Mike.

I can see you are using the newer polyclutches. May i ask what model it is? I have the Slip ease series and they work fine.

My setup had yours as inspiration. :)


March 29, 2020, 03:15:09 pm #4 Last Edit: March 29, 2020, 03:20:32 pm by sagrada737
That's a well done assembly.  Glad to hear that my design inspired someone else.

Regarding the cluthes...   I don't quite remember what the exact model was.   However, it was the lowest torque version.  The adjustment of the clutch is somewhat important.  You want the TQ Lever cable tension to be mildly tight, so that the Lever movement "feels" about what you might expect.   This is a function of the clutch adjustment.

I found that it was not possible to totally eliminate the backlash with a clutch adjustment that gave the Levers the right feel as you move them back and forth.  I ended up with about a 1/4 inch of backlash in the TQ Levers, which for me was not a big issue.   Fred reports a similar backlash with his TQ setup.

Keep in mind that there is a "range" of clutch tension that will allow the small geared DC Motor to rotate and move the TQ Levers under automation, but still allowing the Motor to rotate if the Levers are physically stopped.

What software are you using to control the TQ automation?

BTW --  Schulz has done a great job with his 737-800 Sim Project.

P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Hi Mike.

Thats interesting info about polyclutches. I also would like to try the ones with lowest torque. Then you can finetune much more. My clutches are too tight after the clutch settings is above say 30-40%.

Would be nice to have more room to play with.

Regarding the backlash. This is impossible to get rid of with this setup. If using cheap dc motors and pilyclutches there will be some backlash. But there is also backlash in the real aircraft about 2-3mm (at least the 737CL).

I am writing the software myself with Sketch code and using teensy microcontroller. It works great with great sync and no "hunting".

I would like to setup the speedbrake the same way, motor with clutch. Then you can tune the feel in the handle and code the event of full beake or when the levers goes back.

This TQ really is a neverwnding story. :)


Great content here.

Im my last TQ Conversion, I found the best way to minimise the backlash was to switch to a motor with planetary gears as opposed to spur or worm gears, and also use the smallest possible sprocket at the motor as possible.

Still had around 3-4mm, but much better that the first setup. 

Also with regard to the speedbrake, these are quite stiff to move in the real A/C,  the NG Spec is 18-20Lb lever force to move it.  Mine measures at 19.


Hi Mick.

Yes, the Speedbrake is my trouble right now.

I tried this:

Motor -> Polyclutch -> Lever.

But my polylutch could not tighten enough to hold the position when the speedbrake was almost at the bottom.
Otherwise this would be a great setup since the logic behind the speedbrake is only "Full deploy" or "Full reverse". You dont need to track the position for it. You can just drive the lever to the edges and then slip on the clutch for half a sec or so.

But with a servo i have the problem with manual override. How do i know that the servo has the right feel when it is powerless and i move the lever manually?

Thats where my problem is now.


March 30, 2020, 09:22:04 am #8 Last Edit: March 30, 2020, 09:24:08 am by sagrada737
Hi Guys,

Regarding the issues with the Speed Brake...

We should consider that our Sims are not exactly like the real 737 in terms of forces required to move linkage, such as the Speed Brake Lever.  In my view, trying to make everything exactly like the real 737 is admirable, but unrealistic.  However, since automating is personal preference, I will share what I did for the Speed Brake Lever operation.

I first gutted all the extra linkage connected to the Speed Brake Lever.  I then decided where on the TQ body that I wanted to drive the Speed Brake Lever.  I used the exit hole in the front of the TQ (see the photo that I posted, and constructed a tube length that would allow the Rod-Ends to connect for Servo movement.

At this point, I checked the Speed Brakes Lever freedom of movement.  To my surprise, it had some "galling" on the Lever travel ring.  I used some very fine emery cloth to remove the galling and then lubricated it with a fine oil at all locations where the Lever bar slides across the ring.  The result was very smooth operation.  Keep in mind that I was not interested in having the same Speed Brake forces as the real 737 for moving the Speed Brake.  I just wanted it as smooth as possible so the Servo Motor and its linkage could move the Speed Brake.

The next consideration was where to mount the Servo Motor, and what length its lever should be, and what angle the lever should be for Servo movement from Speed Brakes Lever stowed to full Speed Brake Lever at its full deployed position.  I found a position for the Servo Motor and its Lever, such that it required as much of the Servo rotation as possible to accomplish the Speed Brake Lever stowed to full movement, and I set these values in the SIOC code accordingly.

I am using a Hi-Tech Servo that is controlled by the SIOC TQ software, which at the end of each software controlled movement of the Servo Motor, it shuts down the power to the Servo Motor.  This allows easy manual movement of the Speed Brake Lever, which due to the Servo gearing, provides some sense of force required to manually move the Speed Brake Lever.

To know if the Servo Motor is able to be moved manually, you can easily test this on the bench.  When you have power to the Servo, you should not be able to move the output shaft.   With the power OFF to the Servo Motor, you should then be able to move the Servo Motor output shaft.  Of course, with enough force, you can break anything, so be careful not to force the Servo output shaft when power is applied.   As far as the "right feel" for the Speed Brakes Lever...  It will be what it will be, based on how you designed the linkage, the type of Servo Motor used, and the length of the Servo Motor Lever used.

With the above approach, you should not need a PolyClutch between the Servo and the Speed Brake Lever.

P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Nice Mike!

On my earlier TQ, I 3D printed a pair of arms that had a slip clutch inside to link them, and attached them to the speedbrake lever, and a 747 Linear actuator.

Then the actuator could raise and lower the handle, while allowing for manual control.  Quite similar to how my current/real setup works, except thats all under the floor.

As they say, many ways to skin a cat :)


Hey Mick,

You really know how to build them solid!  Nice work on the 3D printed levers.  Great how you used the 747 Actuator to manage the movement.

Yep...  Lots of ways to accomplish the mission.   That's a great part of sharing on the Forum.

P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Totally agree with the give and take when it comes to have the exact same setup as in the real aircraft.

In the end it is totally unrealistic that my aircraft cant fly and is stuck in my garage! Where do we draw the line? Hahaha

I think i will have some prototype to show soon.

When i made a shorter arm that is connected to the motor my clutch was able to move the spdbrk from the bottom with no problem.

Lots of ways to do things, great for all info. The most fun with the simulator is the build if you ask me. :)


March 31, 2020, 01:16:44 am #12 Last Edit: March 31, 2020, 01:19:00 am by mickc
Quote from: kurt-olsson on March 30, 2020, 11:03:27 pmTotally agree with the give and take when it comes to have the exact same setup as in the real aircraft.

Indeed.  The other option is to actually use the setup from the real aircraft :D



You have to start a youtube channel with your stuff Mick!

You have 1 million viewers in 2 sec!

This is my final setup!
Really happy with it.

Motor->clutch-> pots.


This content here is superb.

Regarding the spoiler automation, my setup will have to be very compact, because there are two 220VAC servos and gearboxes in and on the floor immediately in front of the TQ.

Thanks for some good ideas.



Good morning!

Can we please talk about the trim wheels?  I want to use a motor and a belt with a slip clutch, but am not sure how to control the 2-4 different RPM's of the wheels.

Also, I want to use a slip clutch on the 10-turn pot with a 24:1 reduction gearbox so that the wheel runs its complete 240 turns stop to stop as designed.

I would appreciate any help that I can get.




Hello Kev,

It sounds like you want to duplicate the Trim Wheel function of some aircraft...

A few questions:
- What aircraft is this for?
- What Trim Wheel speed in RPM do you need at the Trim Wheel?
- Are you skilled at programming/using Arduino controllers?
- How much room do you have to mount electrical components?
- How much do you want to spend on the Trim Wheel function?

With those questions answered, it might be possible to offer up some suggestions for your project.

On my 737-800 TQ automation, I simply used a small, low cost 12 vdc gear motor ($12) 30 rpm.  I machined a 24 tooth #35 sprocket to drive the existing 737 Trim Wheel.  In the real 737, the Trim Wheel speeds are very fast, with instant start/stop capability.   I did not want to go to that complexity, so I just have the Trim Wheel move to its commanded location, by turning the gear motor On/Off Fwd/Rev.   Of course, it does not reproduce the same operational characteristics as the 737, but it was cheap, easy to build, and gives a nice effect in the cockpit.

P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Hi, Mike,

Arduino?  Nope.  PLC and robotics, no problem.  But I can learn.

I actually want to simulate the real Boeing motion.  Because I have the pulley box that goes under the radio pedestal, I have some extra room to mount electronics.

When it comes to mechanical design I'm fairly good, and I have access to a full machine shop.  However, where someone has done a good job of the compact throttle actuation in the TQ like you have done, I'd rather not reinvent the wheel and simply follow in other's footsteps where I can.

I have not been able to nail down the trim wheel speeds - I've asked people here and elsewhere but can't seem to locate them.  There are two to four depending on who you ask, but that's as far as I've been able to get.

There are two switches on the yoke.  One is the stab trim brake release, and the other is the nose-up/nose-down switch.  Also want to duplicate that action as well as the ability to hand-crank the wheel at any time and keep track of the trim position while doing so.




Hi Kev

What sim software are you running? Prosim/Sim-A etc?

There are indeed 4 trim wheel speeds. Well 2 sets of 2 technically.  There are 2 different speeds for flaps up and flaps down, and another 2 for A/P engaged or disengaged.  The fastest is A/P disengaged, flaps down. Slowest is A/P engaged, Flaps up.

Actual trim movement speeds for the NG are:

Flaps UP:
A/P On - 0.09 Trim units/Sec - 85 RPM
A/P Off - 0.2  "    "          190 RPM

Flaps DOWN:
A/P On:  0.27    "     "     257 RPM
A/P Off:  0.4     "    "     381 RPM

I use Prosim, and it has gates for flaps up, A/P engaged etc, so it was fairly easy to set up an Arduino for 4 speeds. The ardino drives a motor via an Pololu motor shield that uses regenerative braking so the trim wheel stops straight away.

The motor also has an integrated encoder that is used for manual trim wheel operation. 

Not sure where you heard that the trim switches have separate functions, but they do not.  They are wired in series to prevent accidental trim operation in the event of a switch failure.  The clutch mechanism is automatically activated at the motor drive in the tail with trim commands.


Thanks, Mick!  I appreciate the info.

I am using Pro-Sim.

Is your set-up small enough to fit in the TQ?  Might I request a picture or two?

Many thanks!



Mike, where did you position your motor that deives the trim wheels?
Cant see it on your video.


I placed the 12vdc 30 rpm gear motor as shown in the photo below...

I wanted it to be central within the TQ body, and use the same #35 chain.  It turns out that there are lots of internal brackets and holes that you can easily take advantage of, when making your own mounting brackets, such as the Trim Motor bracket shown in the photo.

Also...  Notice the linkage just below the Trim Motor.  It is connected to a HiTech Servo motor with its own Lever, which connects to the 737 Trim Indicator linkage.  By playing around with the placement of the Servo Motor and Lever length, and proper adjustment to the TQ software parameters, you can get the indicator to show accurately the Trim Value.

P3d v3.x with Sim-Avionics (two computers), FDS MIP, 3-Optoma HD GT180 projection display driven by a single nvidia GTX980.


Mike, I wanted to say that I'm impressed with how CLEAN your TQ is!  Just started on mine and it looks like it has 30 years of spilled coffee, soda, and chips in it!  Yuck!



April 04, 2020, 02:17:14 pm #23 Last Edit: April 04, 2020, 02:38:08 pm by mickc
Quote from: kattz on April 03, 2020, 05:04:07 amThanks, Mick!  I appreciate the info.

I am using Pro-Sim.

Is your set-up small enough to fit in the TQ?  Might I request a picture or two?

Many thanks!


In the Classic TQ I converted it did fit inside the TQ.  I made a pair of plates that sandwiched over the connector mounting plate on the bottom of the TQ.  Then a right angle bracket to mount the motor to, and then the chain tension could be adjusted by adding or subtracting shims from under the motor bracket.

The motor i used was quite a bit bigger than Mikes, as it has the encoder etc and is quite beefy so it handles the instant braking.

In my current TQ i have had to do it differently as I am utilizing all the OEM connectors, and the NG uses solid rods that pass through the bottom of the TQ instead of cables.  Also there is a flap handle position resolver that reduces the available space that isn't there in the classics.

The motor is mounted on the front face of the TQ in this one, and has an idler sprocket mounted above it.  The chain comes out and goes back in through the holes that it does when fitted to the aircraft.


Thanks, Mick!

I need to ask you and also Mike, how are you getting these units so CLEAN inside?  Every type of cleaner that I use just does nothing.  well, it'll semi loosen chunky dirt, but grime etc and residue seems to be a part of the paint!  I have a -500 TQ, BTW.

I see these pictures and have a thousand questions.

I need to figure out that motor type and controller... and how to control it.  Make that item # 1065 on the never-ending list. 

I figure that I've ten years to retirement and another 10-15 years after that, God willing, so I might get to fly it ONCE the way things are going.

Thanks again.


Like the Website ?
Support and Click Below to Donate