drbutton

Hello everyone,
I have a Hitec Laser 6 transmitter, and I need to add an expo function to it for my foray into 3D flying. Obviously, this will require some (possibly heavy duty) modding of the Tx, but if it'll save me some $$, I'm game. Has anyone ever modified a transmitter to do this? Any advice/schematics/websites would be greatly appreciated. I would like to add adjustable expo to all 4 stick channels with independent adjustment. Thanks!

Phil Cole

Two possible ways to do this:

1. Read up on exponential transfer functions in an op amp text book or application manual. Measure the voltage range going into your encoder chip for each channel, then design your expo circuits to give the response you want. The expo circuits would go between the input pots and the encoder somewhere. Adjust the expo and travel would be done with pots. You could easily add dual rate.

2. Use a small micro controller, (like a PIC). You can read the stick voltage with the controller and apply an exponentiated voltage to the encoder, or read the pulses coming from the encoder and alter their widths before passing them to the modulator. Adjusting the settings could be done by reading analog pot inputs, or by using a small display and a couple of switches.

The components for either will cost about the same, but the microcontroller approach will also require some sort of development system.

Either way, if you know nothing about electronics you have a big task ahead of you unless you find someone who has done exactly what you want to a similar transmitter.

Most people will tell you to sell what you have and buy a transmitter that has what you want in it, unless you want to do it for the sake of doing it.

Rick Lindsey

I've thought about doing something similar, except for some reason i was thinking of doing it at the receiver end... have a wee programmable microcontroller with as many pwm generators as I can find (or perhaps an fpga... that way I can do as many pwm generators as I want) and plug the rx outputs that you want modified into your little custom mixer board, and then have a handful of outputs that are customizable to your needs of the moment, based on the inputs from the rx and your algorithm.

drbutton

Rick, that might be a better idea. Have you come up with a design, or found a schematic?

smokingcrater

more complicated than a PIC, but look up the BasicX. granted, now your plane is going to be carrying around an entire operating system on board, but LOTS of possiblities! if you want some hints as to how to do it, check out kd7lmo's site about his near space return vehicle. (he uses rabbit processors on his, much better than a BasicC, and you program everything in C)


http://www.kd7lmo.net/nsrv.html


Ryan
kc0lmo

Rick Lindsey

Nah, I haven't put the effort into it yet to draw up a schematic. I'd first look into the various FPGA and microcontroller eval boards available, though I'm thinking they will likely not be small enough for our purposes. Maybe one of the PIC basic stamps?

-Rick

Lynx

FPGA is too expensive and not worth investigating, trust me I already have. Microcontrollers will do that kind of stuff in their sleep if you're get good at programming in it's assembly language. Do NOT use a basic stamp, they are COMPLETE junk compared to a full featured micro, and cost 10 times as much. The chip designs are robust enough you don't even need a board, you can run TTL compatible logic directly from the chip with no external components. I'm actually in the process of desiging what you're talking about using mixers built into a cable. The chips are smaller than the width of a servo cable itself so my hopes are that it'll be no more than a bump of heat shrink. No xmitter or receiver modifcation needed, the mixers attach directly to the servo's, and also act as line buffers/boosters and can even increase the refresh rate a servo see's (increasing torque) After that I'm planning on making something akin to a matchbox, only smaller and with higher functionality using a small PCB. A simple inline cable mixer will cost less than 10 dollars to build. I've spent the greater part of the last year researching this project and I'm still determining weather or not I'm going to release it to the public as plans or atempt to sell it commercially. I'll let you know once I get the first calibrated prototypes out. I have examined EVERY possible avenue of modifying an RC setup to do this and know my ideas will work once they're material. All I need right now is time to develop the code, I have the entire hardware setup in place already. If anyone has questions you can PM about them. And I'm most definitly willing to take donations for development if anyone is interested =) It's relativly cheap to develop things the way I am, but it wouldn't hurt to have some old servo's to test on. I'm currently using an AM Attack 4 with a 4 channel AM reciever. One of my side projects is to covert that to a custom protocol PCM setup, but I won't have a scope to reverse engineer the PPM signal source / destination for a few more weeks. Micro's have monstrously huge potential in the RC market, can't wait till companies wake up and smell the coffee.

Rick Lindsey

I like the massively parallel nature of FPGAs (plus I design FPGA logic for a living so I've got free access to the development tools) but they don't usually come in as handy packages as micro's do. A little SOIC-16 micro would be way easier to deal with than the QFP or BGA packages typically found in FPGAs. For our purposes, I'm sure you're right about micro's being more appropriate. They're really not that expensive though (~$30 for Xilinx Spartan devices that will have more resources than you'll ever use on PWM generators).

Eventually I want to get my hands on some piezo-electric gyro's and rig them up with some ADC's so i can have a constant attitude readout and use a micro to control not just mixing, but to actually fly the airplane. It should be cake to set up simple things like maintaining a particular bank angle or a particular angle-of-attack (or both simultaneously to turn the plane), maybe get a little electronic compass too so you could have it turn to a particular heading, and fly at a particular throttle for a given duration before taking the next step of the program. At the ultimate end it would have a little GPS receiver and sonic range-finder for altitude detection once you're within a few hundred feet of the ground. Then you should be able to calibrate the GPS coordinates of the ends of your runway and teach the plane to land itself!

-Rick

Lynx

I'm currently designing the inline mixers for 8pin soic. Commericial GPS doesn't have the update rate to keep an airplane in the air. You need inertial navigation AND high rate GPS, both of which are technically speaking illegal for applications such as RC craft. Developing inertial navigation period is SEVERELY limited by the governement, and higher rate GPS isn't possible at a hobby level; and is just as illegal even if you could =) Just try to sell it. Even the FMA and Futaba systems that are out there for horizon stabilization are technically illegal if you hook them up into any other type of subsystem. They're considered 'dumb' systems, and not acurate enough for the governement to worry about. More precise commercial systems aren't legal. They're possible just not legally practical. Even camera stabilization systems are limited to some degree in the way they can be implemented.

Rick Lindsey

Cool, mind if I ask who's micros you use? I just looked atmel's ATTiny avr series micros -- wow, $3 bucks in a 20-pin SOIC, with a built in 10bit ADC to boot! An 8-pin device didn't seem like enough to me, since you'd lose at least 2 of those to power and ground, which leaves only 3 inputs and 3 outputs. I'd like to find something that would accept, say, a 5-channel input, and control up to a half-dozen or more servos on those inputs (that way you could control two aileron servos, plus flaps and your typical R/E/T stuff to boot). I'm thinking the 20-pin job might just do that for me. If I could get the piezo gyro hooked up to the ADC on the micro then I could do fun stuff like "dummy mode" that would interpret any amount of aileron input as "bank left/right" and use the gyro to maintain a reasonable bank angle and maybe even keep the nose up automagically. Of course if you knew how to fly, the dummy mode would be stupid, but if you just wanted to let a random spectator have the thrill of being in control while keeping the danger to a minimum, it might be neat .

Where can one find out the specifics of the laws regarding precision navigation units? I'm not interested in selling them, just building one for my own enjoyment.

Of course, first I've got to find a source for piezo-electric gyros that doesn't have a minimum buy of 100 units *grumble*. I'd pay $20 each for the gyros (that's what the murata ones cost) but I'm not going to plunk down two grand just so i can play around with them! I've thought about using pairs of accelerometers instead, but that would be a calibration nightmare.

Lynx

Don't use piezo gyro's they're too susceptable to drift. Look into Memsic thermal accelerometers. They have digital two axis accelerometer that don't require calibration, have very low noise levels, are smaller than an 8pin soic form factor, and have the added benefit of also providing the tempature as an extra digital output. All with high linearity, low noise and up to 200hz refresh rates, (better than the 50hz a normal hobby piezo gyro puts out) And like I said, building any kind of inertial navigation system is illegal period. The specifics aren't easy to find.
Parrallex sells a memsic pre-soldered to a PCB with headers that makes it about the same size as a 6 pin dip would be. It's expensive as hell, (25+ bucks) but you can buy em one at a time if you want. The hard part is calculating rotational velocities from two acceleration vectors. That math is NOT easy or cheap to do on a micro. Using the tempature output and an ADC pin connected to the VCC of the micro you can precision calibrate the micro for any working environment that it won't break in =>) I may play with those thing once I get the rest of this stuff done but my plates full for the next 6 months as it is. Exactly what I'm using and how I'm using it I won't say cause I'd like to at least try to have this patented =) Programming them is the hard part anyways, you're talking some serious timeing issues dealing with multiple PWM signals, (especially with a PCM system where they may occur simultaniously) And if you're using digital servo's if you can't precision code your program to work EXACTLY right you'll have servo jitter because interupts occur that extend the pulse in accuracys past the deadband of the servo. Even with fast chips this is a problem.

rik756

My Feedback: (12)

I'm thinking;

Time+Labor+Parts+Crash Tests = More Expensive than a new (or used) radio with Expo already programmed....

But then I know nothing about programming or designing that stuff either. Sounds fun for a *what they hay project* but as a money saver, {IMHO} you'll probably spend more time and money on crashed planes or fried circuits testing it than you would have buying a used radio or even one in need of repair and fixing it. The other benefit to a radio with expo is that you could have *free* expo on all your planes without the inventions lol. Good luck though, if you can do it - You're the man!

ZAGNUT

My Feedback: (1)

ace used to make add on accessory boards for their radios, expo being one of them. Eloy Marez, EloyM on rcu, mentioned in previous threads that he has lots of ace schematics. might be worth checking


dave

Lynx

rik756, The parts are very cheap, even the development environment and all the hardware you need to get started cost under 100 dollars. The circuit itself can consist of nothing more than the micro itself because they're robust enough not to need external circuitry for simple appplications. With enough testing on the ground in air failures should be reduced to the same as what you'd get if you were just using an inline servo reverser. I enjoy doing this stuff in my spare time too so it's not really a project I'm doing cause I need to. For servo ganging I think it would be far cheaper than either the Futaba or JR matchboxes and just as reliable.

rik756

My Feedback: (12)

If it's something you enjoy and something you are good enough to make happen - More Power To Ya Man! I wish I knew that much about stuff like that. Half the time I have to look in the book to remember which slot to plug the left aileron into lol. Guess I'm gun-shy on that sort of stuff most due to lack on knowledge about it. I really hope it works for you. If it does, maybe you could market it and save us all some money :-D I have an XF631 that I won't use on several planes due to the lack of expo....

Rick Lindsey

Oooh, fun. The advantage of a piezo-gyro would be that you don't need to calculate rotational velocities, since it's an angular roll rate sensor rather than linear. If I'm not mistaken, the murata part will run at several kHz, you'd probably need an external ADC set up sampling the bejeezus out of it though, so you can integrate the roll rate into attitude. It's all moot though, since that part is unobtanium! Surely someone else out there sells a roll-rate sensor?

Ah, and that's where the fpga comes in *grin*. It can easily handle as many PWM signals as you have i/o and gate resources for. Duplicate logic!!! Of course, then you've got the packaging problem, ohwell! I was thinking about using a little cpld as a pwm generator, actually, to get that load off the micro so you don't have to worry about interrupts. You might still get a time-lag between stick movement and servo pwm width changing, but at least your pwm wouldn't jitter. This starts to get alot bigger than a little SOIC micro, though. trade-offs galore.

rik756 -- I hear what you're saying. I'm with lynx though, I rather enjoy these kinds of problems. Unfortunately I don't have the funds (or really the time -- a 2 year old daughter at home and a newborn due at the end of this month!) to take it beyond the conceptualization phase. "One of these days..." though, I'll get a board together and buy some parts to play with. Just look for the mad scientist fellow with scary hair out at the flying field with his trainer and a laptop!