Motion Software

[IanH1960]

Hi again all,

There seems to be quite a bit of motion platform hardware and software development work going on, great to see and to further add to the mix I've done a bit more work on my motion driver software.

I've released a new version (Beta (still!) v0.9.1) which should now run with FSX in addition to FS9. It uses FSUIPC as its interface so you'll need an up-to-date version installed. Note if you are installing FSUIPC with FSX please read carefully Pete Dowson's installation information which can be found at the top of his forum.

There are several improvements other than support for FSX. There's a new setup application which allows all the previously hard-coded configuration settings to be set by the user and saved in a config file. You can save as many different configs as you want. You can now set output COM port, data mode, start positions, working ranges etc etc as well as control the washout rates, saturation levels, pitch & roll rate cueing etc etc - have a look. There's a new User Manual (which helps!).

There is also a new version of the microcontroller programming I use which works with the setup application and allows the main controller performance settings to be made from the PC rather than by re-programming and re-flashing the chips - this makes it much easier to tune-up the rig.

It's all still Beta though - so anyone who tries it is really a Beta tester. If you do have a go please let me know how you get on. I'll try and find time to make any changes that might be needed.

Something to keep you occupied until the open-source motion software team build up the momentum.

Ian

PS it's all here

http://buggies.builtforfun.co.uk/Sim/sim-downloads.html

[Dutchboy]

Thanks for the update Ian. I'm curious why you decided not to release this version as open source?

Dutchboy

Hi again all,

There seems to be quite a bit of motion platform hardware and software development work going on, great to see and to further add to the mix I've done a bit more work on my motion driver software.

I've released a new version (Beta (still!) v0.9.1) which should now run with FSX in addition to FS9. It uses FSUIPC as its interface so you'll need an up-to-date version installed. Note if you are installing FSUIPC with FSX please read carefully Pete Dowson's installation information which can be found at the top of his forum.

There are several improvements other than support for FSX. There's a new setup application which allows all the previously hard-coded configuration settings to be set by the user and saved in a config file. You can save as many different configs as you want. You can now set output COM port, data mode, start positions, working ranges etc etc as well as control the washout rates, saturation levels, pitch & roll rate cueing etc etc - have a look. There's a new User Manual (which helps!).

There is also a new version of the microcontroller programming I use which works with the setup application and allows the main controller performance settings to be made from the PC rather than by re-programming and re-flashing the chips - this makes it much easier to tune-up the rig.

It's all still Beta though - so anyone who tries it is really a Beta tester. If you do have a go please let me know how you get on. I'll try and find time to make any changes that might be needed.

Something to keep you occupied until the open-source motion software team build up the momentum.

Ian

PS it's all here

http://buggies.builtforfun.co.uk/Sim/sim-downloads.html

[RobiD]

Hi Ian,

I have been waiting for Jim to finish his motion software and saw your post. I have had a look through the manual and it looks interesting.

One big question. Can I use the Velleman K8055 board for positional feedback and the KT5198 board from www.oceancontrols.com for the motor contoller using your motion software?

Your input would be greatly appreciated.

David

[RobiD]

opps, there's an .au on the end of that. www.oceancontrols.com.au

Sorry

[IanH1960]

Thanks for the update Ian. I'm curious why you decided not to release this version as open source?

Dutchboy

Good Morning Dutchboy,

I didn't go open for a couple of reasons. One is that the software is starting to represent an unreasonable investment of time and I thought I should probably take additional steps now to protect my IP rights, and the second is that all the important bits of method that might be of interest to other developers are actually apparent in the existing GPL version (Beta 0.6.5) which is still available on the site. The new version doesn't really add any new method.

Support for FSX is provided courtesy of FSUIPC, Pete Dowson has written it so that applications that use fsuipc3 with fs9 will run on FSX through fsuipc4 unless there are other, non-fsuipc reasons preventing them. The maths of the algorithms is mostly the same with some minor tweaks, and the new GUI coding is AHK specific and not of any use to others writing in C# for example. If a developer is interested in building AHK GUI's there is a great manual and active forum on the AHK website.

So I thought that making it available free but not open-source would still benefit builders who might want to try it and give my time investment a bit of additional protection. And in any case developers can still see the underlying methods in the existing stuff.

Ian

[IanH1960]

Hi Ian,

I have been waiting for Jim to finish his motion software and saw your post. I have had a look through the manual and it looks interesting.

One big question. Can I use the Velleman K8055 board for positional feedback and the KT5198 board from www.oceancontrols.com for the motor contoller using your motion software?

Your input would be greatly appreciated.

David

David,

I'll have to look at the spec of the controller board you mention as I'm not familiar with it - I'll look into it and get back to you.

Ian

[Prof Bill]

So I thought that making it available free but not open-source would still benefit builders who might want to try it and give my time investment a bit of additional protection. And in any case developers can still see the underlying methods in the existing stuff.

Ian[/QUOTE]

Hi Ian,

I believe you are doing exactly the right thing in the manner you intend making your software available and I am a supporter of your process!

As soon as one goes "Open Source" you usually find that "legal eagles" from other domains and disciplines start crawling all over everything you do to see if there is any infringement or added value they can claim based on your hard work. It is a bit more complicated than this but I guess you have already given this serious thought!

I am a great believer in making things available free but I believe a user should acknowledge and attribute to the work of the creator and originator.

Jim, Dutchboy, Wannabeaflyer, Roland your goodself and many others are prime examples of creators of IPR that should be acknowledged and attributed to when used!

Where would progress be without them!!

The whole Academic and University paper publication system is only successful because it operates in such a manner!
The very essence and spirit on which this Forum is operated and works seems to very much support this manner of cooperative working. Long may it continue!

Bill.

[IanH1960]

Hi Ian,

I have been waiting for Jim to finish his motion software and saw your post. I have had a look through the manual and it looks interesting.

One big question. Can I use the Velleman K8055 board for positional feedback and the KT5198 board from www.oceancontrols.com for the motor contoller using your motion software?

Your input would be greatly appreciated.

David

Hi again David,

The short answer is no, not directly.

If I understand correctly the way in which these boards would be used then the software isn't compatible. I guess you will use the K8055 to port the position feedback information back to the PC and then use speed demand output from the PC to instruct the KT5198's speed and direction drive? (cf Jim's new system layout). In this case the drive software needs to do two different jobs - one is to derive appropriate motion cues and the other is to implement a "servo" type feedback control behaviour in which it calculates from the current rig position and the desired cue position the direction and speed at which the motors should be driven.

The motion software does the first of these jobs - the cue derivation. The cues are output in the form of position demands through a COM port to be picked up by an external "servo" control system which goes on to figure out how to drive the motors.

I did contemplate doing both tasks in the PC but in the end couldn't figure out how to protect the platform when my creaky old PC does one of it's frequent hangs. Even a modest pause in processing (say as a bunch of AI traffic becomes active or a new weather update is loaded, or even if the drive software hangs on a bug) might suspend the next update of the speed demand to the motor controller long enough for the rig to run into it's end stops. In the end I went for a system in which the rig position is controlled independently from the PC and any "loss" of position demand signal from the drive software results in the servo system driving the rig to the last demanded position and stopping there. Handy for general control, startup conditions etc etc

I do have a pretty dodgy old PC right enough and perhaps a newer machine would be reliable enough; there isn't much time to play with though if the output to the speed controller is interrupted.

I don't know the KT5198 at all - does it have the facility for some local programming? ie can you program it to cut the drive to the motors if the speed demand input isn't updated within a certain time period? Can it recognise when there is a loss of live data from the PC?

Just some thoughts. Sorry you can't use the software as it stands. I can add new data output formats but they are position demand outputs and can't be used directly to drive a motor speed controller which needs speed demand.

Ian

PS

Hi Bill.

[TronicGr]

The KT5198 needs PWM input to set the speed and the direction of the motor. Thus another microcontroller is needed to read the serial data and pulse the KT5198 with the appropriate PWM signals.

And the position feedback should be connected back to the microntroller to know how to drive the KT5198 with PWM to set the position of the motor(s).

I'm currently building such PWM controller that can cooperate with such PWM driven motor controllers including KT5198. Its not ready yet but you can see the progress of its construction here:

http://www.x-simulator.de/forum/tron...nput-t559.html

It will read data from Ian's BFF driver interface too (now debugging some baud speed issues with Ian on this).

Thanos

[Prof Bill]

Hi Thanos,
A very neat setup indeed!
I am most impressed with the work Ian has done.

I am using RoboteQ Dual DC Motor Controllers Model AX2550.

www.roboteq.com

They are rather expensive but I previously used them in another underwater experimental development rig! I have six of them and they have saved a tremendous amount of development time as you can really play around with their embedded microcode with some excellent tools that RoboteQ provide.

Their site is worth a visit!!

Bill.