For some considerable time, I have been converting a Symulatory throttle to a better specification using dc motors and a clutch. I am now very happy with the result and will in due course post a complete account of the work on Ian Sissons site to add to my previous articles.
Basically, I removed all the internal mechanics and installed 12 volt dc motors controlled by an Opencockpits USBDCmotors card and a homemade clutch.
My specs/requirements were
Maximum 12 volt supply with a current draw of around 1 amp
Sufficient torque to pull the levers back from a full fwd position, if I recall circa 120 in/oz
High gearing to get the torque and contribute to achieving the correct lever speed which is around 5 secs from idle to firewalled.( courtesy of Ian!)
The motors I used are 940D models with a reduction drive of 721 to 1 and the link is here
http://www.mfacomodrills.com/gearboxes/940d_series.html (http://www.mfacomodrills.com/gearboxes/940d_series.html)
There is a full data spec and pricelist on the site
Additionally, I used a belt drive and ribbed pulleys between the motor and throttle lever arm with a ration of circa 4:1. This gave me a range of speed movement that could be changed by either reducing the motor voltage or through the PWM of the controller card.. The motor works quite happily at 5 volt and still with adequate torque
Also, I needed a clutch as the motor gearing prevents manual operation of the levers and I attach pictures of the completed unit and components I made.
For your information, an explanation of the components from top left is:-
Polymer end bearing
Turned main drive shaft with grub screw to secure the motor, note also the "key" on the shaft to stop the pressure plate from turning. Threaded section to allow variable "force" on pressure plate
Polymer end bearing
Rubber friction plate
Spur gear for driving the potentiometer shaft
Ribbed pulley to drive the belt to throttle lever, with polymer bearing between pulley and main shaft
Rubber friction plate
Keyed Pressure plate
Spring to provide variable "force" on pressure plate
Threaded force adjuster
Ribbed pulleys and belts were sourced from
http://www.motionco.co.uk/timing-belts-timing-belts-c-25_38_27.html (http://www.motionco.co.uk/timing-belts-timing-belts-c-25_38_27.html)
I also attach photos of the completed unit and fyi I used three turn wire pots with gearing to ensure that I got full pot movement across the circa 80deg of throttle arc movement – this avoids "jittering" and allows for smooth slow movement. You will see on the "mounted motors" photo that the spur gear on the pot shaft has changed to a plastic one. I actually made the spur gears on my milling machine using a home made hob which was fine for the 80 tooth spur gear. However, my home made 12 tooth spur gear was pushing things to far and it would not engage properly!
I have had a few teething problems with the clutch ie the belt drive is a side load on the shaft/pulley joint and is "stressing" the plymer bearing. I may have to retrofit a needle roller bearing. The actual clutch "break" works very well and smoothly
The future article on Ian`s site will include how I made working trim indicators, spoiler arm with flight detent, backlighting, and automatic parking break release
Hope this is helpful – if you need any further info, just ask
David
(https://www.cockpitbuilders.com/proxy.php?request=http%3A%2F%2Fi24.photobucket.com%2Falbums%2Fc29%2Fjackpilot%2Fstunned.jpg&hash=44bba57911aac6dc9ae8066bde4c26343a2288c6)
whaooo!
David,
Thank you for the info. Right post at the right time.
I was wondering if you can post a little video of the system working.
Thank you
Roberto C
This is some SERIOUS project :o OUTSTANDING work !!!
Can you explain how the manual override would work with FSX (or FS9). As far as I know, if the autothrottle is on, moving the throttle levers has no effect. So, did you incorporate some sort of feedback circuit that shuts off the autothrottle if you do a manual override?
Anyway, your skill level is just plain amazing.
Maurice
All - Thanks for your kind comments and symbols :)
Jetpilot
If I can work out how to post a video, then I will. However, I can assure you that the lever movement is silky smooth with no jittering whatsoever.
Maurice
The clutch is basically there because you cannot manually drive the levers against the motor gearing - so the clutch enables manual operation. I am, however, programming the feature that if the levers, whilst under A/T control, are moved such that there is more than 10 deg of difference between them - then the A/T disconnects and does not reconnect until the difference is less than 10 deg.
This is, I believe, a feature on the real plane and is easily simulated in SIOC which controls the lever logic and through the offsets in Sim-Avionics.
Moving the throttle levers whilst under A/T control has no effect on my throttle, the levers return to the commanded position with no change in thrust
Currently, sipping much beer just watching the levers move :) :)
David
Maurice
Apologies, my original post was misleading about "manual intervention", hence your question. I have now edited and corrected the OP.
Regards
David
Quote from: fordgt40 on May 23, 2012, 11:45:15 AM
I am, however, programming the feature that if the levers, whilst under A/T control, are moved such that there is more than 10 deg of difference between them - then the A/T disconnects and does not reconnect until the difference is less than 10 deg.
David
Very cool feature!
Maurice
Awesome work David,
I am working on my real TQ using the method Rob posted in another thread and will be using OC motors card to drive the motors with SIOC as well . Your work is pioneering and I look forward to seeing more. I hope you will share your code also once done. Again...very nice work!
Scott
Hi Scott
The sioc programme is no problem :)
Just one word of caution re the OC DC card - it only takes 1 amp per motor and does run very hot! I had to put some heat sinks on the ic`s. You also need to solder some of the ic pins together to spread the load, unless you buy it already made and tested. If you need specific details then let me know
Regards
David
Good to know. My board is the premade one. If you have some details that would be great. I have to check my motors to see what kind of draw they have. Looking forward to meeting again next week at the get together if you are coming. Going to be fun discussing all this stuff in person!
Scott
Yep David is on the VIP list!
:D
Scott
Happy to repay the help you gave me when I visited and flew into the houses at the end of Midway runway :) :) Will try to find the info re the OC cards for you
Jack
Surely not a VIP, just an old hack :) though will be great to meet you in person
Regards
David
Quote from: fordgt40 on May 23, 2012, 11:45:15 AM
All - Thanks for your kind comments and symbols :)
Jetpilot
If I can work out how to post a video, then I will. However, I can assure you that the lever movement is silky smooth with no jittering whatsoever.
Maurice
The clutch is basically there because you cannot manually drive the levers against the motor gearing - so the clutch enables manual operation. I am, however, programming the feature that if the levers, whilst under A/T control, are moved such that there is more than 10 deg of difference between them - then the A/T disconnects and does not reconnect until the difference is less than 10 deg.
This is, I believe, a feature on the real plane and is easily simulated in SIOC which controls the lever logic and through the offsets in Sim-Avionics.
Moving the throttle levers whilst under A/T control has no effect on my throttle, the levers return to the commanded position with no change in thrust
Currently, sipping much beer just watching the levers move :) :)
David
That is really awesome work! Amazing... Just wanted to share some info for you on how the autothrottles work in the real airplane. Not sure if i read what you were doing correctly however maybe i can save you some time.
On the real bird when the autothrottles are engaged, assuming there in a commanded hold mode and not just in the Armed mode, you can overide them at anytime. When you do this nothing disengages and the power will respond to your input. Just be aware that the autothrottle motors will continually try to drive the power levers back, or forward to where the FMC is telling them to be. They degree difference does matter. If you override them, you will need to hold it there and you can actually feel the slight force of the motors trying to drive them. Does not take that much force to override them but its there.
I believe one of the challenges when using FS to drive the autothrottles is actually determining where FS thinks they need to be and getting your own software to attempt to drive them back or forward to that position when overriding them. I messed with FS driven autothrottles before and thats exactly what it was.. A mess! Lol. Full power, idle, full power, idle ..all it was trying to do was hold an airspeed lol..
Hope that helps a little.. Let me know if you have any questions..
Mike L
Mike
Great information and many thanks. I have some more programming now :). The problem is now determining when the throttles are being manually changed against the MCP commanded thrust. This could be fun!
Thanks again for your input
Regards
David
Amazing project, David!
This is another way of getting rid of the heat from the OC DC motor controller card.
A small 12Volt PC fan does the trick.
Vidar
Vidar
Great idea, many thanks for that. :)
Regards
David
Quote from: fordgt40 on May 25, 2012, 12:01:21 AM
Mike
Great information and many thanks. I have some more programming now :). The problem is now determining when the throttles are being manually changed against the MCP commanded thrust. This could be fun!
Thanks again for your input
Regards
David
No problem David! Btw, your not the only one that has more programming to do. Boeing does as well.
Few days ago we got a pilot bulliten about disconnecting autothrottles when in mountain wave turb or moderate to severe turbulance. Flying last night KORD to KSEA working around some wx in light turb, started getting a little stronger.. Next thing you know the power starts coming back to idle as we hit an updraft.. We both look at the throttles and realize whats going on.. He disconnects then A/T, goes to max continuous and it literally takes 15-20 sec for the power to get back up there.. Meanwhile were out of the updraft and into the downdraft and the airplane is approaching the yellow stick shaker arc on airspeed.. Had we not disconnected the authrottles, no doubt the airplane would have stalled itself! The bulliten warned of this happening and the slow engine respool time due to a software update.. Nice huu?? Lol..
Mike L
Great work David!
I am installing my 737 TQ into the base now, next step is interfacing, so I am going to pick your brain on the way from the airport to FDS. ;)
Nick
Nick
No probs - happy to help
David
Mike
I have heard about rotor clouds but did not realise they could have that much effect :(
Programming the "realism" into the throttle programming is proving one step too far, as you thought. The SA software author does not think it practical and the best I can do in SIOC is to try recognise manual intervention, save the thrust mode, disconnect the A/T, respond to the manual thrust and then reverse the process. However, how will I know when manual intervention has ceased :(
On balance, I think that I will just ignore any manual intervention and rely on the A/T disconnect button - Maurice you were right :)
Regards
David
Quote from: fordgt40 on May 31, 2012, 07:32:29 AM
Mike
Programming the "realism" into the throttle programming is proving one step too far, as you thought. The SA software author does not think it practical and the best I can do in SIOC is to try recognise manual intervention, save the thrust mode, disconnect the A/T, respond to the manual thrust and then reverse the process. However, how will I know when manual intervention has
Regards
David
Im just getting around to programming my throttles and i have a close setup to yours i believe. DC motors with a belt driven slip clutch system. Ill let you know how i progress. Im using phidget cards and c++. Well see how much custom code is required to get these things working. Well see!
Mike L
Mike
I think that apart from myself, there are at least two others here who would be very interested in what you are trying :)
Keep us posted
Regards
David
Yep, definitely Mike. Me too....
Scott
Looks amazing Scott! Are you making one for me before you clean up your workbench? :)
Quote from: fordgt40 on May 31, 2012, 12:30:11 PM
Mike
I think that apart from myself, there are at least two others here who would be very interested in what you are trying :)
Keep us posted
Regards
David
Will do guys... The way i think you need to approach this is to Not use any of the built in autothrottle functionality in Flight Sim at all. in other words just drive the throttles in your own code as if you were moving them yourself. Thats the overview. Only problem is this could require a complete autopilot rewrite because then autothrottles are closely tied into the autopilot system. Not sure yet but thats the way im approaching it.
Mike L
For anyone wishing to motorise their throttles, I see that opencockpits have brought out a better and more robust dc motors card
Link here
http://www.opencockpits.com/catalog/...ard-p-263.html (http://www.opencockpits.com/catalog/...ard-p-263.html)
Thank you David for pointing it out!
Right on time - I was just going to order OC motor/servo cards.
Nick
Just curious about a thought on this subject, what if you drill a small hole on one or both throttle handles just behind the top and facing forward. Then, you place an optic day/night sensor in one or both; therefore, when you place your hand and fingers around the throttle, you would cover the optic sensor, which would be used trigger the clutch for manual mode. Removing your hand it goes back to auto, as well as you could program a timer on the optic to something like:
if x(optic covered)=time limit, then stay manual mode, else auto mode.
Not exact language there, but enough to make the point.
Otherwise, I thought of a heat sensor to measure small difference from hand on throttle to hand off, but that is slower. Also, thought of pressure sensitive/resistance device in/on throttle, so when you place your hand on it you could do the same programming and work the clutch.
Thus, no real buttons, but of course this would have to be done without physically affecting the handles too much. And, this would not be exact as the real thing with degree differences...
All just thoughts.
John
Quote from: fordgt40 on July 24, 2012, 07:20:29 AM
For anyone wishing to motorise their throttles, I see that opencockpits have brought out a better and more robust dc motors card
Link here
http://www.opencockpits.com/catalog/...ard-p-263.html (http://www.opencockpits.com/catalog/...ard-p-263.html)
Crap, I had just bought their older one a month ago.... :(
You would think they would warn someone in the shopping cart that a new model was pending....
Scott
Quote from: XOrionFE on July 24, 2012, 04:59:51 PM
You would think they would warn someone in the shopping cart that a new model was pending....
.. and be left with a pile of old ones... ;)
John
An interesting idea that would tell the system when manual intervention is being applied. It should be possible, using this, to save the a/t parameters or mode, revert to manual then re engage the a/t mode. There is no need to disconnect the clutch (there is no facility), rather just turn off the drive motor.
The only downside I can see is that it is standard practice for the hands to "cover" the throttle levers at key times during take-off and landing. Could cause some fun if the opto sensor had too wide a beam :)
Good idea and many thanks - this forum has some real imaginative thinking outside the box!
David
I am , by all means, totally ignorant on motorization as my TQ is not (and will not be..)
But seems to me that if you use a clutch like Polyclutch , and a clutch being a clutch, it can be overridden beyond a certain force for manual op until you disconnect AT.
Please explain if I am wrong..
Jack
Jack
You are correct. The issue here is trying to tell the "system" that manual intervention is occurring. The system does not know if you are holding the lever and the clutch is slipping. Unless you can devise some very clever programming to calculate timescales for the lever to move where it is commanded under the A/T and then work out that it is being held - ditto released :(
David
...or just press the A/T disconnect button and be done with it. Not ALL aspects _needs_ to be simulated... ;)
Quote from: vidarf on July 25, 2012, 07:27:33 AM
...or just press the A/T disconnect button and be done with it. Not ALL aspects _needs_ to be simulated... ;)
Yes, of course. :)
Quote from: fordgt40 on July 25, 2012, 06:57:14 AM
The issue here is trying to tell the "system" that manual intervention is occurring. David
Just for my educ:
This is for a brief manual intervention? yes ?
If the system knows, it will not fight back and start where you left the lever when your intervention is over...is that right?
Or..?
I do not see the difference with diconnectint AT Intervening and reconnecting AT.
Feel I will learn smth here ...lol
:D
Jack
By all means disconnect and reconnect the A/T :)
On the real 737 the plane, when in A/T mode, will respond to a manual setting, but when the handle is released it returns to the currently commanded A/T position.
FSX will not do this - although you can move the throttle levers away from a commanded A/T setting (the clutch slips) the actual commanded thrust will not change - unless you disconnect the A/T, in which case it will then follow the manual setting.
I think this now takes it full circle :)
David
makes sense,
I made my life simpler with my custom 737 :laugh:
Thanks David
Sent from my PC from sunny Canada! lol
You're a better man than me Jack, I read this post and am still none the wiser.... ;)
I went the servo route and always disconnect the A/T if I want manual control, seems logical somehow. :) And it saves burning out the servos!
Joe.
Joe
No argument against that :). My experience was that servos did not give me the smoothness of lever movement, hence the use of dc motors and the "challenges" that they bring. :(
David
David, I have to agree. The servos are not as smooth in operation, in fact I bought an OC DC motor board and have the motors to go with it; but I just couldn't get to grips with SIOC.
So I took the easier yet more expensive Phidgets route. :)
Quote from: Joe Lavery on July 28, 2012, 02:18:57 PM
David, I have to agree. The servos are not as smooth in operation, in fact I bought an OC DC motor board and have the motors to go with it; but I just couldn't get to grips with SIOC.
So I took the easier yet more expensive Phidgets route. :)
So you have yours working with Phidgets motor controllers Joe? What a ionics package do you use? I woul like to hear more about your setup.
Scott
Scott I use PM, combined with the following:
Phidgets PhidgetAdvancedServo 8-Motor Board
Phidgets PhidgetInterfaceKit 8/8/8 Board
PhidgetMotorControl HC Board (this for driving the Trim Wheels)
The whole system is controlled by a program written by Diego called DVATX. It's freeware and works very well. I'm using 2 pretty hefty sail arm servos from Hitek and two DC motors from an ebay seller.
At the moment I'm in the middle of a TQ rebuild to encompass this kit, but I've had it all working quite nicely on the bench. The new TQ is more substantial than the one I built for the PC Pilot magazine article. I've taken a bit more time and structured it around a real TQ with some ideas from other builders, but of course not an all metal construction.
I've also been designing a set of clocks for the MIP, they don't have all the features of an original but they incorporate a standard 7 segment LED clock and a 7 Segment stop watch with lap time. I think that's all that's needed really. I've just finished the circuit board design but I'm not sure how many I'll build, it depends if anyone else is interested. They should be a around £100 each, which is about a third of others I've seen. They're designed using PIC chips which are very versatile, however it's taken me some time to get to grips with the programming. Luckily I have a friend who is a master, so I've been having a few lessons.
I'll post some pics when the boards arrive.
Cheers
Joe.
Hi Joe,
Are the MIP clocks going to be all inclusive with the Bezel and buttons, or will they be an attachment unit that could/would attach to FDS MIP Clock Bezels (non FDS working clock unit). It'd be nice to have these mount up with the FDS Bezels because they're already correctly sized for the holes. It would be disappointing to get something that is slightly oversized.
Will you please post some pics of your work on the clocks. If they fit, I will be interested...
John
Joe,
I'm also interested in your clocks... please keep me in mind.
Warren
Hi guys,
I have an Engravity MIP so I've based the clock on that. As I mentioned in my last post they do not function the same as a real clock, so they only have four buttons.
The clock part has 2 buttons, one for time set and the other to change the digits. The stop watch has 2 more buttons, one for start/stop, the other for lap time and reset. In which case the bezel will only have four button positions, not the usual seven as most MIPs show.
John, I am planning to include a bezel, but it will be based on the Engravity sizes because that's really the only reference I have to work with. It would be no more difficult to make a different size bezel (if I have the accurate sizes), but it would have to be a replacement because most come with button holes already cut out and they wouldn't line up with my buttons. :)
Once I have a finished article I will post some pics. I originally intended just to make a set for myself, but when I showed the prototype to a few friends they immediately said "I want one". So it looks like it might be a production run. ;)
As I said the first circuit boards are on order, then once I know everything works OK I can look at the bezel arrangement. I've attached my circuit design for the display board, there is also a second board with the PIC chips and transistors to drive the LED displays. That sits below the display board, like a sandwich. :)
This may not be the final version either, as I need to make sure everything lines up with the holes in my MIP. Yet again I can still change this to fit the FDS MIP if I can get the dimensions.
But naturally it's simply not worth doing for a single set, they would cost more than the price I'm charging. The circuit boards alone cost over £60.00 for a prototype set, So I'll have to see how many people are interested. I can get cheaper boards from China and Canada but once the UK VAT and money grabbing post office add their cut, it's the same price.
Incidentally if someone wants to send me the sizes of the FDS bezel and it's associated mounting holes, I can see how viable it would be to make two variants.
And finally Warren I will certainly keep you posted on my progress.
Joe.
Hi Joe
You can add me to your growing list. Put me down for 2 Engravity sized ones! :D
Sean
Thank you for the info on your TQ setup Joe. I have downloaded and checked out DVATX in the past. It may not work for me because my TQs are driven by motors instead of servos much like David's but I will look again. Otherwise, I also have the open cockpits cards so I can also lean on David's skills to help with the SIOC code :-) . One way or the other it will get done eventually.
As for your clocks, I would be very interested also in an FDS sized set. Looking forward to seeing you finished prototype. Maybe start a separate thread to gauge more interest as I am sure there are many many more that would like a set if they know about it.
Scott
Scott, It looks like there are more people interested that I would have thought. So I think I'll wait until I have the prototype and post a new thread to gauge how many I'm going to need to build.
I certainly don't want to go the route I've seen many other go and take orders (and money) without having the product ready to ship. This only generates bad feeling and mistrust. :-[
My idea would be to make a batch and only then offer them to those on the list that are interested. That way it would only be a matter of waiting for the post to deliver them.
Cheers
Joe.
Nice way to proceed Joe. ;)
Thanks Jack,
There are too many people out there willing to take your money and not so quick in supplying what you've paid for... I don't intend to be one of them.. ;)
I've been on the end of that particular scenario and it's frustrating as hell! >:(