Seeking recommendations for providing mechanical tension

[AAnnillo]

Hello, all.
I have been working on my B727 TQ and have found a workable solution for supplying tension without having to fully engage the friction lock for the thrust levers.  I have 3 aviation-grade steel cables running through 3 pulleys attached to custom-cut poplar 'flooring' on the bottom of the TQ.  The 3 cables are attached to the 3 thrust lever reels secured with steel balls, just like the real deal.  It's just enough tension to make it more realistic and I'm pleased with the results.
I am now seeking a solution to do something similar to the speed brake handle.  Right now, it's very loose and unrealistic...free-wheeling, if you will.  I initially thought of a screen door closing mechanism to provide this tension but that requires constant pressure to keep the handle up.  I then thought I'd try a gas-type cylinder used for lift gates in SUVs, etc. but the same result....it always goes back to the stored (closed) position upon release.  My TQ will not be fitted with any type of spoiler arming system for auto deployment on touchdown...it will be 100% manually-operated.
What am I overlooking or not thought of to provide realistic tension to keep the speed brake handle in whatever position until it is physically moved?

I appreciate your ideas and inputs.

Thank you.

Antny

[FredK]

My speed brake is driven by a servo motor. That provides both automation and the tension that you are seeking.  You could hook up a servo arm just for the tension.

Fred K

[AAnnillo]

Interesting option.
What kind of servo do you recommend? 
I would think it would need to be strong enough internally to handle rapid back-and-forth movements and given this, I'm guessing a servo with metal gears would be the logical choice.

Thank you for the reply and suggestion!!!

Antny

[FredK]

My speed brake servo is a standard type HiTech servo....nothing special.  It has nylon gears.  It has withstood the test of time.  I did my 737 TQ conversion nine years ago.  The installed servos for the speed brake and trim gauges are all working fine.  The speed brake has a very nice tension feel....it would be floppy without the servo attached.

A metal geared servo would be more robust of course, but is really not needed for this application. The speed brake is only used for a couple of minutes on each flight, and even then it is not a repetitive back-and-forth motion.

Also if you do go with this solution it would be easy enough to go the extra step to automate it since the hardware requirement would be in place which is the hardest part to implement.

Fred K

[sagrada737]

Antny,

To further add to Fred's comments...

I too have modified a real 737 TQ for use in my Sim.  I use the same SIOC software and coding to control the TQ for use with P3D.  This setup has been very reliable, and has never failed to function correctly in my Sim.

The photo below will give you an idea of how I installed the Servo for controlling the Spoiler functions with SIOC, pointing out the various components in the Spoiler part of the TQ.

First of all, I am using an OpenCockpits DC MotorPlus card; and an OpedCockpits Servo card to control all the functions on the TQ.  I had some issues using the "on-board" servo connections with the DC MotorPlus card, hence the added Servo Card.

In the photo, from the Top down, here is a brief description:

-  Open Cockpits Servo Card.  This has the ability to control up to 4 Servo motors, and responds directly to SIOC coding.

- HiTEC Servo.   I use the HiTEC Servo Model HS-654-MG
This is a heavy duty, high torque servo with metal gears. 

Note:  It is very important to disable power going to the Servo when it is not in use, as you can easily damage the gears by moving the Servo manually (TQ Spoiler Lever manually) with power ON.  Remember that! 

Thus, the SIOC software control turns Servo power ON and commands the Servo to move to a certain location, then turns the Servo power OFF after the move is accomplished.

If you are going to use SIOC...  I should point out that the SIOC coding will have to be setup for your specific TQ configuration.  Various values and parameters within the SIOC code will be unique to your specific TQ configuration/setup.

- B737 (misspelled in the photo) TQ Spoiler Link/Rod End.   Shown is the linkage unaltered in the B737 TQ.  I simply designed the Spoiler linkage around this part.  Notice the angle...   This needs to be at a certain angle, such that during the course of its Full Movement, it will not interfere with the TQ Frame.  This requires that the Servo-CAM have a specific form factor, and mounting position on the Servo.  It is best to make the Spoiler Linkage CAM first and get its movement correct; then make the potentiometer CAM so that its travel is correct.  THEN, mechanically join them together.

- Servo-CAM.  I fabricated this CAM in two parts, using the standard Servo mounting part (plastic), attached with 4 screws as shown in the photo.  One part of the CAM connects to the Spoiler Linkage, and the other CAM (co-joined) connects to the link going to the 10K Potentiometer. 

This geometry was difficult to sort out, and I recall having to make several prototypes to get it right.  The reason for this is that each connecting point on the two CAMs serve two different mechanical functions to ensure that the travel is correct for BOTH the TQ Spoiler Linkage, and the Potentiometer linkage.

- CAM Link to Potentiometer.  Nothing too special about this linkage, except that it needs to be cut/adjusted to sync with the required movement of the Servo CAMs and the Spoiler Linkage, within the limits of the potentiometer slider.

On this note...  Allow some buffer on each end of the slider potentiometer, so that it doesn't max out -- yielding about 90% usage of the potentiometer.   This too will affect the geometry of the parts/linkage.

- 10K Potentiometer and Bracket.  I used existing TQ structure to attach a Bracket that holds the 10K Slider type potentiometer.  This potentiometer is a 10K carbon type, Linear Slider Potentiometer.

Of course, the connection of the potentiometer interfaces with some type of game controller.  I use a Bodnar Card for this, but any Windows based Game Controller will work.  This then is calibrated through FSUIPC for the Spoiler function, which in turn requires that the SIOC coding correlate correctly with the mechanical position of the Spoiler Lever and the set limits of the potentiometer, as calibrated within FSUIPC.

"A picture is worth a thousand words."   I hope this explanation helps with your TQ Sim Project.

Also, thanks again Fred for all your help with the TQ stuff.  Much appreciated.

Mike

[AAnnillo]

Fred and Mike;
I thank you both for the extremely informative advice, suggestions and how-tos.  This is most helpful and I truly appreciate your input.

I have ordered a HiTec HS-654-MG servo from Amazon and will be experimenting with its placement and linkage to the bellcrank on the speed brake lever when it arrives.  I enjoy working with mechanics and this will provide hours of trial and error!  I'll be looking-out for a linkages with a clevis to use for this. 

Initially, I had thought I might need some type of governor or clutch-type gearing system to provide this tension and although I will not be using the servo electronically, I'm confident it will provide exactly what I've been seeking, based upon your recommendations.

My next challenge with the TQ will be a way to drive the trim indicator and for this, a full-use servo will be employed.  For the trim wheels, I have a DC robot motor and its sprocket for the TQs 35-pitch chain.
I have several Open Cockpit cards, (DC motors, servo, input, output, etc.) as well as a few Leo Bodnar cards, Arduino Megas, Arduinno Micros, et. al.  I am no where close to doing any type of interfacing for the TQ yet but do have three (3) UniMeasure string potentiometers on stand-by, ready to be installed.  After that will be the flaps handle.  Fun, fun, fun but rewarding when it all comes together!
When I first looked at SIOC programming, it is very daunting to me.  The more and more I look at it, the more and more convinced I am that it IS the way to go.  I'm looking at Mobiflight and the former SimVim programs as well in an effort to determine which is the best.

Mike....if you are willing, would you send me whatever SIOC scripts you have so I can begin experimenting?  I'd be grateful.

Again, thanks so much for taking the time to help this amateur/rookie.

Antny

[ame]

But the OpenCockpit stuff is so expensive. And ironically, not Open.

[sagrada737]

Anthy,

Attached is a copy of the SIOC code that I use.   It was originally developed by Manuel Velez, Spain.  This is what Fred used for his 737 TQ.  He was kind enough to share it with me.  Keep in mind that the Variables, Values, Offsets, and IDs, etc., will likely require change for your particular TQ modification.  In any case, you can get an idea of how the SIOC code is structured.   Since SIOC is not the most intuitive language, it's best for you to first read as much as possible about how to use and program  SIOC from the various resources on the Internet.

Mike

[sagrada737]

Anthy,

Here are a few photos showing some of the mods to my 737 TQ.

Mike

[AAnnillo]

Thank you so much for all this incredibly helpful information and please forgive my delay in responding to express my gratitude—had a little health scare which required a trip to the emergency room and a subsequent overnight at the hospital.  Am awaiting arrival of my ordered servo and pushrods/connectors. 
As my TQ will not have motorized thrust levers nor automated speed brakes, your attached SIOC coding will need to be reworked for sure.  I had done some initial studying of the 'how-to' of SIOC previously and found myself saying 'Whiskey Tango Foxtrot' on more than one occasion.  Coming from a somewhat limited C++ background and experience, I found SIOC more than a little intimidating.  The best thing for me to do is just do some experimenting until I get more familiar with its capabilities.  Others have done the same so why not me?
Again, thank you for helping.  I truly appreciate it.

Antny

[sagrada737]

Hello Anthy,

A quick comment on SIOC...   As I mentioned, SIOC is not intuitive, but once you understand its structure and flow, you can make sense of it, and use it effectively for controlling your TQ in your Sim Project.

Do you already have an OpenCockpits DCMotorPlus Card, or Servo Card?

The best way to begin learning SIOC, is to set up a simple breadboard with a couple of Switches, a couple of LEDs, a Servo, a potentiometer, and a 12 VDC Gear Motor.  Of course, you will also need a 12 VDC Power Supply to power the card(s).  Once all this is connected to your OpenCockpits Card and interfaced via USB to your computer, you are ready to begin SIOC programming and testing.  The nice thing about this stage, is that you don't need any connection or mods to your TQ.  This is simply a way to bench test some hardware with your SIOC software in order to learn some basics. 

For this initial test setup, you are simply getting your OpenCockpits Card connected to your computer, and getting its ID established with the SIOC software.  The ID of the Card(s) is important, as it is used within the SIOC code you write.  If it is not setup correctly, nothing will work.

Once you have your OpenCockpits Card talking to your computer and the SIOC software, you can then write a very simple SIOC program to setup your Variables and do something simple like turn ON an LED with a Switch.  Once that is working, you can move on to doing the same thing with the Servo, making it move a certain distance based on a Switch position.  Then you can further make your test program more complicated by including the Potentiometer, making the Servo move based on Potentiometer position.  At that point, you can integrate your test SIOC Code to control the DC Motor based on Switch and/or Potentiometer positions. -- and other test configurations...   You get the idea...

With success in doing some basic SIOC programing as above, you will then be better able to discern the TQ SIOC code I sent you.

The fact that you only want to automate the Spoiler and TRIM function makes it fairly easy.  I would also include the Parking Brake and Fuel Cut-Off Levers, since they are only switch functions.

You will no doubt have some questions along the way, so don't hesitate to ask on this Forum.  There are quite a few guys that have SIOC experience that may be willing to comment.  I have an extra OpenCockpits DCMotorPlus card and Servo card, so I can easily set it up on the bench and duplicate your test setup as a way to help test your SIOC Code.

Just some ideas...

Mike

[AAnnillo]

Hey Mike;

Again, thank you for all of this.

Given the information you've provided and the research I've done, I will probably use Open Cockpits/SIOC for the thrust levers, flaps, parking brake, elevator trim (wheels and indicators), speed brake and fuel cutoff switches...for now.
The parking brake and fuel cutoff switches should be an easy task as there are only two (2) states—ON or OFF.

I had previously ordered and received three (3) UniMeasure string pots to use for the thrust levers...one (1) 'broke' during testing...the string got wrapped around the internal pot shaft gearing internally and is unusable.  I had recently seen an option for thrust levers using Hall effect pots (no strings) and this looks very appealing with less chance of damaging strings.  For the parking brake switch/light, I am considering an automotive electric door lock, which seems like a great idea.  For the trim wheels, a 12VDC robot-type motor with 35-pitch chain and sprocket (all in my possession) to attach to the trim wheels OEM sprocket and shaft.  For the trim indicators, a servo.  Am now rethinking the speed brake handle to possibly use a Hall effect pot instead of the previously-suggested servo—upon further thought, I'm going to need something to 'read' the position of the speed brake handle and send this to the sim as it's now no longer just about tension. The only possible issue I foresee with using Hall effect pots is the need for a semi-elaborate gearing system for reading the position of a given control rod/lever and transmitting this to the pot itself where it will be picked-up by SIOC and sent to the sim.

I do have several Open Cockpit boards...
One (1) Master Card
Four (4) Expansion Cards V2
Two (2) I/O card connector boards
Three (3) USB Stepper cards, and
Two (2) Outs cards
A couple of ribbon connectors (will need to get a few more) and several USB cables
Yesterday, I ordered a DC motors card...I had thought that I had one on-hand but cannot locate it.
The only thing missing is a partridge in a pear tree!

In addition to the Open Cockpit cards, I also have other electronics boards in my stash...a few Leo Bodnar BU0836X cards, Arduino Megs and Micro cards, several L298N modules, a boatload of automotive steppers to use in a variety of instruments/gauges and a couple of 16-relay boards.  I'm well-stocked with wiring of various gauges and colors and other electronics-related apparatus.

For now, as you've suggested, I feel the best thing to do is to hook the OC cards up and start playing with them and the SIOC coding to get a better knowledge grasp on things.  As previously mentioned, I have allowed SIOC to intimidate me...foolish, I know.  The remaining concern I have is the interface between Open Cockpits/SIOC and X-Plane.  FSUIPC is the driving force behind this interface but X-Plane doesn't rely upon FSUIPC and rather, uses 'data refs' which translate into XPUIPC, a cousin to FSUIPC.  I'll be doing more reading on this and am convinced it can be accomplished.

Thanks for everything.

Antny

[sagrada737]

Hello Anthy,

It looks like you are all set with enough OpenCockpit Cards to begin exploring the basics of SIOC coding.  Keep in mind that SIOC is an "event driven" language, where the code responds to changes in the system, eg. switch status, etc..  In such cases, changes in status, may cause subsequent changes based on requirements in your SIOC code.

The OC DCMotorPlus card has all the functionality you will need to control the TQ.  However, in my case, I could not get the Servos to operate reliably, so I added a OC Servo Card, which worked fine.  In your case, I would try to accomplish everything on the DCMotorPlus card.  Perhaps you will have better luck than I had.

As far as XPlane goes....   I don't have any experience with XPlane.  However, if it uses an XPlane  version of FSUIPC, then you should be able to successfully read and write information from within your SIOC code.  This would be easy to test running your Sim.

Concerning using Hall Effect sensors...  I am strictly using potentiometers with my TQ, so I don't know how a Hall Effect device would work, with say your Bodnar Card.  Others on the Forum might comment on this.     You could easily test this within Windows with your Bodnar Card connected to one of your Hall Effect devices.

One thing I will say...  My TQ under SIOC control has been very reliable over the years.  I did have a couple of cases where changing Sim computers caused the OpenCockpits Card "ID" to change, requiring modifying the new ID within the SIOC code -- afterwhich, it worked correctly.  Something to be aware of.

Have fun...

Mike

[FredK]

To add to Mike's comments...

The OC DCMotorPlus card is an excellent choice for a TQ application since it is an "all-in-one" solution. I have had no problems operating my servos off of it. So I would strongly suggest starting there before you make an investment in a separate card.

A Hall sensor will always be a better choice than a pot since there is essentially nothing to wear out. I have replaced pots as necessary with Hall sensors on my control hardware.  That said, after nine years of service no pot has worn out on my TQ.

Fred K

[AAnnillo]

With regard to the Hall effect pots for your thrust levers...how did you do it?

Since mine will not be motorized (e.g., No auto throttle), I'm thinking a series of linkages and/or pushrods and possibly bellcranks.  I initially thought about a rack and pinion system but space will be tight as I have three (3) engines.  Space wise, I think a drive rod off each thrust level reel connected to a circular gear meshed with a gear attached to the pot shaft would be the best solution.  Still undecided at this point.

I've ordered my Hall effect pots and am awaiting their arrival.
Then more shopping for ready-made robotic parts, gears, linkages, etc. to determine a best practice.

Onward and upward, slowly!

Thank you again for the guidance and support.

Antny

[FredK]

Antny...

As I noted in my comments above I did not use any Hall sensors for my TQ, only pots. Those pots are nine years old now and have not failed.  The pots for the throttle lever were the Unimeasure string pots, same as Mike.

As noted above also....I have replaced pots on my control hardware (yoke and pedals) with Hall sensors.

Fred K

[AAnnillo]

I'm thinking I might be misunderstanding the components here...are these the same things called by a different name..?

I'm referring to this:

https://www.amazon.com/Effect-Potentiometer-Encoder-Sensor-Degree/dp/B07SXFBN5D

Is it possible that this is really just a rotary encoder?
No string, mind you, but maybe a similar thing in a different housing?   Thus the confusion here.

[sagrada737]

Hello Anthy,

This Hall Effect device (Sensor) does not have any "stops" in its rotation.  The Sensor shaft can rotate 360 degrees, which can be beneficial in some applications.  However, that's neither good nor bad in the TQ application.

I use a similar device in my TQ for the Thrust Levers, except it is a wire-wound potentiometer.  With this type of "free-rotating" Sensor/Potentiometer, you need to fix the body of the Sensor so that the Shaft is positioned correctly to provide the high and low voltage positions as read by the Bodnar Controller.

When using a Bodnar Controller, three wires are connected to Sensor/Potentiometer.  Bodnar provides a Ground (GND) connection, and a +5 Volts connection.  It also provides an Input connection for reading the voltage base on the Shaft position of the Sensor.  These are likewise connected to the Sensor/Potentiometer connecting pins: GND; +5v; and OUT respectively.

It is best if you buy one of these Sensors and test it with your Bodnar Controller.   Once you see how it works, it will make more sense for how to use it in your TQ application.

In the case of both the Hall Effect Sensor and a Potentiometer, that are being supplied with a 5 volt potential, the center output pin will show 0 to 5 volts depending on the Shaft rotational position.  This is the voltage that the Bodnar Controller reads and converts into a digital value as shown by Windows and/or calibration value in FSUIPC.

Keep in mind that we are not building a Space Craft that requires super reliability.  The Duty Cycle in a Sim TQ application is very low.  Frankly, a standard 10K Ohm Carbon type of potentiometer will serve the purpose just fine.  I use both rotational and slider potentiometers in my Sim., depending on mounting considerations.

Mike

[AAnnillo]

You hit the nail squarely on the head—-mounting considerations.

I had first seen this type of 'Hall effect potentiometer' via a general internet search which led me to a video posting on YouTube and got more and more curious about them.

After spending a reasonable amount of money for three (3) UniMeasure string pots, which were highly recommended to me, and having the first one get tangled-up internally, I became concerned that this could happen again once all is installed and buttoned-up.  This is when I started inquiring about the alternative 'pots.'

Upon further thought, I don't think these Hall effect pots would be a suitable option in my TQ due to mounting and limited space considerations.  I have three (3) thrust levers to operate whereas most sim TQs I have seen thus far have two (2); all contained in the same close quarters.  As an example, the on-center point measurement for my B727 thrust lever reels is a paltry 1-9/16"..from #1 to #3.  This is not a lot of room to play with, mechanically speaking.  The video I had seen using this application included a track commonly used in kitchen drawers with a nylon/plastic rack attached.  A nylon/plastic gear was meshed with the rack and attached to the pot shaft.  Again, very elaborate and effective but not much room inside my TQ.

As a result, I will reach out to UniMeasure and obtain a replacement for my damaged string pot.  Yours and Fred's explanations and guidance have been invaluable on this topic.

This is not to say that I won't try a Hall effect pot for my flaps...as I would only need a single one but that bridge to cross is over the far horizon.

Thanks again.

Antny

[kattz]

Just to ensure I'm thinking right, the spoiler is used in 3 positions, correct?  Stowed, flight detent, and full ground spoiler detent, right?

I was planning on using a polyclutch with a right angle motor and a pot on the other side with the operating cam.  This way I would have the 3 positions by hand and don't have to worry about motor overtravel.

[navymustang]

Many time in flight you adjust the spoiler during decent to an infinite number of positions as you manage energy. So to say it just has three positions is incorrect. You need to have a potentiometer on this, not just sensors for determining the three positions you cited.

[AAnnillo]

On mine, there are no detents, per se.  My speed brake handle is free-wheeling; that is, it slides around an arc from fully up (extended, to the rear) to fully down (stowed, to the front). 
There are position marks on the side for 10 degrees, 20 degrees and 35 degrees.  On the flaps handle, there are detents for the various degrees of flaps.

Antny

[navymustang]

That is correct for all 737 series. No detents other than ARMED for the spoiler.

[AAnnillo]

On my B727 TQ, the only detent for the speed brake is in the fully stowed (forward) position.
When either #1 or #2 thrust levers are in flight idle, a mechanism below pops the handle up, allowing it to manually be moved back to the fully extended position with relative ease.

Antny

[navymustang]

Actually, the "pop up" is not for you to move the handle. it is for the motor to move it. If the pilot does not arm the speed brake, when the thrust reverser is engaged and wheels are on the ground, the speed brake auto deploys.