Single engine collective

Simchair MKIII I2C single engine collective with AB412 right seat head attached
The latest rubber-band based version with the Huey switch panel


1 x 10x10mm aluminum square pipe
1 x 20×10 aluminum rectangular pipe for chair mounting (the mount for the IKEA GUNDE chair is included)
1 x SS495A hall sensor
1 x 10KOhm LINEAR potentiometer
1 x 6x6x4mm square magnet
4 x 3x40mm screws and nyloc nuts to connect the frame to its enclosure
2 x 3x20mm screw for lever connector axis strengthening
4 x 3x45mm screws for tensioner halves contraction
4 x m8x75 bolts, spring washers, and nuts
2 x 608 bearings (standard skateboard bearings)
1 x Arduino Pro mini
1 x Simchair MKIII I2C controller
super glue (cyanoacrylate)


Assembly is very similar to a twin collective and pretty straightforward. Will add the manual soon, meanwhile, if you have any questions, please contact me through comments!

There’s also a very detailed build log by Peter, be sure to check it out!

Download links

38 Replies to “Single engine collective”

  1. Peter
    September 13, 2018 at 23:13 Edit

    I also watched your build video but while it is helpful in some parts, many things remain unanswered. OK, I found out that I can not comment on the respective pages, so here goes. I will ask questions as my build progresses, and I will quote the relevant passages from the “simple collective” and the “twin collective assembly manual”:

    1) “Start with press-fitting an SS495A sensor into its socket.” – The sensor has a broad side and a narrow side. The sensor is placed horizontally, with the broad side facing downwards. Is that correct?

    2) “Solder its wires and bend the legs, fix everything with drops of super glue.” – How long should the wires be – 10-15 cm?

    3) According to this datasheet ( ), when the broad side is facing downwards then the wires from RIGHT to LEFT should be: ground, -, +. Is that correct?

    4) “Then, insert the lever connector into its frame and press-fit 608zz bearings. Insert M3 screws and put nuts onto it to fix bearings in place.” – So on the video you use ONLY ONE M3x45 screw which you put through both bearings, but no nuts? Then you use 2 M3x40 screws to strengthen the construction, but again no nuts?

    5) “Put the magnet into its socket.” – Which socket? Is this the “magnet holder” part? But where does the magnet holder go? It is never mentioned, and on the video it suddenly is on the side of the SS495A… The magnet holder is also an exccentric part, and I need to know how to orient it.

    Sorry for the stupid questions but this is as far as I got…. ¯\_(ツ)_/¯

    Thanks for your help so far! I am looking forward to continue.

    September 14, 2018 at 06:36 Edit

    Great questions! I will add page comments today)
    1) Yup, narrow side towards the magnet (flat side downwards)
    2) for ss495a, yes. Other wires should be longer as they’ll go through the lever. A good idea is to put printed parts on the table to get a rough estimate of needed wire lenghts.
    3) When the broad side is facing down, legs are towards us, from right to left: signal, GND, VCC(5v), or vcc,gnd,signal from left to right)
    4) It will be easier to see with parts at hand. You will need 2 screws to fix bearings in their sockets (do not overtighten!), and 2 short (20mm) screws to insert into the lever connector from both sides
    5) magnet holder goes to the side of the lever connector, where an ss495a socket is on the frame. You fix it (do not overtighten) with one of strenghtening screws you insert to the lever connector. The idea is, it can rotate for mechanical calibration of the lever. You adjust screw until you can rotate it with a slight force, so it wont rotate on its own. Then you insert a magnet, it should hold itself against the screw head. Orientation of the magnet holder: cut side towads the sensor, holder handle towards enclosure bottom lid (you turn it by holding this handle with pliers during calibration), choose default rotation angle so that when the lever is roughly in the middle of its travel, the magnet is above the middle of a sensor.

  2. Thank you for copying my comments and enabling posting on this page!

    With your helpful suggestions, I managed to make some progress. Here is a little gallery of how far I got:

    As you can see, i had to modify the socket for the sensor little bit, because my SS495a is a little too big. I would also suggest making the “slot” on the side (where the legs of the chip go) a little more trapezoid for easier routing.

    Also I would suggest increasing the size of the hole for the horizontal M3 screws (which fix the bearings in the gimbal) because I had to file the holes until the screw-heads were able to penetrate enough to be flush with the surface.

    I will try to continue withe the build and post more comments and questions as I got long. Thanks for everything!

  3. Everything looks good! A little warped though. What are the print settings your friend uses? Looks like either it’s not PLA or not enough layer cooling. If printed in other plastic or without proper cooling, parts will shrink and all holes, sockets etc will be smaller than they should. This is bad, because building the rest of the lever (where you need to put stuff onto the aluminum pipe) may become a pain =) You can try using a plastic part printed with the same settings instead of an alu pipe in this case. Still, I don’t recommend printing it in ABS, as all the tolerances and clearances are chosen with PLA in mind.

  4. You are right about the warped part! I will continue on and see if it is worth the effort of retooling and sanding each part…

    However I have some more questions:

    1) I am confused on how long should the aluminium pipe which goes into the “single collective lever” part be? On the video at around 15:00 you show the pipe, but the length it not mentioned. On the “simple collective” page the length is given as 25cm. And when I imported your models (.sldasm and sldprt) into Autodesk online viewer, the length of the pipe came out as around 17cm.

    2) Speaking of alu pipes: I guess that I will need more pipe to connect the head with the parts “single collective thottle frame p2” and “single_collective_head_i2c_connector_housing”. How long should that one be?

    3) About cable routing: If I understand your explanations in the video correctely, I will have to deal with the routing of 2 cables: 1x from the potentiometer inside the ” single_collective_throttle_frame_p1″-part and 1x with the ethernet cable coming from the bottom of the AB412 head part. My question is: How do I route the ethernet cable invisibly through the alu pipes? I can’t see a proper way.

    By the way: Uploading the .sldasm files was a great help in figuring this out in 3 dimensions!!! Thanks again for your help! I love this project!

    1. You’re welcome! =) You’re right, you have 1 3-wire potentiometer cable, and one 4-wire I2C cable (which is soldered to the ethernet cable and i2c pins of an Arduino board), which is cut of the same rainbow ribbon cable. One super important notice: do not touch screws (that are below throttle frame part 1) after wires are routed through the lever, you will cut them =)

      About the length of the pipe: just put printed pieces next to each other (to see how they will look when assembled) and cut the pipe so it would not protrude too far out of its sockets when everything is assembled. It can protrude a bit, but should not do it much. It’s best to press-fit the pipe into the lever connector before inserting it into the lever frame (it should be pressed until it touches the end of its socket, but if it will be too tight, press it as far as you can, just don’t drop the lever in this case, as lever connector may break). Do not forget to insert all of the screws, try to tighten stuff together to reduce backlash. After throttle 1 frame part, the I2C wire goes through a triangle-shaped cable channel up to its decorative cover, where it is soldered to the ethernet cable again. Use hot glue to secure ethernet cables in corresponding parts.

  5. 1) About the head/cable part: I think I can picture it in my head now. But seeing a picture of the underside of the head/collective connection would be helpful!

    2) Ethernet/IC2 cable connection. As I picture it in my head, you just use a short few cm worth of ethernet cable between the socket in the AB412-head and the collective level decorative cover. Wouldn’t it be a little more “elegant” to use a longer ethernet cable and just strip the isolation from the whole cable, except for the “top” part which is between AB412-head and decorative cover/triangular cable channel? And just cut the 4 wires I don’t need?

    I will post pictures of my progress soon. If you want, you may use any of them to include in your assembly manual (but I totally understand if they are too ugly… 🙂

    1. 1) Here’s how the head looks from the inside:

      2) You will like that ribbon cable as soon as you will get your hands on it =) Trust me, having a lot of single wire cables in a tight space (when you do scale metal stuff in plastic, space is always tight) isn’t what you want, really, but you probably have to try it to understand what I mean =) That ribbon cable just makes cable management easier, bends well when needed, is cheap and compact.

  6. Hello again, Alex! Thanks for your suggestions! I ran into some problems while assembling the alu pipe and screws (imprecise drilling, some broken thread…. have to dremel the screws *facepalm* but it’s absolutely permanently fixed now! *lol* ).

    The picture is very helpful! And after experimentally tinkering around with the parts and some spare wires I totally get what you said about it being easier with ribbon cable! Would you mind posting another picture? I would love to see the “underside” of the completed AB412head connected to the throttle – I need to see the cable routing.

    As always: Thanks for your suggestions! This is the most fun tinkering project I did so far!

  7. That picture was the completed head =) After that, you just put some duct tape around Arduino pro mini’s pins and stuff things into the enclosure =) As for cable connections inside of the “i2c_conenctor_housing” part, you just leave a bit of spare cable (ribbon and ethernet) there, solder 4 wires, again stuff things into available space so they won’t protrude, that’s it =)

    Or do you mean how the ethernet cable enters the head?

    The same thing with a Huey head:
    There’s a TJ8-8P8 socket pressed fit into the corresponding slot in the head.

    I am happy you’re enjoying building the lever! =)

  8. Hey, those are great pictures! Now that part is also clear to me, thanks!

    Unfortunately I hit another unrelated snag. Please see my image gallery (the bottom 3 pictures illustrate my problem):

    I’ll try my best to describe: The throttle grip has a little long tube inside. On one end that tube slides over the potentiometer “pin”, the other end goes into the “throttle_frame_p2”-part until the “upper end”. Now if I insert the alu pipe into the “throttle_frame_p2”, there is not enough room inside to also fit the little tube into there. It’s also impossible to fix the alu pipe inside the “throttle_frame_p2” because the screws will go right through the little tube.

    So here I am, scratching my head, obviously missing something totally obvious. Please help! 🙂

  9. Dont worry, that long tube is just the built-in support for the part! Different slicers handle supports in a different way, so its here to make sure nothing will go wrong. Just remove it, it should break easilly off the part that connects throttle grip with a cylinder ) I have to say though, you’ve done a bad thing – connected throttle frame parts without the throttle grip. It will be really hard to take them apart without breaking throttle frame p1 =(

    When attaching the throttle grip to the pot, push it with something by its inner central cylindrical part, not by the outer one. You may want to shape pot’s handle to the form of a cone for easier fit. Also, i advice you to resolder that pot – these wires will break sooner or later for sure, they may also touch pot housing, which will grant you various weird effects)

    Looks great otherwise!

  10. I tried your suggestions and it worked! After resoldering the pot cables, I filed the inner tube so it would fit very tightly over the potentiometer “pin”. Then I threaded the cables as you showed in the video, and put everything together.

    Luckily the screw which goes through the “lever_connector”-part was the only one I didn’t ruin! So it was no problem to remove the rubber bands holder and put in the M4-nut.

    I used the inner tube to push into the potentiometer pin, then I simply broke it off. You were right: It was easy to remove!

    Against your advice, I wanted to try it with a stripped ethernet cable. I managed to thread the wires through without problems. Now for the electronics part. I will have to contact you with questions regarding that. Until then: Thanks and all the best!

    1. Thanks =)

      Happy to hear you’ve been able to build the lever without too many problems =)

      One bad thing about ethernet cable, it can break if you bend it too much. Let’s hope everything will be fine (anyway, it’s easy to replace if something goes wrong).

      Electronics part is really simple: basically, you flash it, calibrate by adjusting map function values, and it works =)

  11. I hope that you are right!

    About the electronics part: Would you mind taking a look at my gallery (last picture) or this direct link -> . Because I tried parsing your instructions and coming up with the correct wiring scheme. Would you mind telling me if I understand it correctely and that is the proper wiring? Or if I am planning it wrongly?

    Meanwhile I will be moving over to the AB412 head section and starting my stupid beginner questions there…. (insert facepalm emoji here)


  12. Almost right! Two errors in your drawing:

    – Throttle goes to A1, SS495A goes to A0 (it’s better to follow the default scheme so you won’t need to remember anything in case software will be updated, it’s SUPER important for heads =) )
    – colors in the ethernet wire that are used are white-orange (5V), orange (GND), white-green (SCL), green (SDA), in that particular order from left to right while holding the cable with wire looking at you, contacts facing up (look at your ethernet cable). Why? It’s your standard ethernet cable crimping scheme (EIA-568B standard), it provides an additional reference for you to avoid making an error in the wiring. Also, you can use an ordinary ethernet cable for connecting stuff.

    Note that you may need to press the reset button on Arduino pro mini when it says “waiting for a programmer to respond” or something like that to flash it. Also, note that the head has 2 arduinos in it.

  13. Oh man, I knew I shouldn’t draw this when totally exhausted and tired…. Yes, I drew blue/white instead of green/white because of….. reasons, I guess….. ¯\_(ツ)_/¯ 🙂

    About A0/A1: I am a little confused, because on the simple collective page there is written ->

    “A0, A1 SS495A and potentiometer signal wires (A0 – throttle, A1 – collective, or change this in the ino file)”

    Now you say that it is the other way round. Actually I don’t mind, but I just want to make sure that there is no misunderstanding.

    One last question: What is the reason for you choosing to route the connecting ethernet cable through a hole in the base of the collective? I am pondering putting there another RJ45 connector to make it more modular.

  14. These sockets cost some money (when you puy 1-2 pcs) and can be hard to find, and there’s not much point in detaching this particular cable anyway. I will make a mod in a couple of hours, no problem =) As for the manual, I will fix it and check this part in the code. It’s better when unified =)

  15. I see your point! However I bought those parts in bulk at Aliexpress so I have a few spare ones lying around. No need to make a mod though, my trusty file and me will modify the existing opening. Had to file most other parts for hours already, so I am quite confident that I will fix that in no time!

    By the way, I think I found another typo, this time on the “twin engine collective” page. In the step #32 you write:

    “32. Route ethernet cable through a hole in the bottom lid of the enclosure and solder white orange wire to 5V, orange to GND, white-green to SCL and green to SDA wires, then solder everything to corresponding pins of the Arduino””

    Shouldn’t SDA be “blue”? At least according to all the other pages.


  16. Hello again! I am finally back at it, finished the mechanical construction and soldered the wires according to the sketch you approved. Had a hell of a bad time sorting out PL2303 driver problems, and learning how to connect and flash the ProMini correctely – but your youtube video helped! Unfortunately the collective is not working as it should and I am back to scratching me head again:

    1) When calibrating, where should the little “arm” of the “magnet_holder_5x5mm_socket_single_collective7a”-part facing?

    2) After much fiddling and flashing the pro mini about 20 times I manage to get a range of about 205 to 1170 on the collective. Is that correct?

    3) The map function does not seem to work properly. Even when using the syntax “z = map(z,205,1170,1023,0);” and checking in joy.cpl while moving the collective through the full range the displayed movement range is about 30% to 100%. What am I doing wrong?

    4) The throttle worked once with a range of maybe 23 to 350 but after flashing it now moves from around 70 downwards through zero to 63650. It seems obvious to me that this can’t be mapped properly, and when testing in joy.cpl the throttle does not react to inputs.

    I am fairly confident that I soldered everything properly this time and that there are no cross currents. But I will have to check again.

    Do you have any ideas where I should start troubleshooting this? I have no problem fixing the mechanical problems, but the electronics stuff is really frustrating to me…

  17. Hi! No worries, we’ll fix it.

    1) The little arm should point towards the rear end of the lever (you can turn the magnet holder with it), the open part of the holder should face the magnet.
    2) Comment map functions out before printing raw values! (that ADC is a 10bit one, so it can’t output more than 1023) Then, while looking at Serial.print output, move the collective to its lower position. Rotate the holder until you will see 0 on Z axis =)
    3) I think #2 will solve this as well =)
    4) again, that’s because you haven’t commented out map functions. If you will have 63650 after everything is done right, that may be because of resistor jitter. Increase its range in map function (e.g. if your raw values are from 3 to 705 try rz = map(rz,2,706,1023,0)

  18. Thank you so much! You were 100% correct: I forgot the step where i should comment the map function… *insert facepalm* OK, the collective range works perfectly now! Yay! This is great! I flew around a bit in DCS and it was wonderful, even though I have no chair-attachment yet and had to lay the collective down on a chair next to me! 🙂

    I did the same to the throttle but that part is not working completely: The raw range is 860 to 993. Seems a little small to me… I still inserted those values into the map function: map(rz,860,993,1023,0). However, in joy.cpl the throttle movement seems to me kind of like an exponential curve??? For 1 quarter of the travel there is no reaction at all then for 2/4 the raise is slow and in the last 1/4 the raise is very fast. Weird. I’d appreciate your input on this.

    I also wanted to kindly ask you for a mod to the chair attachment – where should I post about this? I tried to do it myself but I am more of a musician/hands-on guy and getting my head around Fusion360 will take way too long.

  19. Hmm, are you sure the pot is wired correctly? The middle pin is usually the signal one. Looks like you may have connected it like an SS495A =) The raw range should be something like 0 – 800 (200 deg of 300 available). The resistor should be of type B (linear, they usually have it written as B10K or B5K or A10K on them), but it will have somewhat the same range anyway. A means exponential voltage output.

    About the mod – no problem, I will need dimensions of your chair =)

    P.S. I am really glad you enjoyed flying with it! =)

  20. You may be right…. I looked at a few websites explaining how to wire a pot and and it it possible that I somehow mixed up VCC and signal wires during soldering. Oh man, I am terrible at this. Will have to check it out.

    One more thing: In joy.cpl I have 3 “Arduino Leonardo” devices. Is it possible to give them individual names like “Simchair MK2 Collective” and “SimChair MK2 AB412 head”?

    About the mod: I think that I could attach the collective onto my chair with a slightly modified “ikea_gunde_frame_collective_frame_mount_left_150_deg_ikea_gunde_frame02” and “ikea_gunde_frame_collective_frame_mount_right_150_deg_ikea_gunde_frame03″ part together with 2x”ikea_gunde_frame_frame_mount_back_part_ikea_gunde_frame04”. However I would need the first two to be not angled but instead vertical.

    Here is a picture I made of my current sim chair:

    Flying is a blast and who needs a throttle if you can adjust rotor blade angle! Thanks for your help!

  21. Sadly renaming joysticks is not possible as far as I know =(
    But these joysticks are placed in a persistent order, all axes and a stick are on the 1st one, heads are on the 3rd one, the second one is for some special stuff.

    I think making a mount for this chair won’t be a problem at all =)
    Will try to do it as soon as possible when time permits.

    You should try flying something like an R22 or Schweiser 300 without the governor, that’s where the throttle is most useful =) Or turn it off in a Huey and try flying formation while keeping needles in green, that’ll make you sweat for sure =)

  22. Thank you! This is a long term project for me and I don’t mind waiting. Am very glad that you are willing to do this after all.

    And I don’t think formation flying with or without governor would be a successful operation. I am glad to stay within the countries boundaries when trying to hover during the landing phase. 🙂

  23. So, I finally had a little time to myself and I flipped+resoldered the VCC and Signal wires on the potentiometer. And suddenly it is working perfectely! I could calibrate it and the axes work in joy.cpl like they are supposed to!

    YESSSSSSSSSSS! I made something that actually works! I could not have done it without your help. Thank you SO MUCH!

    Next thing I’ll start is the AB412 head and an extension for my Saitek X56. So you will hear from me soon on the AB412 page.

  24. That’s cool! You’re welcome! =)
    Have you calibrated it with the map function, just as the collective axis? Joy.cpl software calibration after map function can slightly reduce precision =)

    I will be working on a spherical bearing based cyclic gimbal soon, I think it may be a lot better than Saitek one (the idea is to make a gimbal that will support springs and force applications).

    Also, I am testing quite an interesting rubber-bands removal related mod for the lever =) If nothing breaks under load in a while, I will post it, flying with it feels great.

    Your questions are welcome =)


  25. WOW! I am super impressed with this system you’ve created! I’m going to start building it and am going to have to learn a LOT on the way, but I’m very excited! Just curious, on your thingiverse it says you’re an engineer, what type are you?

    1. Thank you, Ben, please feel free to ask questions here if any =)

      I am a system administrator/programmer/electronics engineer, something more, something less =)

  26. Hello again! Thank you very much for the new mount! I already sent the files to the guy who has a printer and will report back on my misadventures. 🙂

    1. Everything looks great! Will adjust that mount (it may be that bolts in different countries can have slightly different secondary parameters like head height). I strongly advise you to try the pneumatic mod- can’t really imagine flying without it anymore =) I have to say that, with or without the cylinder, a counterweight works better than rubber bands.

      I am now working on adding proper photos to every manual, hopefully, will add photos for the 412 head and a pneumatic mod in a few days.

  27. That’s great! Feel free to use any of my photos and modify them in any way you deem useful. I would love it if I could provide something useful.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.