Does anyone know of, or would anyone be interested in contributing to, an open hardware spec for a USB joystick? I see there is already a discussion about Open Source Joystick FFB / DIY FFB Joystick[/u] but I am after information about just a regular joystick.
I'm currently working on designing a sim-pit for flying gliders and am less than impressed with the current selection of joysticks available to use as a control stick. I see there are a few good cyclic controls for helicopter sims that would no doubt work but these are very expensive and would only be an approximation of what I'm after, so I'm considering building my own.
I have spent a considerable amount of time looking for resources to use to build my own joystick but so far haven't really been able to find what I'm after. Currently I am trying to find a CAD model of a gimbal setup that I may be able to use and it was VO101MMaister's excellent design in [u]this thread that has drawn me to this forum:
Another thread I've found useful has been Cataclysm72's thread DIY Controller and Throttle in the "Elite: Dangerious" forum however, even though it's meant to be a DIY thread he stops short of actually making the plans available�though he has uploaded an early prototype of his design to ThingiVerse: Simplicatron Joystick and Throttle mechanics[/u]
A third resource I have been pulling design inspiration from is the curious product known as a [u]heli-chair:
Of these three design inspirations the heli-chair is the easiest to comprehend (and can probably be made with just an angle grinder and welder), Cataclysm72's has the most information available, however VO101MMaister's design is by far the best.
even though it's meant to be a DIY thread he stops short of actually making the plans available
I just shared my thoughts and my concepts, DIY simply means that I did everything myself, it was never meant as a detailed instructional build, so don't spread talks like this about me. This design (and my updated spring return) took almost 2 years of my spare time, and here you accuse me of 'not sharing the designs'.
We've spoken about this on the Elite forums and private messages, you'll have to put in some hard work yourself.
The manufacturer method used seems the more accessible for DIY, as avoid use of CNC, solder machines... the parts is laser cut in steel, a service more affordable than CNC.
even though it's meant to be a DIY thread he stops short of actually making the plans available
I just shared my thoughts and my concepts, DIY simply means that I did everything myself, it was never meant as a detailed instructional build, so don't spread talks like this about me. This design (and my updated spring return) took almost 2 years of my spare time, and here you accuse me of 'not sharing the designs'.
We've spoken about this on the Elite forums and private messages, you'll have to put in some hard work yourself.
Rob, I am not "spreading talk" about you or "accusing" you of anything. It is true that you have an excellent DIY thread in the other forum (which I have linked to as I think it is, well, excellent) and it is also true that you are not willing to make instructions available or discuss details about the design, and that's your own prerogative. I'm not actually being critical about that here.
What I am trying to do is have a discussion about designing a joystick and making the information available such that people are able to build it themselves. If you don't want to contribute to such a project that is perfectly fine but don't take it as a personal slight. If however you were willing to contribute I would be delighted and your input would of course be most welcomed.
Keep the peace brother, And I love your work, Slarti.
Take a look at Russian forums, are there that the "Next Generation" gimbals born with use of CAM for center. ...
Thanks for the links Sokol1. I seem to keep coming across Russian threads in this project for some reason but unfortunately it all ends up sounding like gobbledygook to me, even using GoogleTranslate. With translation turned on you can kind of get an idea of what you think is being discussed but you keep coming across translations like:
Quote:
... 5. Baur makes steel 2mm (3mm under bearings), what do you think of duralumin is much sad? Although that fact that that I do not. ...
Which I think means he is using material that is 2mm thick for the most part but uses 3mm material for parts that hold bearings however such material is not available in aluminium—but who knows, I could well be way off track. It's all a bit of a crap-shoot at this end of the translator I'm afraid
The Servocity images you post look very encouraging though:
Are these images from another thread at all, and are there any tools one can use to workout how to construct things?
I appreciate the idea is to no doubt physically 'play' with the pieces and fit them together like lego bricks but if I'm going down the Servocity road I'd like to be able to roughly know what I need before ordering the parts. Are there any resources I could use to try and determine exactly which parts I might need?
Slarti, PM me your email address and i'll send you a bag of pictures i scaped from the net while doing my research, including some early gimbal excercises from me. I have no energy to put them all in imgur etc (being sick, recovering from cancer et), but i'm sure you know what to do with it. I very much support your project i wish i had the energy to go full out.
Perhaps i'm just an old fashioned guy who can't get used to opensourcing everything please don't take it personal.
Glider pilot simdork here (I'm Sisu1a on the Condor forum :))... if you are making a sailplane gimbals, the pitch axis should be about 50mm longer than roll.
The centering mechanism is what defines the kinematics for the most part. Listen to Sokel1, he is pointing you the right direction for this. Most other methods of centering are rather unsatisfying...
I have a design for a FF sailplane stick that I've not pursued in a long time, but if Condor 2 ever rears its head I'll actually make it, or at least follow Roland van Roy's MSFFII hack since it works so well in that sim and has a great library of effects to draw from.
I also have a design for a static centering gimbals, that copies the trim mechanism from my SZD-59, which uses a long torsion spring acting on the stick in pitch, which you can adjust by setting it in one of several notches fore/aft along a plate adjacent to the stick, though it kind of digs into your thighs the trimmable speed range with it is from stall to a speed higher than I'm comfortable letting the plane fly itself at.
I've spent some time measuring various ships and the stick in it is pretty typical of a high performance sailpalnes, with 25cm in roll and 30 in pitch, from axes to trip of grip. I think it moves +/- 17deg in roll and +10/-25 in pitch but I'd have to look again.
Here's some rudder pedals I made from cheap $20 Aeronca Chief pedals off ebay and some scrap metal/pushrods... it still says my Logitech Attack III is ready when I plug them in :p
ps. a Schwinn 205p is a great recumbent exercise bike to convert into a semi-supine cockpit I gather you want to build http://imgur.com/a/UBiV6
Last edited by LocNar; 11/20/1611:02 PM.
#4313856 - 11/21/1606:44 AMRe: Open Hardware Joystick
[Re: robv]
Joined: Nov 2016 Posts: 26S_Bartfast
Junior Member
Perhaps i'm just an old fashioned guy who can't get used to opensourcing everything please don't take it personal.
No offence taken Rob.
I can appreciate how strange it can seem to simply give away completely for free what has taken years of heartache and tears to produce but that is pretty much how the open software/hardware movement works. Personally I think it's amazing that the movement exists at all, but I love it. If you had of told me 20 years ago that there would be 1000's of highly trained and professional people all over the world working in their spare time for years to produce software with the explicit intention of giving it all away for free, I would have said you were crazy, but sometimes the world can be stranger than fiction
In many ways the whole of the Internet can be considered a remarkable experiment in anarchy in that there is no governing body that orchestrates how things are to be conducted (especially in the beginning) and the only reason it works at all is because all those using it essentially just agree to communicate with each other in an agreeable manner. There's no real reason the Internet is as great as it is, or even that it even exists in the first place, it just happens to be here because there were enough people interested to communicate with each other and enough others to create content, and it was seen the content was good so here we all are today. The whole thing is astounding really if ever you stop to think about it. Anyway, the Open Software/Hardware movements are just a part of all that, and I love it!
Glider pilot simdork here (I'm Sisu1a on the Condor forum :))
Hello, LocNar. I've actually come across your pedals on the Condor forum before and must say they look most authentic. When I get to making the pedals I was thinking to hit you up for details
I have also started a thread on the Condor forum about the build but it hasn't received much love.
Originally Posted by LocNar
if you are making a sailplane gimbals, the pitch axis should be about 50mm longer than roll.
Interesting, thanks. I would never have realised that.
In regard to gliders in particular do you know at all what the dimensions of a typical control column are? Going by this image:
It appears the stick is swept back 120mm and about 240mm tall, which is smaller than I would have guessed but does that seem about right to you? I've done a lot of searching and ask in several forums but those above are the only dimensions I can find. I'm also thinking to make it out of 25mm tube (1"). Does that seem about right?
Originally Posted by LocNar
The centring mechanism is what defines the kinematics for the most part. Listen to Sokel1, he is pointing you the right direction for this.
I'm afraid I didn't quite follow what Sokel1 was saying there. I followed the links and tried to make sense of what I was seeing but I'm afraid it was all a bit lost on me.
Is this the part of VO101MMaister's design Sokel1 is referring to when he (I assume he) says:
Quote
the "Next Generation" gimbals born with use of CAM for centre
Originally Posted by LocNar
Here's some rudder pedals I made from cheap $20 Aeronca Chief pedals off ebay and some scrap metal/pushrods...
Yes, I do very much like these however I'm not so confident I'll be able to find the same on eBay myself. Fortunately though they don't look to hard to make from scratch, the actual pedal part that is.
Originally Posted by LocNar
ps. a Schwinn 205p is a great recumbent exercise bike to convert into a semi-supine cockpit I gather you want to build http://imgur.com/a/UBiV6
Yeah, that does seem to lend itself to conversion surprisingly well, but again I'm going to try not to rely upon "GumTree Bargains".
Non-Joystick Chat: While I don't want to divert the discussion from joysticks for too long my current plan is to hack a "Pello" armchair from IKEA: These are one of the cheepest armchairs in store and that single 'bucket' piece looks like it will mount in a cockpit with out too much trouble. I also won't mind cutting cushions and hacksawing a channel for the control-stick if I find I need the stick closer (control sticks in gliders do have a tendency to reside surprisingly close to the family jewels ;)).
What ever chair I opt for I intend to fit it to a base somewhat similar to the heli-chair mentioned at the top of this post:
This base will then fit into a frame that holds the monitors and has previsions for placing a keyboard and mouse etc:
The ergonomics of this designed was based off the image shown above:
By the way, the reason I'm keen to have the 'heli-chair' removable is that I intend to make it possible to drag the chair down to the local flying park, plug it into a transmitter and use it to fly remote controlled gliders with an FPV setup.
Stick should be ~300mm in pitch/250mm in roll from axes to tip of grip. In a glider it will look shorter than this because the axes are *below the seatpan.
As far as the rear offset goes, that varies a lot from ship to ship. It's to allow you to make full fwd pitch without hitting the panel, which needs to be close enough to reach with your hands to fiddle knobs. Too much offset and it will mash your balls when you pull back. My SZD-59 has about 5" of rear offset, and the grip is 9" from the panel, so the base is 4" back.
1" is good size, older ships had 7/8" (22mm). I've tried to find inexpensive bending, but no luck. I have found some aluminum crutches (polio type) sometimes have the perfect section you can easily cut out with a tube cutter of hacksaw and have a perfect glider stick though.
Those Aeronca pedals come up more often than you might think... search "pedals" in the 'aviation parts' category of ebay motors, I see them come up time to time still. Sure you can make some, but this is much easier and you end up with real chromoly airplane parts and by the time all is said and done might be cheaper than trying to fabricate them.
One problem I see with your cockpit layout is ingress/egress. Also, if you have your monitors coupled to the rest of the cockpit you are sitting in, there will be a lot of wiggling of monitors and it won't go away unless you overbuild everything with large metal tubing, well into the zone of diminishing returns. Just make the component holding monitors separate from the part you sit in, and if it can 'swing' into place or roll/slide into place after you are already seated it solves your ingress/egress issue.
I see you plan triple monitors... what glider sim that is worth flying supports this? Condor is really the only sim with realistic enough flight model to do actual glider flying justice and being stuck on DX7 is not capable of triple monitor. Of course there is the much rumored Condor V2 which has been promised by 'christmas' since 2007, though no one has said what year. The original thread about it got nuked, but I started one a while back that is currently in General that has lots of teases by OXO suggesting it's coming soon.
I really hope so, because I want to get my sailplane pit back in order and start virtual soaring... it's such a perfect scenario for VR, but not until we are out of the DX7 dark ages...
Looks to me that might be a tad bit difficult to get in and out of without having a bunch of swing away arms.
Yes, I havn't yet got to thinking about ingress and egress.
As it is though there is space for one to swing their leg between the chair and that upright holding the monitors: Possibly then with your legs astride the chair you may be able to put your weight on one leg and with your hand on the top of the chair for balance one may be able to draw the other leg over the seat, kind of like getting off a motorcycle. It probably won't be that graceful but hopefully it will work. I'm not exactly sure, and to be honest I haven't given it a great deal of thought but I'm sure I'll work something out.
Stick should be ~300mm in pitch/250mm in roll from axes to tip of grip. In a glider it will look shorter than this because the axes are *below the seatpan.
Are you saying the pitch and roll axis are actually hinged at different locations?
I understood from what you said before that the two axis had differing ranges of movement (as in differing degrees of deflection when at full lock) but do you actually mean they are 'hinged' at different heights. Kind of like this example from a Piper (but around the other way - ie. short roll and long pitch): Is that what you mean?
Originally Posted by LocNar
Those Aeronca pedals come up more often than you might think... search "pedals" in the 'aviation parts' category of ebay motors, I see them come up time to time still. Sure you can make some, but this is much easier and you end up with real chromoly airplane parts and by the time all is said and done might be cheaper than trying to fabricate them.
I'll keep an eye out. I suspect eBay's aviation section is not nearly as well resourced in Australia as it is in other parts of the world but I'll keep my eyes open. At present there is a pair of Tiger Moth pedals going for 180 AUD but something more suitable may well show up. We'll see.
Originally Posted by LocNar
One problem I see with your cockpit layout is ingress/egress. Also, if you have your monitors coupled to the rest of the cockpit you are sitting in, there will be a lot of wiggling of monitors and it won't go away unless you overbuild everything with large metal tubing, well into the zone of diminishing returns. Just make the component holding monitors separate from the part you sit in, and if it can 'swing' into place or roll/slide into place after you are already seated it solves your ingress/egress issue.
The monitor stand is actually a seperate piece to the chair piece:
It's kind of hard to see in the rendering but the idea is that the chair piece slides into the stand and the stand ends up being anchored by the chair itself. With this approach one can have the screens as close or as far as they please:
As I mentioned before though, the idea of having the chair removable is actually to facilitate taking it to the field and flying RC aircraft rather than aiding ingress/egress as such. I'm also considering making it possible to mount wheels to that back cross piece to make it easier to drag down to the park. I don't know how feasible it's going to be to realise this part of the project but, well, it's still in the plans at the moment
Originally Posted by LocNar
I see you plan triple monitors... what glider sim that is worth flying supports this? Condor is really the only sim with realistic enough flight model to do actual glider flying justice and being stuck on DX7 is not capable of triple monitor. Of course there is the much rumored Condor V2 which has been promised by 'christmas' since 2007, though no one has said what year. The original thread about it got nuked, but I started one a while back that is currently in General that has lots of teases by OXO suggesting it's coming soon.
I really hope so, because I want to get my sailplane pit back in order and start virtual soaring... it's such a perfect scenario for VR, but not until we are out of the DX7 dark ages...
Arr, sigh, yes. I have read that thread as well
There is of course nothing preventing this setup being used with only one monitor and that is how I will start as at present I have but one monitor to give but I figured I might as well design it such that it can accommodate 3. However until hell freezes over I mean "Condor V2 is released" the multi monitor setup will be restricted to flying FSX or XPlane, or what ever else.
By the way, although the sim-pit's ergonomics are fashioned after that of a glider it by no means means I only intend to fly gliders in it. After reading robv's (Cataclysm72's) thread I have been inspired to check out Elite: Dangerous so may well end up becoming an extremely laid back Commander in that universe (though I would use a regular HOTAS setup rather than the gliders flight stick). Although it would be a little odd regular aircraft and even helicopters would be perfectly viable to fly from within the 'bath-tub' posture of this pit, though I grant it would feel strange
Thanks Sokol1. I have tried to read through that thread but I'm afraid I just can't comprehend what it is talking about. There were these two interesting images which obviously relate to the image you've indicated:
These are clearly discussing the same design as one of the images Rob sent:
But I'm afraid the actual discussion is all lost on me.
I assume those two independent levers is what they are referring to when they keep talking about the "Top of the heart". Also what does "zagruzhateli" mean? I've read it several times in the translation but I can't workout what it means.
Well, I have just spent far too much of my night reading this Russian thread: КОН...;#1090;ик BRD-DS It's about the joystick made out of plate I posted the image of above, and I must say this badboy looks really good:
Unfortunately there is not much information in English about it however I did come across a related thread on this site. While the thread's mainly about his KG13 handgrip project there are mentions made of the gimble, and it's alot easier to understand than the Russian translation:
He also has a short thread about a equaly inspiring set of rudder pedals:
While neither of these are the Open Hardware project I was hoping to find/create they are nonetheless incredibly inspiring 'hacker-esk' products worthy of consideration.
Well, read Russian forums through Google Translator is really confuse, need do some "interpretation"...
"Zagruzathelli/lya" is what we call CAM mechanism.
Be use two lever in the mechanism simplify deal with asymmetry of one lever, basically one half of levers - identical in two sides - distend the spring.
Unless some specific made for helicopter piloting gimbals based on damper for retain the stick position if you want a return to center force over your stick, you need rely on springs.
All commercial joysticks available use a center mechanism base on "pincer" (CH, Cobra M5, VKB Gladiator, Cougar, early Tm) or "piston" (Saitek's, Logitech , Warthog).
Both system provide a "hard" (clunk) center position, and the "piston" has a additional drawback of require more force to leave center position, and no require additional force for move from X to Y axis or vice-versa after leave the center position.
A alternative method is opposite springs in "X" - what was used only in Cougar MOD (EvenStrain, NN_Dam, UberNXT), and some sticks for military simulation (HighRev SIm now Bugeye) this provide a very "soft" center.
CAM mechanism is way to deal with this inconvenients, providing a less clunk but tactile center position, can have a more "hard" or soft center based on CAM profile as well a progressive force through stick movement.
Yes, the length of the moment arm the stick makes from the tip of the grip to the center of the axis pivots. Longer for pitch -very small pitch inputs result in massive speed changes so the stick is longer in this axis, which is conducive to these very fine inputs that you tend to hold for lengths of time.
Roll is much shorter because you do much more vigorous inputs in that axis, and much more often. The shorter length accommodates this type of input. It also makes any normal joysticks that have x/y axes on the same height feel nothing like an actual glider, so if you are building from scratch you might as well copy the actual kinematics, it's quite an important ingredient in a sim.
You should pick up a pair of VKB's pedals. They are quite reasonably priced though I see now they are discontinued. Shame, a used pair just sold for $150 not long ago... They are the only off-shelf pedals that will feel like a sailplane, and they use cam centering so you would be able to study/rev engineer the mechanism for your stick. Can probably find a used pair for sale, they only have 1 axis so many get them as introductory pedals and upgrade later.
A alternative method is opposite springs in "X" - what was used only in Cougar MOD (EvenStrain, NN_Dam, UberNXT), and some sticks for military simulation (HighRev Sim now Bugeye),this provide a very "soft" center.
Okay, so I have a first sketch of a design for this stic.
This is what I have at the moment:
There are 3 components coloured dark, medium and light grey which are connected with 6x10x3mm bearings.
The dark grey piece can pitch forward 10 degrees and pull back 25 degrees inside the medium grey piece, and pivots 50mm below the roll axis:
The medium grey piece has 17 degrees roll available inside the light grey piece:
And this is how the stick looks in the top-left corner:
I'm not too sure of how the centreing mechanism's going to work just yet but as a first pitch I've just got three pegs and springs are to be stretched between these pegs:
I don't know how well this is going to work as it is just a first draft but you know, it's something to build on:
Looks good, though that sharp corner (Shempp Hirth?) is aimed straight at your balls :o
If you use this type of spring centering you will need very hefty springs which will take much more space than you have allotted for them, and req much more robust pins to support them.
Design-wise, you want the angle the springs form relative to the direction they are pulling to be much shallower than that. There is an exponential decrease in efficiency (spring applying force the direction you want) as you diverge from pulling perfectly in-line (180deg), and at the angles you have you are well up that curve already.
Looks good, though that sharp corner (Shempp Hirth?) is aimed straight at your balls :o
Hehe, I thought about that too, and momentarily thought how it could be potentially disastrous in a crash - then realised it's just a simulator
This curve isn't actually based on anything it was just a means to draw the grip back 5" (125mm), though I did partially get inspiration from the grip in the heli-chair. The idea though is to have the pole removable so that one can easily exchange it if so desired:
Originally Posted by LocNar
If you use this type of spring centering you will need very hefty springs which will take much more space than you have allotted for them, and req much more robust pins to support them.
Design-wise, you want the angle the springs form relative to the direction they are pulling to be much shallower than that. There is an exponential decrease in efficiency (spring applying force the direction you want) as you diverge from pulling perfectly in-line (180deg), and at the angles you have you are well up that curve already.
I agree, the centering mechanism is very poor. I did have the springs more opposed to each other at first but was then concerned that when one spring was extended their wasn't actually enough room for the other spring to compress so rearranged them to how they are now but I agree, it's a pretty poor implementation.
I've been trying to think of a neater/better way of working these springs and will probably opt for a CAM design (as all the serious projects seem to do) but at the moment I'm just having a little trouble visualising how exactly it works. I figured I'd just start off with slapping springs directly to the pins and figure out how to attach CAMs in due course
By the way, are their any joysticks that have mechanical trims built into them? I figured the trim in a real aircraft would actually adjust the neutral/resting location of the stick (as in cause the stick to come to rest somewhere not necessarily in the center of its travel). I take it the digital trims we use in our flight sims just adjust what input the neutral location of the joystick represents without actually physically changing the location of where the stick comes to rest. Is that correct? Are there any joysticks that do change the resting location? I figure it could be done by placing those CAM/Spring pins on a plate that the operator can rotate with either a wheel or a lever of sorts:
Possibly fly-by-wire aircraft actually do use digital trims that do not affect the stick location and for those it would be fine to simply use the trim in the simulator but for a real glider I'm pretty sure the rest location of the stick would have to be affected by the trim setting.
By the way, are their any joysticks that have mechanical trims built into them? I figured the trim in a real aircraft would actually adjust the neutral/resting location of the stick (as in cause the stick to come to rest somewhere not necessarily in the center of its travel). I take it the digital trims we use in our flight sims just adjust what input the neutral location of the joystick represents without actually physically changing the location of where the stick comes to rest. Is that correct? Are there any joysticks that do change the resting location? I figure it could be done by placing those CAM/Spring pins on a plate that the operator can rotate with either a wheel or a lever of sorts: https://dl.dropboxusercontent.com/u/84843025/Image%20Cache/Cockpits/50mm-offset/Trim-800.jpg
After re-reading LocNar's first post I see he was discussing something simular to this but using a "torsion spring" rather than a modifying the balance points of CAMs. Do you have any more details LocNar? Your approach sounds relatively straightforward.
Here's what LocNar said again to save hunting back:
Originally Posted By: LocNar
I also have a design for a static centering gimbals, that copies the trim mechanism from my SZD-59, which uses a long torsion spring acting on the stick in pitch, which you can adjust by setting it in one of several notches fore/aft along a plate adjacent to the stick, though it kind of digs into your thighs the trimmable speed range with it is from stall to a speed higher than I'm comfortable letting the plane fly itself at.
By the way, are their any joysticks that have mechanical trims built into them? I figured the trim in a real aircraft would actually adjust the neutral/resting location of the stick (as in cause the stick to come to rest somewhere not necessarily in the center of its travel). I take it the digital trims we use in our flight sims just adjust what input the neutral location of the joystick represents without actually physically changing the location of where the stick comes to rest. Is that correct? Are there any joysticks that do change the resting location? I figure it could be done by placing those CAM/Spring pins on a plate that the operator can rotate with either a wheel or a lever of sorts: https://dl.dropboxusercontent.com/u/84843025/Image%20Cache/Cockpits/50mm-offset/Trim-800.jpg
CH joystick gimbal has mechanical trims in similar way to your drawing, although they are just a DOS era legacy, when they are need for calibration process, since CH are all "vintage" (design, electronics..) from 1990's, are more cheap continues produce the trim that change the molds for remove.
You find CH soft "guru" Bob Church post (there on in CH Hangar, don't remember) recommend don't use that trims in USB sticks, seems because are liming the axis resolution in one side of the gimbal.
Since you have R/C background, CH gimbal use the same mechanics of old Futabas gimbals, before the "digital" trim.
Too bad that Sukhoi forum are now dead, or you will see that all this center mechanism, spring positions that you are "studying" there are done in that forum at ~10 years ago and the conclusion: CAM Gimbal.
BTW - Talking in CH gimbal, this "thing" rusting there is a "scaled up" CH gimbal done at 10 or more years ago by a colleague and a mechanic machinist used to build experimental aircraft.
Was done for dogfights in IL-2 Sturmovik, and for me result horrible for this purpose, very "cluck" and the big leverage amplifies gimbal play. I think with some tuning are suitable maybe for FSX "flights".
In a way I've inherited it - the guy leave there for install new USB circuit (at time use gameport) but meanwhile lost interest in "computer flights"... By myself it goes to the trash, but the garbage truck does not carry it. I just remove their Thrustmaster FCS B-8 like grip. "Rust in pieces".
Curiosity: Industrial joystick gimbal inside view, with "piston" spring system Saitek like and big... pot'.
Well, after a reasonably successful weekend of angle grinding and welding I've been able to knock together this somewhat functional prototype of the design:
These are the dimensions of the components from design:
After making this prototype I began to get concerned that 10� forward pitch might not be enough. 10� really is a small deflection and it feel strange to think that is all that's available. It's no doubt correct for a glider but it would seem hard to fly other types of aircraft with such limited forward movement, so I decided to do a bit of a re-design.
I ended up deciding to make the joystick capable of providing symmetrical deflection but implementing bolts that can be adjusted to limit the range. The idea is that the more you screw the bolts in the more the joystick's movement is restricted in that given direction:
This shows the setup for a limit of 10� forward:
Is there a reason that 3 of the bolts have the jam nut on the outside, but 1 on the inside. Is it to provide a longer stop at the top?
In short, yes.
The idea is that the holes for the bolts through the 3mm plate will in fact be threaded (such that the bolts will actually 'screw through' the plate) and the additional nut on the outside is, as Falstar correctly surmises, a 'jam-nut'. Once the bolt has been threaded through enough to limit the joysticks throw to the correct amount the jam-nut is synched up to the plate to provide tension and prevent the bolt from winding in or out unintentionally.
The two lower bolts have the jam-nuts on the outside because there simply isn't space to have them on the inside. The two upper nuts were originally on the inside (as their wasn't enough space on the outside) but when I came to wind that rear bolt out to allow the 25° deflection I found the stick was hitting the nut rather than traveling the full range that was intended. The simple solution to this was to place the nut on the other side of the plate however this necessitated cutting away some of the lighter-grey piece to provide clearance for the nut, which is why the upper bearings are now in the centre of those semi-circular 'ears'. After I made the modification I then found when I wound the front bolt all the way in to restrict the forward travel to 10° their wasn't enough space between the head of the bolt and the plate it was screwing into for the nut, so this nut was relocated back to the inside again where there was plenty of space. As these nuts are 'free-spinning' nuts (in that they aren't actually attached to the main work piece) and don't carry any significant load they can be easily swapped around or replaced with thinner low-profile nuts which is common for jam-nuts, or even left off entirely—it really doesn't matter.
By the way, these bolts are actually M5 bolts and the ones depicted are 30mm long. These bolts are surprisingly tiny when you come to hold them in your hands. Infact many of the components were smaller than I was expecting when I came to make them. I know what the dimensions were when I was designing it and I've thought about them a fair amount but you get so use to looking at zoomed in images on a large monitor that you tend to forget how small 3mm actually is The good news though was that although components were at times surprisingly small it all seemed to function as I'd expected
Well, I started a new job recently which is great because not only will I be able to feed myself again but should be able to afford to buy joystick stuffs. However, this has unfortunately also meant I havn't been able to devote as much time to this project as I would like, but I have been thinking things over when I can.
The next thing to work out has been a proper way of implementing the centering springs. I have actually spent a lot of time pondering this and while I very much like the look of the "Zagruzathelli/lya" system I'm not so sure I'll be able to make it work myself, so have been considering a 'Scissor' design as used in RC transmitters:
A user called 'ZAGNUT' on the 'RCGroups' forum kindly uploaded some detailed designs of a gimbal mechanism as used in RC transmitters. While it has taken a lot if thinking to get my head around (and I'm still not convinced that I understand how it works fully) it has nevertheless given me inspiration to design my own system:
The idea is that the two sides of the movement are meant to be able to swing independently but there is a spring across the top pulling the halves closed. There are then going to be pins on the stick which will pry the mechanism open as the stick is deflected away from the centre:
Here is an image of the mechanism mounted on the stick in the upright position:
Full forwards:
And full back:
And finally here is an animated GIF of the stick being thrown from full back to full forward:
I have no idea if this will work properly on a full sized joystick but it seems to work well enough in RC transmitters. Another fortunate attribute of this design is that both the fore and aft movement is resisted by the same spring so there wouldn't be any asymmetry concerns as mentioned by Sokol1. We'll just have to try it and see if it works
Classic 'pincher', 'tweezer', 'opposing rocker arm' design, found in many devices, mostly because it's easy to manufacture however this type of centering will be very un-glider feeling, at least in roll where you cross the center frequently.
This method creates a pronounced/sharp 'clack' as you cross the center as the load transfers from one lever to the other. Softer materials can make it less pronounced but it will never go away. Net result: it would drive me -crazy- trying to center a thermal, feeling and hearing tack tack tack tack tack tack tack...
Honestly the plain old simple opposing spring idea you were going with a while back was a better design for what you are trying to accomplish, especially if you also incorporate damping with it as well. Ether hydraulic (motorcycle steering dampers are popular but bulky, rotary ones like on VO101MMaister's cost a fortune) or simple mechanism of sliding parts under pressure using damping grease ('thick' does not = damping) like Nyogel 767a, like VKB has adopted on their mechanisms. Almost as important for realistic feel.
Unless I'm mis-remembering what why it was brought up, asymmetry concerns relate to cam centering, specifically designs like VKB where the bearing drives the cam and the cam pivots from one side. Those can not use a simple symmetrical profile because of the changing leverage depending on if the drive bearing is moving towards or away from the pivot (lever getting bigger vs lever getting smaller), so the profile is modified to offset the variable leverage.
However, if the cam drives the bearing like VO10Maister (and later Monoblok/VPC), the cam is then symmetrical because there is no changing lever length in the tensioning mechanism regardless which way it swings.
There is also Baur's mechanisms, which originally used 2 simple opposing cams originally to keep loads symmetrical (though his later cams were given offsets to be actually symmetrical), and another version that I've only seen a few examples of, is Romer's design here: https://youtu.be/JtroF-BjAc8
Hmm, I see what you are saying. With this design there definitely is contact and separation of components as one crosses the centre. While I would think that if you were to get the tolerances just right there should be a constant tension across the center there would however undoubtedly be a 'tack' sound and quite possibly a small tactile 'knock' as the components come into contact with each other. How irritating! I would think that it should be possible to eliminate the slop in the center with very precise engineering/tolerances but there is no way of avoiding the fact that components are coming into and leaving contact with each other across that center. How disappointing .
So I've been scratching my head about things again and have spent the last week or so pondering exactly what the curve of that CAM lever (or whatever) should be, and I'm pretty sure that even in VO10Maister's design the CAMs should be at least partially asymmetrical. This is because as the bearing is pushed up it will also be swung away from the centre line, precisely because the lever arm stays the same length. As in, as the stick is pulled forward (or pushed backward) and the CAM on the stick pushes that bearing up, the bearing will not only rise due to being pushed up by the CAM but it will also be swung away from the centre line by the lever arm and will therefor no longer be directly above the pivot point of the joystick. This means the curve of the CAM will likewise need to be 'shifted' to the side by the same amount in order to cause the same deflection of the lever arm. If the bearing were simply pushed up and always remained on the centre line the CAM profile would indeed be symmetrical however as it is both CAMs profiles will need to be 'bent' in the same direction. The degree of asymmetry may well be less than that required by the 'CAM lever' design but the curves on the CAMs will indeed need to be 'bent' towards the 'hinge' of the lever arm at least a little as the CAM profile gets 'higher' in order to maintain similar spring tension. This will have the effect of flattening out one side and raising the apparent steepness of the other, thus making the CAM profile asymmetrical :s
I implemented the mechanism on my joystick but have it inverted such that the CAM is on the end of the stick pointing down rather than on the side of the stick pointing up.
This is what it looks like at 10� forward with the CAM leaver up against the bottom of the stick:
And this is what is looks like with the stick 10� back with the CAM leaver in exactly the same position:
As can be seen the bearing is clearly separated from the CAM in the second image meaning if the CAM were to be swung up such that the bearing were to make contact with the CAM there would be significantly less spring tension acting on the lever.
Does anyone know a good way to calculate exactly what the profile of that CAM curve should be? I've been trying to work it out in my head but I'm afraid my talent for calculating such equations is not as strong as it otherwise could be. Ideally I would like a curve that for every degree of deflection of the stick will result in exactly 1mm of spring tension. Hopefully if I can get an equation for the curve I'll be able to print it out on a printer then glue it to a piece of steel then ever so carefully carve it out with an angle grinder .
However, if the cam drives the bearing like VO10Maister (and later Monoblok/VPC), the cam is then symmetrical because there is no changing lever length in the tensioning mechanism regardless which way it swings.
I believe that V0O101Maister and Monoblock profile is not symmetrical, maybe not too noticeable due their straight profile. VKB use progressive profile - by user request - to create "feel" of weight across the movement. Soon will offer more "linear" profiles as option.
In Baur double CAM only the side in extremity force against the spring, the other side became just a stop, so you have the same profile in each side.
German,Japanese and USA aircraft all used a mixer mechanism for pitch and roll on the bottom.Top rod slides back and forward for pitch and the bottom rod pushes on a bellcrank for the elevators.Top rod rotates for roll and has an arm at the front to rotate the aileron control.
Does anyone knows, how should desing and plan the CAM profile? Why assimetric the profile's left and right side? Anyone give me some infos the whole process? I have CAD softver, but I don't know how should start .
Does anyone knows, how should desing and plan the CAM profile? Why assimetric the profile's left and right side? Anyone give me some infos the whole process? I have CAD softver, but I don't know how should start .
The profiles are asymmetric because in one direction the mecahnism is moving towards the pivot point while the other direction is moving away from it, thus changing the length of the arm doing the work. This changes the leverage depending on which way it's traveling. To compensate for this, you need to offset the profile to maintain something close to symmetrical spring resistance.
There is a document floating around, though I'm not sure if it's supposed to be public knowledge or not so don't know if I'm supposed to share it but the offset is essentially determined by applying the involute of a circle to the profile.
Here's a translated thread that gets further into it: http://translate.google.com/translate?hl=en&sl=auto&tl=en&u=http%3A%2F%2Fwww.gavca.com%2FphpBB3%2Fviewtopic.php%3Ff%3D49%26t%3D24042
I think It's a little bit difficult me . The google translate doesn't help me a lot. There isn't a easier way to know what datas have to me for the design, if I would like to draw a own CAM profile?
I don't know, I follow CAM for joystick "born" since 2007/8 (RiP Sukhoi) and this discussion in Avia-forum is the more informative I see - but difficult to follow, and for me incomplete.
Looks that in sim communitty who know the matter don't reveal because think in makey money with this. No "Opensource CAM joystick".
Are (free) books around about CAM (more for automotive use), but the matter is complex, involve advanced math.
Does anyone knows, how should desing and plan the CAM profile?
Oh, I sure wish I did!
This really is what's holding me up on this project at the moment. At first I thought I'd be able to thrash out the maths without too much difficulty (after all I did manage to scrape a "PASS" in geometry while a school :P) but oh lordy! So I ended up drafting a mega-brain at work into helping me out on this one, and tell you what, it has us both flummoxed —though we have made some headway .
In particular, after reverse engineering this photo: It seems the relationship between "spring stretch" and "joystick deflection", rather than being linear, seems to be more quadratic.
Previously I had thought there would be a constant amount of extra spring tension produced for every additional degree of stick deflection (as in 1mm of string stretch for every 1° of stick angle), ie. a "lineal" relationship, but it in fact seems that not only is the spring tension lighter near the centre but the rate at which this tension increases is also a lot less near the centre, and I think this is even evident in the second half of this video:
It's very hard to see but to me at least it appears that the spring is indeed being stretched more in the final few degrees of movement than it is in the first few:
Edit: Oh, and by the way, here is a translated copy of the Russian text mentioned above if anyone is interested: Calculation of the loader It's was only translated using Google's translation service but if you can manage to read between the lines some you can half guess at what it's going on about. It definitely seems that someone has done some fine work in that document but I am not yet convinced that the Involute Curve is actually related to this cam problem. The Involute Curve is most definitely related to the shape of the teeth on gears (as demonstrated romantically in this video) but I am struggling to see how it relates here. Perhaps my Russian just isn't good enough but I don't as of yet see the connection I'm afraid.
It seems the relationship between "spring stretch" and "joystick deflection", rather than being linear, seems to be more quadratic.
Yes, this is characteristic of VKB CAM, to add (based on simmers wishes) the "feel" that control surface (e.g. aileron) became heavy as you move the stick, the CAM profile angle change according for add this "feel". Some VKB buyers don't like this "feel", they wish for more linear "feel".
The CAM document (of avia-sim) purpose use Involute Curve as start point based on this premises:
- The force of the centering spring should be transmitted to joystick lever through the CAM. - This force transfer should be as linear as possible. - The mechanism should not increase the effort required to move the joystick lever more than the center spring resistance. - The effort required for move the joystick away from the center position should increase smoothly and the return to the center position should not cause mechanical shock. - The return of joystick to the center position should be tactile for hand.
![[Linked Image]](http://farm8.staticflickr.com/7075/7180914928_a369dc6ae2_z.jpg)
![[Linked Image]](https://dl.dropboxusercontent.com/s/75t9wiajlckzo5f/Gimbal.png)
![[Linked Image]](http://www.heli-chair.com/images_public/product_images/chair_bergen_pavement.jpg)
![[Linked Image]](https://dl.dropbox.com/s/w4wwukw0a1f1cr6/cockpit_sizing_3.jpg)
![[Linked Image]](https://dl.dropbox.com/s/nog8r7wymjf3mko/NextGenCAM-600.png)
![[Linked Image]](https://dl.dropbox.com/s/yxhjbpngj9ts3wb/pello-armchair_IKEA.JPG)
![[Linked Image]](https://dl.dropbox.com/s/zww3sm2xy68fllz/Heli-chair_timber.jpg)
![[Linked Image]](https://dl.dropbox.com/s/cjbtkyjuvibyuv6/Cockpit_timber.jpg)
![[Linked Image]](https:/dl.dropbox.com/s/vrllfwoq2zcars9/Cockpit_3.jpg)
![[Linked Image]](https://dl.dropbox.com/s/toqvjastutd9pdg/TimnerMannequin_flipped-trim.jpg)
![[Linked Image]](https://dl.dropbox.com/s/hc4fnuoqu99e0fx/Simpit_front-ish.jpg)
![[Linked Image]](https://dl.dropbox.com/s/125ptpgrb0p6ube/Piper.png)
![[Linked Image]](https://dl.dropbox.com/s/78f2vo6v2k1rpuj/Screens.jpg)
![[Linked Image]](https://dl.dropbox.com/s/jq04rn4e5qdtf0e/FarScreen.jpg)
![[Linked Image]](https://dl.dropbox.com/s/p64oenpz1oyl4o7/%D0%91%D0%B5%D0%B7%D1%8B%D0%BC%D1%8F%D0%BD%D0%BD%D1%8B%D0%B9.JPG)
![[Linked Image]](https://dl.dropbox.com/s/z9qrkh9h7nftt3j/1482565.jpg)
![[Linked Image]](https://dl.dropbox.com/s/ipn98ta5oibcb4w/RobsImage.jpg)
![[Linked Image]](https://dl.dropbox.com/s/i5aov13qu0k3hsm/DS-1.jpg)
![[Linked Image]](http://avia-sim.ru/forum/files/_11_245.jpg)
![[Linked Image]](http://forum.il2sturmovik.ru/uploads/monthly_06_2016/post-842-0-60705100-1466626967.jpg)
![[Linked Image]](https://dl.dropbox.com/s/g8lt5yqwcay7trf/Centre.jpg)
![[Linked Image]](https://dl.dropbox.com/s/c6qn7iz9aybhjlq/Dive.jpg)
![[Linked Image]](https://dl.dropbox.com/s/8vta7cvkzez8c5r/PullUp.jpg)
![[Linked Image]](https://dl.dropbox.com/s/5lg4l6nezda3af8/Roll.jpg)
![[Linked Image]](https://dl.dropbox.com/s/1rhg2y1zkd86zc4/Spin.jpg)
![[Linked Image]](https://dl.dropbox.com/s/td1dvyskj2vnckf/Spring-neutral.jpg)
![[Linked Image]](https://dl.dropbox.com/s/gtq5b44k909fv5o/Spring-forward.jpg)
![[Linked Image]](https://dl.dropbox.com/s/5p6emdq4zhrh4ma/Spring-back.jpg)
![[Linked Image]](https://dl.dropbox.com/s/jrvnozexexmtlg6/Sticky-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/1pc448hswoeqdfv/Trim-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/uqvqjvvfj9v5hbi/InThePaw-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/bnxmy8l5t69b6kb/Ring-dimen-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/i3i8v5dee2gon9j/Bucket-dimen-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/7vs9506elw0aiy3/Sticky-dimen-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/onl5z7q0xwwh227/Pieces-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/qhjqvd6fk20546f/BoltLimit-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/1qm1zk093t8lekw/BoltLimit-dive-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/ie4y9956vdba8fk/BoltLimit-climb-800.jpg)
![[Linked Image]](https://dl.dropbox.com/s/qeywuc2zmk5672q/BoltLimit-roll-800.jpg)
![[Linked Image]](https://static.rcgroups.net/forums/attachments/3/5/3/3/0/a9557344-35-FRAMEASSEM.jpg)
![[Linked Image]](https://dl.dropbox.com/s/1c6c0bmlmy01ynm/Scissors%20-%20high.jpg)
![[Linked Image]](https://dl.dropbox.com/s/v3oy1uzy0d088k6/Scissors%20-%20low.jpg)
![[Linked Image]](https://dl.dropbox.com/s/sl333x4nlqvtzlx/ScissorsPins.jpg)
![[Linked Image]](https://dl.dropbox.com/s/hy1s3sxb21x47yk/scissors-upright_800.png)
![[Linked Image]](https://dl.dropbox.com/s/vfil1fycad445gz/scissors%2B10_800.png)
![[Linked Image]](https://dl.dropbox.com/s/6i0h6r6dy34kvzc/scissors-25_800.png)
![[Linked Image]](https://dl.dropbox.com/s/hjrupg9juf55vof/scissors.gif)
![[Linked Image]](https://dl.dropbox.com/s/neyab4o9vq2c11e/Spring%2B10.png)
![[Linked Image]](https://dl.dropbox.com/s/c9jdg21j2cvc2yk/Spring-10.png)
![[Linked Image]](https://s22.postimg.org/tcp6o9m5d/CAM_Evolvente.jpg)
—though we have made some headway ![[Linked Image]](https://forum.il2sturmovik.ru/uploads/monthly_05_2016/post-12483-0-43064700-1462309677.jpg)
![[Linked Image]](https://dl.dropbox.com/s/rwjyt473wp59crp/QuadCam.gif)