question regarding f 14 spider detent (emergency wing sweep handle)

[Chronocidal]

I've been messing with the idea of making something much less complicated (no motor drives, and a simple pot/hall effect for the wingsweep axis), and had the thought of printing the solid portion of the wingsweep cover with a hollow track, and then gluing a flexible piece of clear plastic into it.

I think I have a small scrap of lexan that would probably do the trick, but I also make a habit of collecting large flat scraps of transparent plastic packages.

Otherwise, my plan has been to print the entire wingsweep handle, and model in a spring-powered detent tooth that will lock into place in the lowered and raised positions, while activating a tiny pressure switch at each location.  It'll be a simple slide.. but I think I can see how you could make the entire handle function.

The lower handle has to just be a box for the upper handle to slide in, with built-in limits for the sliding motion.  The upper handle needs to be hollow in the center, to make room for a rotating shaft attached to a tension spring, with a tooth on the lower half that will spring out and keep the handle extended when pulled out.  The upper end of the shaft needs to be hidden in the yellow handle, with an angled T-shaped cap that will limit the rotation of the shaft in one direction, while letting you use the side button to press the other half of the T and rotate the shaft back so the tooth will disengage and allow the handle to collapse.

Maybe a quick model will help.

This doesn't have any of the part geometry worked out, just the basic concept.  It'll need the interior of the handle designed with room to hold the springs, button and shaft, but that's the sort of mechanism that would do the trick.

The one part of the handle I've never quite understood is what is going on with the odd angles.  The handle cap seems to be tilted, and the entire lever doesn't actually appear aligned with the rotation axis, as if it's offset forwards, and then the entire lever is leaning backward.  No one seems to have enough clear pictures of the handle itself to make that shape make sense.

Edited June 12, 2023 by Chronocidal

[Gareth Barry]

Chronocidal, that's an excellent illustration, and thank you!! And it looks a lot like the original! 

[Chronocidal]

I'm surprised my 1am musings made that much sense, but I'm thinking I might try and make it now.  

The individual parts need fleshing out, and the tension spring can probably be attached anywhere inside the handle.  I think my main concern is making sure that spring in the handle will give enough force to push the tooth out at the bottom.  The shaft rotation will have to be pretty precise.  Some kind of torsion spring on the shaft (possibly inside the tooth) might work better, but the upper handle has the most room for a spring, and using that see-saw action with the sideways spring is probably much simpler.

[Gareth Barry]

OK so, the summary of what I have done to this point. For the last few days I have been checking and rechecking clearances etc, this is because my CAD skills are limited at best. Still, we do the best we can. Still have to;

-engineer the push switch which allows the handle to be stowed, as well as tell DCS that the handle is being stowed

-engineer a detent system for the different flap positions, ie up, down, emergency up and emergency down (even though emergency down does nothing as far as I know)

-decide on how I am going to mount the magnets for the hall sensors - there will be a few of them

-decide on what power supply to use for the motors; at the moment I have drawn a hole for a kettle plug input, which would mean that the 12v DC for the motors would all be done with internal circuitry from a mains supply - I am not sure whether to just go with a generic commonly available 12v input and save myself the bother

-decide on either 3d printing the flap lever or laser cutting it from aluminium and then bending it.

 

Other things I should mention;

I am thinking of using commonly available gt2 belts that are used in 3d printers for the pulleys between the motors and the levers, however, I may go with gears.

I am probably going to make the parts as they are currently drawn, check that they all fit, and then correct as needed. I will update the files with any corrections that I make. My plan is to get it functioning without the motors, ie as just a basic input throttle, and then get the motors working. 

At the moment, I am getting ready to go on vacation so it probably wont be until mid July that I begin laser cutting and 3d printing parts. At that point, my hope is that I keep receiving suggestions, criticism as far as construction choices are concerned, so as to improve the design. In any case, the CAD files are will be here as at least a starting point for anyone who wants to make something like this.

 

f 14 throttle.zip

[Chronocidal]

Without having the time to look through your folder of things, a couple of suggestions based on my recent DCS BIOS experiments.

- The switch to stow the handle should be easy to achieve in terms of DCS code.  It's a simple toggle, you'll just need contact switches for when the handle is extended and retracted.  You can probably wire some very tiny contacts into the design I suggested.  I was planning to cut small recesses in the sides of the handle with stoppers at the slide limits that would mount the contact switches.

- The flap detents... don't make any sense.  Unfortunately, they are not currently programmed correctly.  I spent the past couple of days working out the DCS BIOS code to achieve a pushbutton flap axis, which just activates the individual detents.  Well.. turns out the detents are not directly connected to the flap axis at all.  

There are four button mappings in the sim to give the flaps their four "positions".. but these do not correspond to the handle positions.  I tested assigning them.  "EMER UP" and "UP" both put the handle in the fully up position, "EMER DN" and "DN" both put the handle fully down.

On the other hand, the flap axis has a full range of motion.. and so do the flaps, when you move the handle that way, which is not how they should work, as far as I understand.  My "pushbutton axis" is programmed to manually set the flap axis at values of 0, 21,845, 43690, and 65535 (the full range of the axis).  That does indeed move the handle to the visible marks on the flap control, but the outboard flaps move to match the handle position, while the inboard sections drop fully once I move from "EMER UP" to "UP."  Obviously, not ideal.

I don't think this is going to get fixed unless Heatblur modifies the model and animation routines in the cockpit.  Until there's a better option, you really only need two flap detents, UP, and DN.  The handle in the cockpit is not going to accurately represent that.  My flap handle is going to just have to trigger full up on the top two detents, and full down on the bottom two.  It'll feel right, it just won't look right if I glance at the lever in the virtual cockpit.

- For the flap handle, I would definitely go with a cut/bent aluminum handle if you have the ability, with a 3D printed lever cap.  The mechanism for the flaps (and throttles for that matter) dealt with shifting the handle sideways to move past detents.  If you want that action represented, you'll need a flexible handle, and a printed lever probably isn't going to do well at that.

[Despayre]

OMG, what is that video? There are children here!

[draconus]

On 6/12/2023 at 9:12 AM, Chronocidal said:

Shouldn't there be a lock in the lowerposition too? Or no lower spring?

[Chronocidal]

2 hours ago, draconus said:

Shouldn't there be a lock in the lowerposition too? Or no lower spring?

I don't believe there's a lock down, the spring in the bottom is in tension, pulling the handle down.  The catch is to hold it in the extended position, until the button is pressed, releasing it back down.

When I make the parts to print, I'll be adding a small detent with a switch to detect the lowered position, but the button is only to release it from being pulled up.

[draconus]

4 minutes ago, Chronocidal said:

I don't believe there's a lock down, the spring in the bottom is in tension, pulling the handle down.

Oh, that explains it, thanks.

[Chronocidal]

A small update to that lever design, this one should actually be physically printable and usable.. once I figure out the potentiometer linkage.  

Funny, I was busting my brains over how to fit two switches in the lever, when I realized the limit switches I have include both open and closed options, so I only need one to trigger both events (handle up and down).

 

[Gareth Barry]

OK so, after a holiday in the UK and the Netherlands....(btw, I HIGHLY recommend the Soesterburg airforce museum in the Netherlands, great value for money, and ticked off many of my aviation bucketlist items to see...)

So,  have made some progess. Chronocidal, your CAD model looks great - my CAD skills are not nearly as good. However, I may try to implement your design.

In the meantime, I have begun 3d printing the throttle handles - pretty happy with how they came out. I will begin wiring the up this weeekend. I haven't included models for the hat switches, as i thought most people would be able to do that themselves? However, I may do that anyway (include them). Already, I am seeing some improvements that could be made to the model, for instance, the holes for both ha-switches should be bigger - I will get onto fixing that. Again, I will provide all of this for free, hoping that someone else will use my models, and perhaps build themselves this throttle. I am also unsure of the size of the flap handle - it seems to scale according to the pictures but it really is small?

I bought the black acrylic to begin laser cutting the templates for the throttle body. I ordered 5mm acrylic, which I am now thinking might be overkill- perhaps 3mm would have been a better choice. I don't want to waste the money I spent on the 5mm sheet however, so I will probably go with that for most of the chassis, and just use 3mm for the curved part on the top.

All in all, I am happy with the overall shape and fit, especially considering that I am only learning all this stuff.

[Gareth Barry]

So my plan is, to first build the throttle as a basic axis, without the servo actuated autto-throttle and wing-sweep mechanism, and then try to add in the motors and drive units for those. The whole unit will obviously be built with the accommodation for ser-motors etc in mind.

Now, regarding the system to motorise the throttles and wingsweep- I have come up with 2 ideas, and I am unsure as to which is best- I may experiment with both.

idea 1 (the one I have done the most planning for) ; is to use moderately geared 6v motors, driving the throttles and wingsweep handle by way of gt2 belts. The 'stalks' of the handles have gears that I have drawn that will accept a 6 mm gt2 belt for a 3d printer. Each handle will have its own hall sensor to monitor position, and use a PID (had to learn what that is) to move control the motor which drives each handle. There would need to be some sort of condition that when met 9from a person shoving on the handle) trips out the PID loop and disengages autothrottle/wingsweep. My one thought was to use a current sensor to register a spike in current due the motor being stalled, and use that to trip out the PID loop. However, I would far rather a clever solution in the code that would take care of this without the use of something like hall current sensors.

idea 2; use a clitch servo from DFrobot;

DFRobot 2kg Clutch Servo | 300°, 2kg, Integrated Clutch (diyelectronics.co.za)

or maybe a lighter one, such as

DFRobot 9g Clutch Servo | 180°, 1.2kg, Integrated Clutch (diyelectronics.co.za)

This would then drive the handles, which would still have analogue hall-sensors to monitor position regardless of whether the servos are driving them or not. 

I am unsure which of the two ideas would be simpler, bearing in mind I have still yet to code anything in arduino, and have no idea how DCSbios even works !!

But, that is half the reason that I am doing this project, firstly or the challenge, and secondly to learn something. Heck, this project has already taught me what a PID is and how a servo works!

As always, I appreciate feedback and thoughts. Rustbelt, here's seriously hoping to hear from you...

[Gareth Barry]

OK so, the 3d printed parts have worked well - the laser cut acrylic, not so much, sadly. I was shocked at just how brittle the stuff is, and the laser cutting didn't quite get all the way through the 5mm sheet. I tried to bend the parts out, only to crack through most of the parts at the edges. It looked very pretty before that though!

I am probably at this stage going to redesign the main chassis of the throttle to be 3d printable - would probably be easier for most people. Oh well you live, you learn!

Still chugging along in this project, hopefully it will be of use to simpit builders.

 

[HILOK]

great project, thanks so much for sharing this!!

cant speak for the tomcat, but other airplanes, where autothrust actually moves the thrust levers, e.g. B737, use some kind of clutch system, allowing the pilot to override autothrust inputs. if this is the case in the F14, this could also be implemented with your 'idea 1', if the belt is flat (no teeth).

[draconus]

5 minutes ago, HILOK said:

if this is the case in the F14

Yes. https://www.heatblur.se/F-14Manual/general.html#throttle-controls

Quote

To allow selection of auto mode the throttles need to be between 75 to 90% rpm, the gear handle needs to be down and with no weight on the wheels. If these criteria are no longer met, the throttles are manually overriden with force or the Cage/SEAM button on the left throttle is depressed the solenoid releases the switch and it reverts to boost.

 

[Gareth Barry]

So, an update for anyone who might still be interested. I have got some of the main parts 3D printed finally. The delay was trying to figure out an efficient way of engraving text in Freecad. In the end I gave up, and might instead laser cut and engrave a sheet of acrylic to act as an overlay. In the end, I will supply all of the raw 3D models so that anyone can make any changes as they see fit.

Finally ordered some of the electronics, so will be making some progress in terms of wiring everything up shortly. Still got to finalise the model of the wing sweep handle and print the stalks for the handles. Nevertheless, I can see an end in sight. On one of the pictures, you can see the '4-degree notches' that a spring-loaded protrusion from the wing sweep hand will slot into. You can also see the large slots at the fully forward and overs swept positions to allow the handle to be stowed.

 

[Gareth Barry]

In terms of the autothrottle, I am having some second thoughts, I must be honest. Whilst I would like to solve the technical challenge of getting it to work, and I think i would learn a lot, I think it would be a heck of a lot of effort for something that I never use....

I was wondering, how many people actually land with autothrottle? I tried it for the first time the other day and I absolutely hated it! I was fighting it the whole way round the pattern. Personally, I use tiny increments of throttle, backwards and forwards constantly, in order to make glide-slope adjustments. I can see how applying backstick pressure is functionally the same as manually adding more throttle, but I just could get the feel of it. It made me wonder, how many people, either in the sim or real life, actually landed with this thing? Would be interesting to know.

[draconus]

55 minutes ago, Gareth Barry said:

I can see how applying backstick pressure is functionally the same as manually adding more throttle

It's not the same. Both will correct the GS and change AoA but aft stick will slow you down.

No, I don't use AT either but it felt good and easy to fly with last time I checked.

[Gareth Barry]

OK maybe I phrased that wrong, and maybe I am flying the pattern wrong in auto-throttle, dunno? But for me, when I start the final turn from just aft of the beam position on the carrier, I need to add a bit of throttle to compensate for the 25ish degree angle of bank, otherwise my sink rate is too high, and end up at 300 feet or so, having to intercept the glideslope later than I'd like. It's nice to be at the correct height more or less as soon as the final turn is finished and Im lined up, so that I can begin retrimming the aircraft as soon as possible for that final approach on the ball. So the only way i could get similar results using autothrottle was by putting a bit of back-pressure on the stick, and yes it would make the aircraft slow, which would then force the system to add some power, which is what I needed anyway? Anyways, I will try and use the autothrottle again later today and see how it goes. With all things, practice makes perfect I guess!

Would be interesting to hear from Victory 205 as to how often auto-throttle was actually used in case 1 recoveries.

[draconus]

18 minutes ago, Gareth Barry said:

So the only way i could get similar results using autothrottle was by putting a bit of back-pressure on the stick, and yes it would make the aircraft slow, which would then force the system to add some power, which is what I needed anyway?

Oh, in the AT case, yes, you're right.

[Lt_Jaeger]

From what I get, no self respecting aviator would fly a case 1 in autothrottle. Case 3,on the other hand is something totally different and as soon as we talk ACLS, there is no way around AT. Too bad that a bug with FFB flys me right into the drink of I use ACLS. 

 

[RustBelt]

Just 200 more serial numbers to go until Heatblur can get their Rhino FFB stick delivered . At 3 a week......

[Lt_Jaeger]

24 minutes ago, RustBelt said:

Just 200 more serial numbers to go until Heatblur can get their Rhino FFB stick delivered . At 3 a week......

As soon as they have that thing, development will stop. They will fight for a slot to fly their products in VR with the Rhino bouncing them around. 

Love the rhino. 

 

 

[Gareth Barry]

Regarding force-feedback sticks, I have actually put some thought into that. The way I would probably implement it is by using a spring+lever, with a fulcrum that is movable, and the position of which is controlled by a stepper motor+lead screw. The other end of the lever is attached to the joystick, such that moving the joystick applies force against the spring through the lever. As indicated airspeed increases, the stepper motor moves the position of the fulcrum such that more force is needed for a given stab deflection, ideally so that the same force is needed for a given pitch rate regardless of indicated airspeed. I suppose a similar mechanism could be used for roll, but having read through the manual, it seems that artificial feel in the Tomcat was limited to springs in the roll axis and bob-weights in the pitch axis. BTW, this whole moving fulcrum with a spring+lever isn't really my idea, I was reading how artificial feel was implemented in some jets. For stick shake i might use pneumatic/air jets or rotating counterweights on a motor to create vibrations, as has been done cheaper 'force-feedback' sticks. This might seem somewhat crude, but I think I could get by with a much smaller motor doing it this way, since the motor wouldn't be operating in a stalled condition, essentially turning the current into heat. Also, I am not really clever enough to deal with things like 'anti-cogging' of motors and the like, whereas this alternative way seams quite simple. 

In any case, this is still a long way in the future, I still need to finish this project, then move onto 3d printed pendular pedals, and get my simpit with various buttons to be functional and match the real thing. Once I've got that, it'll probably be time for a PC hardware upgrade and then finally VR.

Still working on this, waiting for the person who does my 3D printing to finish up the next batch of parts. In the meantime I'm still editing the CADs, cleaning up any errors etc.

[RustBelt]

The implemented FFB in the Tomcat is tuned to direct motor FFB sticks like the Brunner, Rhino, and the old Sidewinder FFB2. The feel is good even through a backlashey 20+ year old FFB2. Bobweight could be stronger, But then so could the FFB2. 

Using a system like you described will need a translator to convert the DirectInput FFB info to your Control Loading system. Or completely ignoring the Very well implemented FFB Heatblur has written and going full self coded reading from DCSBios.

I can HIGHLY recommend the VPForce KITS which are cheaper than a prefab Rhino, and also readily available as oppsoed to a year back wait list. Anticogging and everything pre solved and DirectInput native. 

Edited October 12, 2023 by RustBelt