Futaba M6 Fr Sky 2.4 ghz DIY Conversion

[Phil Green]

I found the calibration was easier if you get it to print the raw adc values before the mapping Paul, rather than after as per the your code - ie you hold full right and full right trim, note the raw adc value, hold full left with full left trim, note the raw value again, then plug these two numbers into the map line. Repeat with all channels.

To do this you'd comment out the whole "chan01=map(analogRead(A0),247,796,1000,2000); etc block,

and temporarily replace it with a block of "chan01 = analogRead(A0);" etc (just while you're doing the calibration), then the serialprint shows the values you need for the mapping lines. Plug the raw values into the map lines, uncomment the map lines, comment out the chan01 = analogRead(A0); lines and Bob Sherunkle.

I also put a "/n" at the end of the serialprint block so the adc results are nicely formatted rather than in one long line.

During the calibration you'll need to comment out the noInterrupt() for serial comms to work.

A neater and easier way would be to use a compiler switch, to set "calibration mode" on or off, to select the raw adc and serialprint block, or the mapping block.

Regarding electronics on RC Groups, maybe its just a personal thing but I feel a bit out of place there, the skill level of most of the contributors has me awestruck! Some of their projects are incredible. Multiprotocol transmitters. Orbiting Satellites (yes really). DIY Flight Controllers. All the reverse engineering work done by PB and others. Amazing!

I think this has a negative effect as 'ordinary' projects seem unworthy amongst these mega complex contributions. Also some of the 'regulars' can be quite curt with new contributors tresspassing on their territory!

The UK forum RCMF has a good 'down-to-earth' electronics section, though obviously much smaller.

Cheers

Phil

Edited By Phil Green on 30/05/2014 15:40:31

[Chris Bott - Moderator]

To calibrate I read the raw stick values via serial monitor using e.g Serial.print(analogRead(A0));
but I didn't comment out the mapping lines.

I read the raw values at stick extremes but didn't move the trims, so I'll have a little more movement than 1000 to 2000.

[Paul Luby]

Hi Phil

Regards mapping, that's what I'd been doing, obtaining the raw values and then popping them into the map command.

When putting stuff on forums for guys that may not familiar with this kind of thing, I'm always struck between explaining things in some depth or posting just enough so they can make it work. You can't believe the number of questions I've had in the past regards what I'd consider simple issues.

At the minute I think the sketch, and I hope you agree, is fairly easy to follow for most modellers. The best way for any article, will I think be to write another sketch that can be loaded that allows guys to record the pot values and then open the operating sketch, insert the pot values and then upload the operational sketch. I think I said in an earlier post, I could make the PPM assembly routine a lot smaller by using a loop type routine but the way it is makes it easier to follow I think, the sketch is I hope a good step by step routine. Maybe if someone mods a Tx, they can do a bit of research and put in a loop type routine themselves once they've are pleased they have it working with the standard sketch.

I think including a routine that allows reading and storing of pot values, whilst ideal, could cause confusion for software unfamiliar types. Maybe by making it easy to follow the first time round guys may decide they want to introduce an included calibration and pot value storage routine and write it themselves?

If one or two guys mod a Tx as a result of any resulting article then that's great. If one guy gets the electronics bug and talks about it down at his club spreading the good word, then that's one in the eye for lack of understanding.

God I sound like a teacher.

Mainly its so I don't get lots of "it's not working and I don't know why" e-mails.

Paul

(AKA Veri-Gash)

[Paul Luby]

Hi Chris

The "do I include, do I not include trims in the cal" question is down to the individual I think.

Most servo's in my experience have a range of 0.75mS to 2.25ms so you pays your money.

Spektrum radio, if you look closely, with trims centred, give less servo movement at full stick deflection than do Futaba and Taranis. I think its to stop over running servo's cause of all the mixing you can do and because modern ARTF modellers don't bother re-centring after test flight like us older guys do. They tend to just throw the gear in the model and use the radio to set it up. I think you can tell its a particular pet hate of mine.

Paul

(AKA Veri-Gash)

[Paul Luby]

Hi Phil

Just seen the edit and you've hit the nail on the head.

KISS, keep it simple stupid, should rule.

There's nothing worse than looking at a project you've found on the internet, want to get involved in and thinking WTH are they on about. It puts guys off from having a go straight away.

That's exactly what I want to avoid.

Some of the guys on some of the RCGroups projects get peeved when they get asked the same question again and again.

My missus says I can come across as abrasive, I do try not to be but I suppose 30 years in the military has programmed me. Now if somebody could write a be less abrasive sketch I could upload.

There's a great book I read recently called "How Apollo Got To The Moon", written by an engineer for guys curious about the mechanics of how it was done. I'm an aircraft electrical/avionic engineers and that book fascinated me and scared the crap out of me. KISS ruled.

Its well worth a read, if you want to borrow it, let me know and I'll bring it to the PANDAS meet.

Paul

(AKA Veri-Gash)

 

Edited By Pete B - Moderator on 31/05/2014 11:57:14

[Paul Luby]

Hi Guys

Here's the latest code version.

Enjoy.

Paul

(AKA Veri-Gash)

[Paul Luby]

Hi Guys

Here's some of the expo results from the algorithm I used.

Again enjoy.

Paul.

(AKA Veri-Gash)

[Chris Bott - Moderator]

The noInterrupt line kills my jitter completely

Thanks chaps..

[Paul Luby]

Hi Phil

Just re-reading some of the post and that pic u/c door controller is sweet.

A mate of mine who's into pic's does a gadget for air u/c that automatically lowers your gear if the pressure drops to a pot adjusted preset level. He saw them on a site in the USA and they cost about $80, he knocks the out for a few quid. The most expensive bit is the sensor.

Regards Dropbox, I've tended to keep my on line profile down a little cause of all my RAF travels and stuff I've done. I have a Facebook profile but that's friends and a couple of closed groups only.

Don't you have to invite people to access your Dropbox files? It would be great if files could be uploaded to this forum so guys could download any code or instructions that are included with any articles don't you think?

I'll have to have a closer look at Dropbox, RAFMAA uses it for its documents but I'll have a closer look to see about open access settings.

Paul

(AKA Veri-Gash)

[Paul Luby]

Hi Phil

Interesting that you came up with summing and averaging a sequence of stick pot readings.

I already use a similar method for an older friend with mild Parkinson's. I use a rolling average to generate the pulse length for each primary channel. Its allowed him to carry on flying for a little longer.

Expo on a bought Tx just didn't cut it. The rolling average with timing tailored to his, for want of a better phrase, shake frequency has been quite effective.

It sounds wacky but it is effective. I think of it as a dynamic electronic damper.

Paul.

(AKA Veri-Gash)

[Paul Luby]

Pete B

A Spektrum mAh sensor would be nice.

There's a lot that could be done with Spektrum but unfortunately Horizon are in bed with Apple as their latest AS3X Rx's that can be set up using an iphone show. They also seem to have adopted Apples information control policies.

Check out this link for info regards how tight Horizon control their telemetry protocol information.

I tried to get some Spektrum telemetry protocol info for a project I was working on last year and it was full of non disclosure agreement legal blah blah blah.

So now I've changed the project to work with Taranis. So much better working with the open source community.

If you want any bespoke telemetry sensors or types that Spektrum don't make then buy a Taranis, there'll be someone in the Taranis community doing what you want.

Paul

(AKA Veri-Gash)

[Chris Bott - Moderator]

Something like this I guess for the Spekky A sensor

Here's one I made earlier.

Shame it's someone else's design really. Maybe we could persuade Jurgen to allow a piece in RCM&E?

[Paul Luby]

Hi Chris

Have you got an internet link for that Spektrum A Sensor.

Paul

(AKA Veri-Gash)

[Paul Luby]

Hi Chris

Never mind.

Found it.

Paul.

(AKA Veri-Gash)

[Chris Bott - Moderator]

There's some detail of the protocol on the pages for his early version **LINK**

[Paul Luby]

Aw crap

50 today.

Paul

(AKA Veri-Gash)

[Phil Green]

Happy Birthday Paul !!!

[Chris Bott - Moderator]

Happy Birthday from me too.

[David Tweddle]

Well I'm almost 8, . Happy Birthday Paul, also my Mums... or would have been, got a bottle of red for her so I guess I will drink it in her honour Cheers.

David, Year of the Monkey

[Keiran Arnold]

You're an old git now Paul ;0)

Edited By Keiran Arnold on 31/05/2014 17:18:02

[Phil Green]

Where did the pow() value of 0.1 to 1.99 come from Paul?    ie map pot to 1-199, div 100, then pow(result) ?
Must admit, maths isnt my strong point!

My code's a mongrel of bits of RCG stuff, condensed bits of yours, bits of mine, and bits of my experiments.
Heres my expo on ch1 and 2, I dont use a sign variable but I did pinch your pow() values: (ta!)

This is expo without a switch - just set it to zero (ie  '1' ) on the pot if its not required.

The 'map()' function makes me smile every time I use it. So easy compared to pic assembly!!!
Same with pow() I've ordered some tiny itsy bitsy ones to play with next!

So how did the birthday celebrations go?
Cheers
Phil

Edited By Phil Green on 03/06/2014 00:37:35

[Chris Bott - Moderator]

Well you've completely lost me now chaps, but, I've proved that you don't have to fully understand this stuff to make good use of it.

News is that the Waltron Tx was "maidened" yesterday. After a range test where I gave up at 45 paces, it performed flawlessly as I had three great flights with the Mini Tyro.

I'll have to get used to checking mechanical trims again, and I may have to take my thumbs to the gym so I can move those sticks against such strong springs, but it was great to be making use of the old Tx again after all these years.

Many thanks

[Paul Luby]

Hi Chris

Great isn't it, having your old Waltron down the field again. Hope it caused a few raised eyebrows and questions.

You can connect one of these little fellas up to the FrSky DIY module and get telemetry for pennies.

http://www.t9hobbysport.com/frsky-lcdfld02-lite-screen

That'll cause a stir Chris, a Waltron with telemetry, whooooooooh.

Phil

The Pow() value of 1 to 199 came from a little number crunching. If you divide 1 to 199 by 100 then you end up with 0.01 to 1.99. Any power between 0 and 1 gives you negative expo, ie a curve that gives more servo movement around centre stick and less at the extremes of stick travel whereas a power between 1 and 1.99 gives the expo we're all used to, a curve that gives you less servo movement around centre stick and more at extremes of stick travel. I worked the algorithm out so that with zero trim input, any expo selected gives the same amount of servo travel at full stick throw. I didn't want to use a bigger number than 1.99 because it was getting near to the servo end stops if 1.99 expo selected with full trim. Of course I could put a travel limiting function in the code, but

Hope that answers the powers question.

I've looked at my channel 2 code and it looks and performs ok, even looked at the picture of the code I posted and it looks ok.

Here's the pictures of the Futaba M6 stripboard modified with the pots and switches for the rates, expo and reversing (not on throttle, I know it says throttle on the sticker but that has to be redone) as well as the instruction sheet I've put on the inside of the back case. Centre of expo pot travel gives an expo power of 1 which is normal and centre of the rate pot travel gives 50% rate.

 

 

 

That's definitely as far as I can take the Futaba M6 I think. It now performs as well as a well expensive set of gear but looks way cooler.

If you want me to post the code, let me know.

Enjoy

Paul

(AKA Veri-Gash)

Edited By Paul Luby on 02/06/2014 21:36:38

[Phil Green]

Sorry I misread something there

I'm rubbish at maths but I thought the expo routines would affect the servo throw, I'm being thick I know but I cant see how the throw remains constant over the range of the expo pot.
Say for example you've got expo set to 2, so that gives the channel value to the power of 2

Aha I see its just clicked ! After dividing the channels by 500, their values are always between 0 > 1. One to any power is one so the max throw remains the same. Senior moment, sorry!

I'll never use negative expo (except in the Futaba sense, which is positive!) so I'm just going to have it unswitched & configured so the pot fully anti-clockwise is '1' ie no expo, turning clockwise to gradually introduce expo.

On a completely different tack (off-topic alert!) are you a member of the Selby club Paul? or any other nearby club? I might be looking around for a club with a thinner rule book than ours

Cheers

Phil

 

 

Edited By Phil Green on 02/06/2014 21:44:44

[Paul Luby]

Hi Phil

Glad you got the math thing, bit of a math junky unfortunately.

I have a linear brain so don't use expo on my fixed wing stuff. However, on my indoor helis I use the negative expo on trainer helis to make them twitchy before you move onto 3D helis which are twitchy by nature.

Come to think of it, I'll never use the negative expo on the Futaba M6 so I'll probably turn it off as well. Got sucked into the math fun there didn't I.

As for the Selby club, yep I'm a member, we like to keep it small and informal so by all means let me know if you want to pop along and I'll introduce you. We aren't a club for stacks of rules and are small enough to run it as a bunch of mates. Cause we keep the numbers down there's never any of that clique stuff you get at larger clubs.

Paul

(AKA Veri-Gash)

Edited By Paul Luby on 02/06/2014 21:59:57