Universal military aircraft homecockpit project

[Viper1970]

And ready!

The second one is just in print...

It fits perfectly over the tube from a bike saddle, that I use for my twist-collective and the heads as a base for the mounting. How this should look like if all parts are printed and installed can be seen as 3D models in my thread. I will use nyogel to dampen the twist-grips in the rotation. The gear segment is for a matching gear at the pots I use for the throttles input.

 

Edited February 28, 2022 by Viper1970

[Eugel]

Your output and craftsmanship are simply amazing.

 

[Viper1970]

Thank you!

Mostly the good results in printing are a matter of time. This twist-grip is 40mm x 80mm in dimension and a relatively simple shape concerning the print itself. There is nothing what makes the printer doing long travels or anything else that is very complex and it took nearly 10 hours with a layer-height of 0.12mm, an lined infill set to 60%, and a print speed of 50mm/s to finish it. On the other side there is for sure a lot of time saved at the reworks. It's mostly ready for painting.

My craftsmanship is far away from perfect . I have searched in some model maker forums for a good way to do paint works during the cold weather season without using aerosol cans or a airbrush gun inside the flat. There are some guys that do paint works with a simple brush that are really mind-blowing! I have tried it with all the tipps and tricks they told me, but my stuff still looks like rubish. I have to wait until summer and have to use spray paint for all the painting of the bigger surfaces at the parts.

My solder works are also far away from anything good, what makes me angry sometimes. I know that I have done some real good solderings in the past, but now I can't remember this. It's for sure a matter of my disease, but also has to do with missing patience. I want to get this damned dream now finished, after nearly 25 years and three other pits I started.

I hope I'm able to do get it done this time, but who knows what new BS the RL brings again. The overall situation is not looking very promising at the moment .

Edited February 28, 2022 by Viper1970

[No1sonuk]

Is there a reason you use straight gears rather than helical?

Helical gears have far less backlash.

[Viper1970]

Yes you are right! I totally missed this option during the 3d construction .

I hope it works precise enough for the turbine rpm. Normaly you rarely use it after the start up.

I also have no mechanical idle stop. I will do this function with a simple, but fully programable digital button. Not realistic and I thought about a mechanical stop, but the way I constructed my collective with changeable hats, making a mechanical stop is to complex and the stop mechanism would be to chunky.

Edited February 28, 2022 by Viper1970

[Viper1970]

Twist grips and the little spacer between them are done.

The spacer will be simply hold with two little screws in place if the rest of the parts is adjusted and I know where to put the holes exactly.

The spacer I had to rework a bit with my lathe to get it rotateable fitting inside the gaps of the grips.

Now I will switch back to my electronics of the command unit and also have totally forgotten that I still have to print some parts for my F-14, F-15 and F/A-18 throttles. That's the crux if you work at many building sites in parallel, to get things finished .

 

Edited February 28, 2022 by Viper1970

[Viper1970]

Doing further at the electronics of the command unit. Soldered all wires to the buttons, switches and encoder. The subd-cables for the matrix conneting the consoles are also done and soldered. Now I have to bring a little of system in that chaos and solder the wires to the matching inputs on my Pokeys board .

I mainly used already used parts from my old pit, that I had totally dissambled last year, for the command unit electronics. The new unused parts I want to store for the real pit.

 

 

Edited March 1, 2022 by Viper1970

[Viper1970]

UFC for the com-unit is done and working!

Now I have to make a new usb-cable for the programable cherry numpad and solder the subd- sockets at my console parts to be able to connect them with the main electronics in the UFC again.

The UFC has the fully programable cherry numpad with 21 functions, 24 button functions, 1 switch, an analog ministick and 7 encoders available (the 24 buttons are incl. the push functions of the mini stick and the encoders).

At the consoles there are 12 functions at the left one (3 push buttons / 9 switch functions) and 15 functions at the right one (15 switch functions / 3 rotary functions). 3way swtiches or rotaries only use two real functions, the middle position will be programed to a release function. This makes 3 position switches or rotaries fully useable with only 2 digital inputs.

With the Pokeys cards I can program every input as a DX-button or a keyboard macro. Even the analog axis are programable with DX-button presses or keyboard inputs. This can also be done to individual sections of the axis. The complete config can be saved as file and uploaded to the Pokeys card within a second. So it's possible to have different configurations depending on what you like to simulate.

 

Edited March 1, 2022 by Viper1970

[Viper1970]

!!!

Cable break at the heat-bed sensor! Fortunately I did a last look at the print before I went in the living room for TV watching. If I wouldn't have noticed it, the heat bed would have overheated, cause the sensor always detects 0° and so the printer will permanently heat the bed.

All of this must also happening in the middle of the biggest part of my F/A-18 throttle .

I will let it finish this print and will repair the printer for the third time in the last weeks now. I hope that I get no warping at this part, cause this already happened once, as i thought that I can shut down the heat-bed after the first layers, because I'm only printing PLA.

Although the results of the prints are really good, those little hobby machines aren't made for permanent use.

Edited March 1, 2022 by Viper1970

[Viper1970]

So, the printer is working again .

Here are the parts for my Warthog throttle convertions to F-14, F-15 and F/A-18 I already made so far. The last big part for the F/A-18 throttle is printing right now at the moment.

There are still many small parts missing like hats, switch-caps etc., but I want to print all the grip-shell parts first, to be able to do the spray filler and paint work as soon as the weather is warm enough. If I'm not able to finish those parts this year, I have to wait another year to be able to spray paint them.

That's the reason I do all the grip parts at the moment and let the printer run nearly 24/7, instead of finishing one HOTAS after another. If all grip-parts are ready and painted, things will slow down and I will mainly work only at one control at once, until it's finished. But it's important to have all the shells finished so that the work is not depending on the weather if you just need a finished grip-shell to get further, or in the worst case have to wait a few months to continue the work on it.

All the small parts like hats or similar things can also painted with a brush or even with a spray can if the weather is not perfect, but this is not possible for the larger parts. Here I need perfect conditions to get good results.

Some parts on the photo are already roughly sanded, while others are still in the condition like they came out of the printer and not even correctly deburred yet.

 

 

 

and the F/A-18 throttle parts put loosely together for a quick view. Nothing is deburred yet.

 

Edited March 2, 2022 by Viper1970

[rel4y]

Nice work and really clean looking prints!

[Viper1970]

Hi rel4y,

thank you!

Now I have another problem . The matrix I want to use for my consoles in my com-unit is not working correctly. At first I thought I made a mistake somewhere in my "high-tech" circuits I made at the backside of my console panels , but it looks like it is the Pokeys itself.

It works with the first part of the splittet matrix (the panel with the gear lever) without any problems or ghosting even with multiple switches activated, but when I connect both parts or even only the second one, I have multiple keypresses with flickering.

The diodes are working correctly and the "circuit" on the second console is also ok. So I did a test at the Pokeys directly with a simple short between the column and row pins and left both SUBD connections to my consoles unpluged, but still the same, The SUBD cables are also ok, I already measured them twice, to be sure they have no short.

All is working until I connect row 5 & 6.  Don't know whats happening here. I already used splitted matrix in my old cockpit to be able to control different panel parts at different locations with only one matrix. Have I made a mistake this time?

 

Here is the schematic I used. Column 1-5 of both panels are connected to the column 1-5 pins on the pokeys. This is done with an "Y-wiring" where eg. the column A from the left panel is soldered to the column pin A at the Pokeys card and the Column A from the right panel is also soldered on that same pin.

The rows are splitted into 1-3 for the left panel and 4-6 on the right panel. This makes an overall matrix of 5x6. Some functions aren't used, but the next smaller matrix (5x5) would have not enough functions available. The UFC is missing here in the schematic, but this part is working correctly and also did not use a matrix. I need the matrix for the consoles to have enough functions available for all the parts together with the use of only one Pokeys card. The colored functions are the different 3way functions, of which the middle position will be programed with the use of release actions.

 

Edited March 2, 2022 by Viper1970

[Viper1970]

So, I found out what's the problem and what I did find is real BS . It seems that the Pokeys doesn't like longer wires if you connect a matrix to it.

I tried to find a solution for this some hours now, but the only circumstance which makes it fully funtional is to go not further as arround 50cm with the wires connected to the matrix. This is a real problem cause I need arround 1m for the side consoles and also 1,5m for my connection between the throttle of my modded HORI HOTAS and the stick. For the stick I had planned to use a 5x5 matrix, which had given me 25 functions for the buttons and hats. With the additional 4 wires for the pots I had needed 14 wires and so only a sub-d gameport-cable to the stick would have been required.

Now it looks like I can use the matrix functions only directly in the proximity of the Pokeys card. I never had such problems in the past, cause I never had the need to use longer wires in my old pit. Although I have a lot of the Pokeys cards, cause I was able to get 10 Pokeys56U for low in an ebay auction some years ago and additionally bought 5 Pokeys57U, 13 of them are already reserved for my pit.

The remaining two I want to use with the command units HOTAS and the panels. I also need something like a Pokeys or an Arcaze, to be able to program the devices with keyboard macros to have them fully functional under an old OS like Win98 too, in case of my retro simulations. Here the HOTAS (the stick, the throttle and the rudders) must also be one device to use it with older simulations in older operating systems. So my first thought to buy two additional 57U cards for the pit and transfer two of the old 56U to the com-unit instead is also no solution. I have a real problem now!


But there is also some pleasant. I have printed the first part of my AH-64D Apache HOCAS (the right grip shell of the cyclic) and it looks really nice. I haven't done a lot of rework yet, cause I was fiddling arround with the matrix problem, but I'm very satisfied with the result at the moment. It was a bit of pain to get the supports out of the print, but with a little patience, the help of a cutter knife, a nail scissors and a little screwdriver it is possible .

 

Edited March 3, 2022 by Viper1970

[Viper1970]

Maybe there is a solution for the additional input problem with my Pokeys boards. I will try to look if I can get 74HC595 shift registers modules and put four of them in serial to have 32 additional inputs available. I hope that it's not the same sh.. with shift registers as with the use of a matrix relating to the wire length.

If this is also not working, I have to redo the complete electronics and make a new plan for making my HOTAS and the panels working in my command unit .

[Viper1970]

So, part number 2 of the cyclic is done and the last big part is in print .

The parts are really more smooth as I thought. At first I really considered to print them with 0.08mm layer height, but after giving 0.12mm a try I'm totally satisfied with the results. 0.08 would be a little bit nicer directly out of the printer, but it would also cost a lot of additional print time.

The parts are smoother as they seem at the photos, cause the reflection of the light let the layers seem much more noticeable as they are. There is not much sanding needed before coating them with the filler. Mostly only a quick pass to remove rests of the brim or other small residues of the print is needed and finally make a good adhesion for the filler.

The biggest part of the cyclic I'm printing right now, the left grip shell, will take arround 24 hours with 0.12mm layer height, a cubic infill set to 60%, supports set to 55° and a print speed of 50mm/s (10-15mm at the first layers). Wall thickness is set to 1.5mm and top and bottom are set to 1.56mm, the nozzle size is 0.5mm. The nozzle temp is 224° and the bed temp is 55°. As slicer I used Cura 4.13.1.

Edited March 4, 2022 by Viper1970

[Viper1970]

and its still printing... 

It's at 16:49 hours at the moment and there is still a lot of the grip missing

Reducing the time for a print arround the half would really be a great advantage, but this means printing with 100 to 120mm/s with the same quality. I have already thought about a second printer, an Ender 7 with a core x-y head system to achieve this, but reviews were not so promising in case of quality with higher speeds. Even arround 120-150mm/s the prints aren't very good, not to mention the quality at the ridiculous reachable speed of 250mm/s, Creality advertises it.

All the reviews I read reported that you have to reduce the speed most of the time to below 100mm/s to acheive good results and that isn't also the doubled speed to 50mm/s, cause at very fine details and a lot of short travels the printer does not reach this speed most of the time. To have an advantage here, you need big parts with wide surfaces at which the print-head really can reach those speeds. My HOTAS parts are mostly the opposite things in this case and the machine costs also still arround 700€. Cause of this circumstances I quickly dropped the idea again.

Another solution could be a bigger SLA-printer, but here you have the problem of cleaning and hardening the print with an UV-cabin after the print, which also takes a lot of time if you do more parts, even with a wash & cure system. And there is one problem the FDM doesn't have and that's the problem of the resin which must be able to run out of your parts during the print to avoid cured rests inside the cavities of the print. For this you must set additional holes during the planning in the slicer or even during the 3D construction. For HOTAS parts this isn't a nice solution and this prevents me from buying a bigger SLA printer on the other side.

In terms of quality the results are really adequate at the moment. Overall shorter print times are not really reachable, if you consider all steps for a print until it's really finished and ready for the filler, so I will still stay with my little brave Ender 3 Pro and hope that no more bigger failures appear until the mega print session is done .

If I really purchase a second printer somewhere in future (this is also a problem about a space for mounting it anywhere - only the pit-room would remain to do this), I would go for an Ender 3 S1. In my opinion it's one of the best solutions at the moment. It's based on the proven design of the Ender 3 series, has all the advantages I did on my own, like dual z-axis, a better print bed and so on, but it has the advantage of a double geared direct drive print head of better quality, a better firmware and a modern touch-screen system. For prints with SLA or PETG you don't need a closed building space and if the prints aren't extremly big and heavy a printer with a moving bed is still adequate enough, but mostly they are much cheaper as other system and also much easier to repair if something happens.

 

Edited March 4, 2022 by Viper1970

[Viper1970]

Ready!

Just a quick test how it looks if putting all parts together. I only used double sided tape to "mount" it.

Nothing sanded, not really deburred and also a lot of the support material is remaining.

All the small parts like hats, triggers, parts of the sockets etc. are also still missing at all my HOTAS parts.

I'm only printing the big grip parts at the moment, nothing else.

P.S:

My cyclic is the one the newer E-models now use, which has a momentary three way CMS switch (similar to the one in shape that is used for mode setting at the F-16 TQS) at the place the flare button is placed in the D-model. There is for sure also the option to use the button only, but I want to have as much functions as possible and always design my HOTAS/HOCAS after the one the latest models have. That's also the reason I did an F-14D throttle instead of the B version (and also made an extra additional D-model flight-grip). You can do all the functions of the earlier models, but there is still the option to use the controls with later models also, if they should be available somewhere in future.

 

Edited March 4, 2022 by Viper1970

[Viper1970]

The matrix I want to use for the electronics of my side/front panels at my command unit are running!!!

I have really tried everything! The last step was to solder pull-up resistors from each column to 3.3V. I got the tip with the pull-up resistors from rel4y (many, many thanks for this! - without you I never would have been able to get all running ) and it worked as long as I tested it at the plug itself which connects to the socket of the side-panels. But as soon as I connected the side-panel, things got worse again.

I measured my circuit at the panels again and again, changed the diodes to be sure there isn't a fault in them, measured the cables to exclude any sort of short and .... Nothing helped.

Than after endless hours of brooding and changing things, I decided to put the resistors to the 5V lane instead of the 3.3V. I had the fear to fry something on the board, cause the old manual of the similar Pokeys55U boards advises to use only 3.3V for pull-up resistors if even needed. For Pokeys56U and 57U there isn't such an advise anymore in their manual at all. All pull-ups are integrated, but it seems that this isn't enough.

So

I soldered them to 5V and all is running perfectly how it should!!! No ghostings no strange behavior, even if all of the buttons and switches are closed at the same time. WHAT A NICE DAY!

Edited March 5, 2022 by Viper1970

[Viper1970]

First Apache collective part is printed right now. Cause not much filament is remaining on the spool, I will now print some smaller parts of my twist-throttle collective. As soon as the new spool is attached I will print the remaining parts of the Apache collective.

The electronics of the UFC and the console panels look really ugly now, after desolder and resolder everything a lot of times. . Not a work you have to be proud of, BUT it works, so who cares . And you could not see it, once it is built-in the command unit, so !

I will get such a little china CNC as a birthday-gift this month from my wife. So in the future I can produce simple PCB's for such applications. I have an etching machine also, but it's not so comfortable to use and you have always the problem with the acid etc.

For the simple PCB's I want to make such a small CNC is ok and much more practical.

 

Edited March 6, 2022 by Viper1970

[Viper1970]

Not much happend here. I have some small troubles with my printer, if I print parts at which the head has to make longer travels between the sections to print. This is no problem with lower temps, but as I print with relatively high temps I got a lot of stringing at such sections.

Increasing the retraction helps, but this leads to underextrusion at the beginning of the next section sometimes. There is to much play in the bowden-system. Now I will upgrade to an direct-drive extruder with a dual gear. First I wanted to do this and some other additional modifications on my Ender 3 Pro after all HOTAS parts are finished, but the stringing makes me crazy.

So I will do the modifications now. Just have ordered the needed parts. Most will be done with stock parts, cause I'm a little bit low on money at the moment. The whole TM stuff has made a hole in my poket, which want heal very soon . I also wanted to upgrade to a silent board cause I'm still using the original old mainboard with the A4988 stepper drivers, which make an awful sound, especially during night-time prints. But cause of the financial situation this has to wait till the next money is available .

 

Edit:

And the first parts for the upcoming printer-upgrade. The missing belt tensioners I never did (the tensioners can be found at Thingiverse)

I will also have to change my spool holder inside the cabinet, to have a better feeding for the direct drive and do some other changes to the guidance of the eletrical cables going to the print-head.

I will do all these prints now, as long as the printer is still useable, so no more HOTAS parts until the conversion is done.

 

Edited March 7, 2022 by Viper1970

[Viper1970]

Y-Puller screw already broken as I wanted to strain the belt.

I had the fear that this would happen, cause printed parts are not so good at tensile strength. Printing the part horizontal to have the layers angled 90° to the tensile what impacts to the thread isn't possible with a 0.5er nozzle, cause the thread won't be exactly printed (and I have no desire at the moment to change it to a different size and adjust the whole printer configs again ). Now I do a very high temperature print with 245°, a layer height of only 0.08mm and an 100° infill.  Looks really good and smooth until now.

This would also be perfect for the HOTAS parts, but this little tensioner-part did already cost 4 hours of printing time. A grip shell, which now takes already arround 24 hours, would be arround 35 hours if I change to 0.08 LH and 100% infill. But it would really be cool, cause I think the difference between such a printed part and a mold wouldn't be big after all. Maybe I'll try this at one of the grips , but first I want to have most of them finished.

Edit:

and the second puller got broken just during the print of the new Y-puller . I will switch back to the original belt system without tensioners. I rarely adjusted the tension in the past and never had any problems. As useful as it could be if it would work, I think those parts should not be printed. Maybe PETG or ABS is more resistant, but I will try to do a metal or alumium solution for this somewhere later.

 

Edited March 7, 2022 by Viper1970

[Viper1970]

My birthday gift from my wife arrived . One week to early, but I couldn't withstand to open and begin to assemble it .

It's the smallest of those little china CNC's, it's only a 1610, but sadly even the 2418 is to big to fit at the place in my workshop, not to mention the 3018. I will use it for making some simple PCB's most of the time, so the 1610 should be enough, I hope.

I will 3D-print a few parts for it like a case for the controller-board, cause I did not like the idea of an unprotected mainboard at the back of the frame of a mill. I also will try to add end position switches for each axis if possible.

The printer makes big troubles at the moment and I hope that it will work flawlessly after the convertion to direct drive again. I think the extruder gear-box is damaged (spring and gear-wheel worn out), so I ordered a new metal dual-gear extruder. Cause there are some other things, like the convertion to direct drive, I already wanted to change, I will do it also just now. I have already started to make a 3D construction for my own type of a direct drive. I hope the printer will keep up long enough to print it's own repair and upgrade parts somewhat useable.

I also ordered a set of metal belt tensioners for the 3D-printer. They are so cheap, that doing them self makes no real sense at all.

If all those things are done, I'm back on the track with full steam ahead to get my HOTAS parts done until the "painting-weather" appears, I hope.

[No1sonuk]

Nice.
I have a 3040 and I added a laser.
You might want to figure out some kind of cover for those rails to stop dust collecting on them.

[Viper1970]

Yes, there are a few things I want to change or add, IF the printer works again. Those little china CNC's are far away from perfect, but they are really cheap and with a few addons they are good enough to do some hobby works. I was really surprised of the overall quality of this kit, considering the price.

And the new direct drive extruder in theory, the way I would like to have it I used some things from Thingiverse here and adjusted them for my needs. Screwings still missing. The head is very similar in shape like the old one was.

 

Edited March 9, 2022 by Viper1970

[Viper1970]

Still working on the modification kit for the direct drive at my printer. Had some trouble with it, which prevents me from printing the needed parts for the convertion.

I want to do this convertion, because I want to be able to print flexible filament and it's also better for high temp prints. The flexible filament could be nice for textured grip inlays, like eg. the original Thrustmaster X-Fighter B8 style grip had (or even the smaller Top Gun version).

Edit:

First parts of the convertion kit for my printer and the new version of it as 3D-model. I want to have the bowden guided to the back of the printer, cause I can not put the spool on top of it without a bowden. That's the problem with my printer in the cabinet.

I have to put the spool beside the printer mounted at the cabinets wall. I have constructed some parts to be able to do this, including a new adjustable spool holder with bearings and a guide for the bowden with an integrated filament filter.

Not a coventional direct drive layout and I will also loose a lot in height. But I never used nearly the full height of my printer (I always try to arrange things horizontal if possible), so that's no big deal.

 

 

Printer-head prints printer-head

 

Edited March 12, 2022 by Viper1970