SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
Old Dec 26, 2012, 02:18 AM
Registered User
Yashiro's Avatar
chiba, Japan
Joined May 2004
64 Posts
thanks!
Yashiro is offline Find More Posts by Yashiro
Reply With Quote
Sign up now
to remove ads between posts
Old Dec 26, 2012, 04:55 AM
If it has wings it will fly...
ckleanth's Avatar
United Kingdom, England, Birmingham
Joined Mar 2011
1,443 Posts
Quote:
Originally Posted by noobee View Post
the v2 has 4 input (ail, ele, rud, aux) and 4 output (ailL, ele, rud, ailR) pins.

to support flapperons, you would need another input (ail2) pin. this extra input would probably need to come from one of the 6 dip switches.
thats what I was saying
for normal mode the aux woud have the off/on function
for flaperons you can use the aux as input so you wont have the off function without mods
for quadcopter 4 in 4 out as you said but so you wont have the off function without mods
ckleanth is online now Find More Posts by ckleanth
Reply With Quote
Old Dec 26, 2012, 02:52 PM
Registered User
Joined Dec 2006
1,441 Posts
Quote:
Originally Posted by ckleanth View Post
thats what I was saying
for normal mode the aux woud have the off/on function
for flaperons you can use the aux as input so you wont have the off function without mods
for quadcopter 4 in 4 out as you said but so you wont have the off function without mods
ah i see. in your case, on the v2, you would prefer to use the AUX-IN for AIL2-IN, lose the remote master gain input but do not require any modification. i'll look into that when implementing v2 support. the permutations can get interesting:

rx3s v1 single_aileron - use AILR-OUT pin for AUX-IN
rx3s v1 flapperon - use MOSI/MISO pins for AUX-IN/AIL2-IN, keep AILR-OUT unchanged

rx3s v2 single_aileron - no pin reassignments needed (use AUX-IN for master gain)
rx3s v2 flapperon option 1 - solder wire from a dip switch for AIL2-IN
rx3s v2 flapperon option 2 - use AUX-IN pin for AIL2-IN, lose AUX-IN function (your suggestion)

does that sound right?

it seems like flapperon (or "real" dual aileron) support is discussed often (in the other thread too). did not understand why HK would support two AIL(L/R)-OUT but only a single AIL in. i suspect they weren't really planning to support flapperons, but provide two AIL out for convenience (don't need a servo reverser in some, very few, setups)
noobee is offline Find More Posts by noobee
Reply With Quote
Old Dec 26, 2012, 03:08 PM
If it has wings it will fly...
ckleanth's Avatar
United Kingdom, England, Birmingham
Joined Mar 2011
1,443 Posts
something like that..

I have now received the flasing tool to direcly connect on the cpu pins so as soon as i have my v2 I will give it a go. I havent checked the flightstab site site did you put the source code there? I want to have a read through it when the kids go to bed

regarding the HK 5V programmer i think there is a jumper in order to self power the RX3S. shall I put 3.3 on any input/output pin?
ckleanth is online now Find More Posts by ckleanth
Reply With Quote
Old Dec 26, 2012, 04:20 PM
Registered User
Joined Dec 2006
1,441 Posts
yes, the source code to support rx3s v1 is on the github site (linked from 1st post). feel free to browse through and suggest changes.

i have not got to v2 support yet (on vacation). but you should be able to connect to the v2 CPU and read the fuse/lock bits to see if the usbasp can recognize it.

i think you need to remove the jumper on the usbasp so that it does not power the chip at 5v through it's vcc line (the atmega168 is fine, it's the itg3205 that should not receive 5v). then, i think you need to power the rx3s (through the RX connection for example). there is an on-board regulator that will provide the 3.3v needed by the atmega168. i've attached a pic to show the connections.

i have not tried this method (using a 5v usbasp), there was a post in the other thread that mentioned the chip would be 5v tolerant. i just used a 3.3v capable usbasp programmer.



Quote:
Originally Posted by ckleanth View Post
something like that..

I have now received the flasing tool to direcly connect on the cpu pins so as soon as i have my v2 I will give it a go. I havent checked the flightstab site site did you put the source code there? I want to have a read through it when the kids go to bed

regarding the HK 5V programmer i think there is a jumper in order to self power the RX3S. shall I put 3.3 on any input/output pin?
noobee is offline Find More Posts by noobee
Last edited by noobee; Dec 26, 2012 at 04:57 PM.
Reply With Quote
Old Dec 26, 2012, 06:06 PM
If it has wings it will fly...
ckleanth's Avatar
United Kingdom, England, Birmingham
Joined Mar 2011
1,443 Posts
Quote:
Originally Posted by noobee View Post
yes, the source code to support rx3s v1 is on the github site (linked from 1st post). feel free to browse through and suggest changes.

i have not got to v2 support yet (on vacation). but you should be able to connect to the v2 CPU and read the fuse/lock bits to see if the usbasp can recognize it.

i think you need to remove the jumper on the usbasp so that it does not power the chip at 5v through it's vcc line (the atmega168 is fine, it's the itg3205 that should not receive 5v). then, i think you need to power the rx3s (through the RX connection for example). there is an on-board regulator that will provide the 3.3v needed by the atmega168. i've attached a pic to show the connections.

i have not tried this method (using a 5v usbasp), there was a post in the other thread that mentioned the chip would be 5v tolerant. i just used a 3.3v capable usbasp programmer.
I was just wondering if it would be worthwile (or desirable) to offer a modified firmware that just address the gains but retaining the original configuration, should one wants to go back to how the module used to work but with more liniarised gains. then the beta/alternative versions is the custom code

last a sugestion on how the gains affect the servo movement, when you change the pulse width for the servo and you add the oscilation, the gain being the amplitude, we need to put some math/logic how the pulse is modified so that if one uses the maximum gain value and the module has servo stick input the total pulse wont force the servo over the +90 - 120 range. this will dramatically increase servo life.
ckleanth is online now Find More Posts by ckleanth
Reply With Quote
Old Dec 27, 2012, 07:48 PM
Registered User
Joined Dec 2006
1,441 Posts
i find that the single AIL input with dual AIL output of the original RX3S V1 firmware makes little sense. it is better to support either
- single AIL in and single AIL out (thus using the extra AIL out for AUX in), or
- dual AIL in and dual AIL out (flapperon), which is what i'm working on now.

i need to refactor the code quite a bit since the use cases to reassign pins around seems more common (especially for v2). the current servo out code is quite rigid.

all servo output are always clamped at [1000us, 2000us] to prevent them from going "out of bounds". what kills the servo is probably the rapid corrections occurring continuously. i left my servo jittering and it was quite hot after a couple of hours.

Quote:
Originally Posted by ckleanth View Post
I was just wondering if it would be worthwile (or desirable) to offer a modified firmware that just address the gains but retaining the original configuration, should one wants to go back to how the module used to work but with more liniarised gains. then the beta/alternative versions is the custom code

last a sugestion on how the gains affect the servo movement, when you change the pulse width for the servo and you add the oscilation, the gain being the amplitude, we need to put some math/logic how the pulse is modified so that if one uses the maximum gain value and the module has servo stick input the total pulse wont force the servo over the +90 - 120 range. this will dramatically increase servo life.
noobee is offline Find More Posts by noobee
Reply With Quote
Old Dec 27, 2012, 08:43 PM
Registered User
Portugal, Lisboa, Cascais
Joined Sep 2012
365 Posts
Hi

I went and had a look at the excellent code you've developed, and I believe that you should keep it modular as it is now.

With some conditional compiling, it would be possible to have a free assignment of all the ports (pins, switched and potentiometers).

What I mean with this would also require a companion software (PC based?) to choose which function is assigned to which port. Imagine as a stupid example that someone wants to use 5 groups of pins to control servos of which pin group A is assigned to gyro X, pin group B is assigned to gyro Y and pin group C,D,E are assigned to 50% gyro X and 50% gyro Y for some specific behaviours and only input would be assigned to one single pin group.

Or a simpler case, someone that would only want to stabilize Z axis with Gyro Z with one input and one output but all the other axis would act on passthrough (could be a model ship with stabilizer fins like in real life to counter list)

So on one side we would have all the ports and on the other side all the available actions/behaviours. There would be restrictions to which behaviours could be assigned to which ports (no sense on assigning a variable rate to a dip switch) In the end this would generate a conditional compiling file that would then be burned to the controller.


I can elaborate some more on this idea, but I need to find a PC with Visio to describe this "would be nice to have" setup.

brgds

luis
lvale is offline Find More Posts by lvale
Last edited by lvale; Dec 27, 2012 at 08:56 PM.
Reply With Quote
Old Dec 28, 2012, 08:23 PM
Registered User
Joined Dec 2006
1,441 Posts
thanks for the feedback.

a goal was to allow flexible assignment of pins so it'll be easier to support various devices. the servo output code was still quite rigid and i'm almost done making that flexible as well.

flapperons (dual ailerons) seem to be quite popular in the other thread, so i started on that too. enabling flapperons means that i cannot enable the stick proportional gain for the aileron channel since the neutral positions can change.

the processing flow right now is pretty linear (and fixed) but the physical pin assignments are flexible. would be interesting if you could elaborate the use cases for a more flexible processing flow.

thanks!

Quote:
Originally Posted by lvale View Post
Hi

I went and had a look at the excellent code you've developed, and I believe that you should keep it modular as it is now.

With some conditional compiling, it would be possible to have a free assignment of all the ports (pins, switched and potentiometers).

What I mean with this would also require a companion software (PC based?) to choose which function is assigned to which port. Imagine as a stupid example that someone wants to use 5 groups of pins to control servos of which pin group A is assigned to gyro X, pin group B is assigned to gyro Y and pin group C,D,E are assigned to 50% gyro X and 50% gyro Y for some specific behaviours and only input would be assigned to one single pin group.

Or a simpler case, someone that would only want to stabilize Z axis with Gyro Z with one input and one output but all the other axis would act on passthrough (could be a model ship with stabilizer fins like in real life to counter list)

So on one side we would have all the ports and on the other side all the available actions/behaviours. There would be restrictions to which behaviours could be assigned to which ports (no sense on assigning a variable rate to a dip switch) In the end this would generate a conditional compiling file that would then be burned to the controller.


I can elaborate some more on this idea, but I need to find a PC with Visio to describe this "would be nice to have" setup.

brgds

luis
noobee is offline Find More Posts by noobee
Reply With Quote
Old Dec 29, 2012, 09:39 PM
If it has wings it will fly...
ckleanth's Avatar
United Kingdom, England, Birmingham
Joined Mar 2011
1,443 Posts
just found this.. maybe it will shorten development time?
HappySundays ported OpenAero to the HobbyKing i86 / Eagle N6 board.
ckleanth is online now Find More Posts by ckleanth
Reply With Quote
Old Dec 30, 2012, 03:06 AM
Registered User
Joined Dec 2006
1,441 Posts
cool! didn't know there was another project for custom flight stabilizers. looks to be based on the multiwii project from the gui app. the mixer looks pretty sophisticated too. am even surprised at the similarity of the goals/features (like stick position gain, stick controlled rotation rate).

Quote:
Originally Posted by ckleanth View Post
just found this.. maybe it will shorten development time?
HappySundays ported OpenAero to the HobbyKing i86 / Eagle N6 board.
noobee is offline Find More Posts by noobee
Reply With Quote
Old Dec 31, 2012, 12:19 PM
Deviant
sbstnp's Avatar
Joined Jul 2012
1,183 Posts
I got a V2 en route, so when you are ready to test things on it, I'll give a hand. Have a USBASP already.

Oh, and Happy New Year!
sbstnp is offline Find More Posts by sbstnp
Reply With Quote
Old Jan 01, 2013, 05:54 AM
Registered User
Joined Dec 2006
1,441 Posts
i'm still on vacation and will work on v2 support when i get back

in the meantime, i've been checking changes into the git dev (not master) branch (dual aileron support, updated servo out code, again and a bunch of cleanups).

and a happy new year too!


Quote:
Originally Posted by sbstnp View Post
I got a V2 en route, so when you are ready to test things on it, I'll give a hand. Have a USBASP already.

Oh, and Happy New Year!
noobee is offline Find More Posts by noobee
Reply With Quote
Old Jan 04, 2013, 11:10 AM
If it has wings it will fly...
ckleanth's Avatar
United Kingdom, England, Birmingham
Joined Mar 2011
1,443 Posts
got my all stuff and I'm ready to flash a v2 as soon as you have something mate.
mind you with all the discussion about how the v2 orange stabiliser works, and unless the math for the servo pulse compensation can handle 22ms and 11ms receiver frame rate I was wondering if one if the many switches on the v2 can be used for selecting the different frame rate.
ckleanth is online now Find More Posts by ckleanth
Reply With Quote
Old Jan 04, 2013, 06:56 PM
wjs
William
United States, MI, Brighton
Joined Oct 2012
1,966 Posts
Quote:
Originally Posted by noobee View Post
i find that the single AIL input with dual AIL output of the original RX3S V1 firmware makes little sense. it is better to support either
- single AIL in and single AIL out (thus using the extra AIL out for AUX in), or
- dual AIL in and dual AIL out (flapperon), which is what i'm working on now.
I agree. They probably should have instead duplicated the one Ail1 output to Ail2 to give people the std Y experience, and save the need for a Y cable. The Rev switch, should have then just PWM reversed 1 of the Ail outputs to give it the behavior of today (if needed). Reversing both would then just a simple matter of a tx change.
wjs is offline Find More Posts by wjs
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Sold Flight Stabilization System with Programing Card Woody_99 Aircraft - Electric - Helis (FS/W) 1 Mar 23, 2012 12:50 PM
Sold Flymentor Flight Stabilization with Field Programmer Woody_99 Aircraft - General - Radio Equipment (FS/W) 0 Mar 17, 2012 07:43 AM
Sold Totally Tricked out 400 size with flight stabilization installed, BNF Woody_99 Aircraft - Electric - Helis (FS/W) 2 Mar 12, 2012 06:56 AM
Wanted FY-30A Flight Stabilization System Casey_S FPV Equipment (FS/W) 0 Mar 05, 2012 02:40 PM