LowPowerLab Forum

Hardware support => General topics => Topic started by: rjkirk on August 03, 2013, 01:22:43 AM

Title: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: rjkirk on August 03, 2013, 01:22:43 AM
Hi Felix,
You've done a very nice job on your board projects, products, website, forum, and YouTube videos. The website is nice and clean...one word...WORDPRESS...

I'm into RC airplanes and I stumbled across one of your YouTube Videos while researching Hope RF modules. Your Moteino is very compelling. This concept of Hope RF module married to the underside of an ATMega boards is now very hot in the Remote Control Airplane and Helicopter arena. Especially with all the interest in drones and FPV flying. There's lots of products hitting the market.

Your product would provide a great base for both the Ground Radio as well as the airplane Receiver. Two way communications provides a telemetry downlink, as well as an I2c bus on the receiver to interface with remote boards containing GPS, Gyro, Barometer, etc.
Here's a video of a radio transmitting to a receiver which is driving a servo, as well as sending back gyro(tilting) telemetry.
http://www.youtube.com/watch?feature=player_embedded&v=PqsIxCU548o#at=38
 
Most of today's products are using the 433Mhz RFM22B and RFM22BP modules(SI4432 chip) which I know are outdated. As you know, the latest chip is the SI4464 used on the HopeRF RFM69HW module(replacement for RFM22b). I don't know if HopeRF is producing a large power module like the RFM22BP yet.

I also like the fact that you are going to offer the ATMega32u4. Adafruit has one as well as the Teensy 2.0. Of course, neither board combines a radio.

When you marry the ATMega32u4 (or better yet the AT90USB) and the RFM69HW module... then I'm a buyer.

What size boards can you design with your Eagle Software?
Do you have the free version of Eagle that limits your size?
How is it dealing with Hackvana.com for your boards?
How do the bare PCB prices stack up to ITEAD...etc?

I watched a YOUTUBE video of the owner of Hackvana:
He wants to become a DIY kit sourcer. It makes sense...and the RC community would eat up radio/receiver kits like popcorn...nobody does it. Surface mount is fine if you keep the component size to 1206 or 0805. Dave's EVV blog has 3 great YouTube soldering tutorials including SMT soldering.

Here are some links to show what is available commercially in the RC World:
http://www.hobbyking.com/hobbyking/store/__27096__OrangeRx_Open_LRS_433MHz_9Ch_Receiver.html
http://www.hobbyking.com/hobbyking/store/__37329__FPV_433Mhz_Radio_Telemetry_Kit_100mW.html
http://www.flytron.com/openlrs/146-openlrs-receiver-v2.html
The beauty of this whole thing is that all the hardware makers have to do is provide the hardware. Software is all open source. The Hobby King guys simply cloned the open source hardware designed by Flytron. And both hardware makers rely on the open source software guys for firmware. The open source software community can easily take their software and make it run on a Moteino board.

There's lots of opportunity...especially with these new modules. Whoever comes out first with the new modules will have a jump on the competition. The new HopeRF module footprints are not backwards compatible with the old modules. The firmware will need rewritten as the internal registers seem to be very different.
 
I see some new DATA/I0 Pins that are not present in the old chips. I notice you leave the Data/IO pins unconnected...which is just as well, as the "FIFO Packet Handler" is the only way to go.
Title: Re: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: Felix on August 03, 2013, 11:26:37 AM
Hey rjkirk,
Thanks for the feedback and extensive writeup!
Yes my intent as stated many times before was to make a radio enabled arduino that would be very compact yet still have all the functionality/pins exposed for other makers to include in their project. Both R2 (RFM12B) and R3 (RFM69W/HW) have their target markets and applications. The RFM69 comes in a high power variant which has the same pinout (RFMHW). This will be available in my store very soon, it has 20dBm output vs 13dBm of the RFM69W. So it will have that extra power boost for longer range, at the expense of more TX current of course.

I believe the RFM69HW would be the equivalent of RFM22B/BP, not really sure, but believe me ... with the HW you will run out of space in open air before you can run out of range on a drone (see the RFM69 intro video i made).

I would love Moteino to get into drone projects, I just don't have the time to spend to market it and such. I am 1 man operation still, a side job, not a main job. So I am relying on the community to spread the word and slowly it picks up if it's a quality product which I am hearing a lot that it is. Lots of people praise Moteino for what it is, which is both exciting and humbling for me.

On the Atmega32u4 front ... that is an odd cat and it was hard to make by hand. So not sure how much I would be able to mass produce it. I have a much easier time to produce R2/R3 than Leo.

I use the free Eagle at the moment, all my projects are small boards, which is something I'm always looking for because small PCB = low cost to make.
Hackvana is great, I mean you may not get the top quality you get at a US reputable maker, but you also won't pay outrageous prices. I think the prices are very competitive and compelling. I have ditched Itead/SeedStudio for hackvana, hopefully Mitch will keep up and improve his service. I think he has a lot to gain given so many startups and individual makers need a consistent reliably and good quality PCB maker, which in my opinion was an issue before hackvana. I tried many services, 4pcb, batchpcb, itead, seed, oshpark....

OSHPark is good too, but it's a prototyping service, so it's good for checking your boards, mass producing becomes VERY expensive at OSHPark unfortunately, and you're stuck with purple (which I heart he will expand to more colors ..) and ENIG finish which I personally dislike and really prefer HASL over ENIG (yeah .. gold looks nice on purple but it's a huge myth floating around that it's easier to solder by hand).

I cannot provide too much software because there's a million cool projects out there and I'm trying to stay focused. What I can do is provide hardware. I hope the RFM69 will take off and become a good alternative to RFM12B. The lib i released I believe is the first for the Arduino platform so I hope/expect others to test it and improve it and build stacks on top of it for drones and other applications.

There's several DIO (interrupts) on the RFM69, of which I only use the single one that's relevant to packet handling. That's cause I'm not really interested in continuous and other modes this module offers. Those modes might be ok for video or streaming or things like that, but they also don't have the cool features of packet handling which I really want. Packet handling opens the door for a lot of things I don't want to code for in firmware, which means more reliability.

I hope I addressed all your questions, but the discussion is open ended, thanks again :)
Title: Re: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: Felix on August 03, 2013, 11:29:41 AM
Oh by the way, you said I should come up with a kit, would sell like popcorn :)

I'm not really into drones even though I always read and watch videos about drones and X-copters.
Any more specifics?
To me it almost sounds like the Moteino could just be plug and play (abstracting firmware which I think is easily adaptable), and perhaps a matter of a few people to pioneer some drones with Moteino.

Well.. knowing that some copters are now very advanced, have arm MCUs ... maybe i'm missing something, but any advice appreciated!
Title: Re: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: rjkirk on August 03, 2013, 03:32:11 PM
Hi Felix,

You said that:
"On the Atmega32u4 front ... that is an odd cat and it was hard to make by hand. So not sure how much I would be able to mass produce it. I have a much easier time to produce R2/R3 than Leo."

Why is that? I thought USB would be nice to have...but it's not really a show stopper.

I'm single and retired, so I have the luxury of dabbling into these things all day. The project I'm brainstorming is single footprint radio/receiver board that will work as a general purpose ground radio as well as an in air receiver. And it's not strictly for drones. The drone piece only comes into play if you add I2C accessory boards containing those necessary chips(gyros, barometer, compass, GPS). So this is a general purpose board that can be enhanced. It's also two-way communications.

Most of today's RC radios evolved in the one-way mindset. So most drone systems are actually adding a totally separate radio system for the downlink. This is crazy, especially when you consider today's capabilities of two way automated FIFO transmission.

But I'm getting ahead of myself, as I don't know how familiar you are with RC Radios and Receivers.

Essentially the RC Radio runs off a general purpose chip like an Atmega64. That chip monitors two joysticks which consist of 2 potentiometers each. So the two joysticks tie up 4 analog pins. Also from an analog standpoint, some radios have one or more a rotary dial potentiometers or slider potentiometers. So that's a minimum of 5 analog pins used. Everything else is digital. You have two types of switches. Toggle switches can be read at any time, and each toggle takes up a digital pin. Momentary tactile switches also take up a digital pin, and must be constantly monitored in a round-robin fashion to catch the switch push. Some radios have LCD displays, some have a buzzer or speaker(pwm), some have SD cards. The final piece of the puzzle is controlling Radio's RF chip. I won't go into great detail here. Suffice it to say that for the sake of backwards compatibility, this linkage is also caught in the dark ages.

Up in the air the RC Receiver is usually a smaller CPU chip that controls the Air RF chip and feeds the 4 to 12 servo channels with the proper PWM square wave signals to drive the servos. This receiver is usually powered by 5 volts that is derived from the ESC (electronic speed controller) which is turn is powered by a lipo battery. Historically there is no transmission in the reverse direction.

That's it in a nutshell.

Rather than make this post any longer, I'm going to start digging through your Git files. I want to look at the hardware schematics. Also I need to start reading the SI Labs and HopeRF datasheet of the new RFM69HW module. I believe the RFM69W and RFM69HW are compatible from a footprint and programming prospective...aren't they?

Any firmware programming and proof of concept can be done on the Moteino with the RFM69HW chip. And I totally agree about hardware vs software. Software should be left totally in the hands of the community. Anything past the bootloader and a simple communications test program should be an open source software responsibility.

Just as an FYI, I use Studio 6 rather than Arduino. I'm also a proponent of putting a 6-pin ISP on the board in addition to the pins for TX/RX bootloaders. I've seen too many forum posts where bootloaders get trashed, and need to be re-flashed....or cases where CPU fuses need to be adjusted.

Also my project would probably require a 64 pin chip along the lines of the 64 pin versions of the ATMega128 or ATMega256. Also, yes...lots of systems are moving to Arm Chips. It's totally unnecessary in my mind. The RC Radio transmits data packets at a rate of 50 packets per second. So the radio/receiver combo repeats the same set of tasks over and over again....one pass every 1/50 of a second. At 16 MHz, 1/50 of a second is 320000 CPU clock cycles. I think the task list is simple enough that you will spend lots of time sitting in a "wait loop" at the end of each pass.

The bottleneck will always be the "over the air baud rate". I'd like to shoot for a data packet size of 32 bytes at an air baud rate of 56K to 60Kbaud.

I'll keep watch for any of your boards with the RFM69HW module.

Title: Re: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: Felix on August 03, 2013, 11:28:22 PM
Ok let me try to answer your questions.

About Atmega32u4 ... first of all it's hard to make, and test. The board is small and I had to use the VQFN package with 0.5mm pad pith. Yeah .. I don't have a PNP machine, real pain to assemble by hand, and time consuming to bootstrap and test. Perhaps the TQFP version would be easier to asseble but might enlarge the board which was something I wanted to avoid. My intent was to satisfy the needs of some users that asked for a Leonardo type Moteino, but it didn't pick up so much steam as I had hoped. Also the parts are more expensive so I have to invest more $$ to keep parts in stock. My desire was to keep Moteino pretty cheap so you can leave it in all your projects, Leo doesn't really fit that idea. Those are the main primary reasons, and I'll leave it at that.

Would definitely need to document myself a little more on RC circuits. Unfortunately I have so little time to spread in other directions. I hope some community people will play with these and submit suggestions if they have working solutions. I just don't have the time to invest in coming up with new solutions for markets where I don't have much vested expertise.

More capable Moteinos can be made around other chips of course. It's just that each design requires a lot of time, and also money in making prototypes, testing. The question is, if I am not primarily interested in more specialized chips, who will?

I wanted to keep Moteino small, any eliminate redundancy, which the ISP header is. All ISP pins are broken out on breadboard friendly headers, so they can be reflashed any time either using a breadboard or using a jig similar to what I use to flash them (http://lowpowerlab.com/blog/2013/06/27/illustrated-guide-to-making-simple-jigs-for-programming-and-testing/). I posted the bootloaders I use in my github repository (https://github.com/lowpowerlab).

I just made the RFM69HW available, both as standalone and bundled with Moteino. It has the exact same layout/footprint. My lib is hardwired at 55.5kbps, but can be changed in the initialize function. 50 short messages per second would probably be the top limit as I see it now, unless you raise the bitrate to 100kbps or more. They are capable of up to 300kbps I think but I suspect the range would suffer significantly at such high rates.
Title: Re: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: rjkirk on August 04, 2013, 12:41:31 AM
Super!!!
If the RFM69HW is available attached to the Moteino Board, I'll buy two.
Problem is, I don't see it as a bundled product on your store.
As soon as it is, I'll pull the trigger...along with an FTDI board.

One last question, then I'll quit pestering you.
I'm going to download the Arduino IDE. Should I just get the most current version?

Now it's time to start printing out and reading the datasheets.
I'm excited.....

 
Title: Re: Come up with a kit for the Remote Control Airplane/Copter Hobby
Post by: rjkirk on August 04, 2013, 02:39:51 AM
Opps Felix...my bad...

I looked a little closer and noticed the RFM69HW is in a dropdown for the product as an option.
Now I need to decide if I want to add the extra EEPROM, so I have to do some research.

You probably should tweak the product title to say Moteino R3 (including RFM69W or RFM69HW).