Developing a universal module

More
21 Mar 2013 15:21 #7961 by PhracturedBlue
Developing a universal module was created by PhracturedBlue
I've brought this topic up several times over the last year, so I thought I'd start a thread dedicated to it in the hope that it'll lead to forward progress.

I want to design a universal module that can be used to run all protocols that Deviation can support. I'd like it to be low-cost and a drop-in replacement for the CYRF module on a Devo radio.

The goal would be to design a board tht can incorporate multiple transceiver chips with a single power-amplifier and single antenna. And to keep the overall production cost as low as possible.

The CYRF module includes the following pins:
5.5V supply
GND
SCK
MISO
MOSI
CSn
Reset

If we also use the Debug port, there are 2 additional IO lines available as well an a 3.3V supply.

Here's my thoughts on a parts list (prices are estimated at low quantities):
CC2500 QFN-20 package ~$3.35
CYRF6936 QFN-40 package ~$3.20
nrf24L01 QFN-20 package (no US supplier)
A7105 QFN-20 package $2.40
8:1 Switch for GPIO (something like a MAX4581) ~$2.00
3.3V converter (various options) ~$2.00
RF Switch. Something like a BGSF18DM20 looks pretty good, but I'm not sure about the impedence matching and using it as the input to the PA/LNA. ~$1.50 There seem to be pother options at the SP4T and SP6T levels as well
Power Amp ~20dBm. Several options around ~$1.00
A bunch of R/C/L components for impedance matching and biasing
Possible need for a small uC to to control the switches (AVR/PIC) ~$2.00

Virtually all of the above parts are SMT and cannot be reasonably hand soldered without special equpment. So this is not a 'kit' project.

I think I'd develop the above as a module that can directly replace the Devo module.

In order to make this module more attractive, it would be valuable to develop a PPM->SPI module around an STM32 MCU. This would host the Deviation protocol stack as well as some interface to control the protocol (perhaps bluetooth). Together with the above it could be used with any transmitter to provide all protocols. That is yet another project on top of the 1st.

I'll need to look at software options for development. While I ue a bunch of EDA tools at work, they are all proprietary and not designed for board-level development. I only do this kind of thing every couple years so I don't keep up with the current software much. In the past I used Eagle which works fine for small boards, but has the most atrocious user interface. If anyone knows what state-of-the-art in freely available circuit design and board layout software is I'd be interested.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 17:28 #7965 by sbstnp
Replied by sbstnp on topic Developing a universal module
Take a look at:

www.zenitpcb.com/
sourceforge.net/apps/mediawiki/tinycad/i....php?title=Main_Page
www.freepcb.com/


I have no idea if they are suited for the task or complete/good they are.

Devo 10 + 4in1
Spektrum Dx9
FrSky Taranis + TBS Crossfire

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 19:13 #7968 by Wene001
Replied by Wene001 on topic Developing a universal module
3 boards for me please... B)

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 19:27 #7969 by PhracturedBlue
Replied by PhracturedBlue on topic Developing a universal module
Thanks, I'm aware there are lots of free solutions out there. I'm interested if anyone has actual experience with any of them. Spending effort to evaluate each of the options is not on my 'fun' list :)

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 19:31 #7970 by nerwin
Replied by nerwin on topic Developing a universal module
How about something akin to a kickstarter. If a bunch of us all pre-pay for modules, at a predefined and slightly inflated price, we can essentially fund the outsourcing of the module creation, and let the talented team of coders here work on the software side of things.

I know I would be willing, as I love the work done so far (I have been a long time lurker, and have owned a 8s, and now a 12s only because of the existence of this firmware).

PB, I know your anti dontation, but this is more like spending others money to achieve the desired goal.

Then you wouldnt have to bother learning any layout software, just defining the standards/requirements, and finding a offshore design house to create. I know there are a few native chinese speaking folks who could likely assist in finding a manufacturer/designer offshore.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 20:32 #7971 by PhracturedBlue
Replied by PhracturedBlue on topic Developing a universal module
It is more than a question of funding. There are potentially legal ramifications for monetizing an endeavor like this in the US. If I ever complete the process, it is very likely I would not be willing to actually be involved in the manufacture or distribution. There are also FCC regulations that come into play in the US and elsewhere.

So the best case is probably that I develop a BOM and a board design and make that freely available. That would leave it up to some entrepreneur to actually manufacture and sell the boards (as well as reap any profit). Becasue of the nature of the design, it is unlikely that this can be a DIY type of thing.

The alternative is to design a board that does not contain any RF, but instead just has solder pads for all known RF modules of interest. This board could be manufactured to use only through-hole components and could easily be sold as a DIY kit. The downside is that you will have a bunch of antennas to deal with, and overall a lot more expense (since modules tend to cost $10 to $15 each). The overall platform would likely be a lot larger as well.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 20:43 #7974 by nerwin
Replied by nerwin on topic Developing a universal module
What about the guy who makes the anytx module, does anyone know who is and speak his native tongue?

Looks like he could assist from layout right through to production.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 22:10 - 21 Mar 2013 22:11 #7976 by blackmoon
Replied by blackmoon on topic Developing a universal module
I'm starting learning Kicad, it's from my point of view a lot more friendly than Eagle..

You could contact th9xer (th9xeratgmail.com) from the openRCforums (ex-er9x forum), he has already developed a ppm2spi converter that runs your protocol stack on AVR, currently Flysky and Hubsan.

openrcforums.com/forum/viewtopic.php?f=5...f720d99d90741b5204f2

code.google.com/p/rc-ppm-2-spi/

He also uses Kicad.
Last edit: 21 Mar 2013 22:11 by blackmoon.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 23:10 #7978 by RoGuE_StreaK
Replied by RoGuE_StreaK on topic Developing a universal module
I actually find Eagle to be quite good now that I've become (semi) used to it; here's my latest board that arrived the other day, Seeed will make these sub-50mmx50mm boards 10x for $10
File Attachment:


While I like the idea of the multi-module board, I think there would be real issues if anyone wanted to fly with it somewhere that they are strict on compliance; that said, are any of these separate modules certified for use? I know some clubs won't let you fly with non-tick equipment, but I guess running deviation probably already makes it non-tick anyway?

Not that it matters to me, as I don't fly in any of those places.

RE: the module-connecting board, would need to get good close shots of each of the devo TX boards; I know the 7E definitely doesn't have the header that rbe2012 mentioned in the skyartec thread, but also haven't looked to see if the pads that are there match those of the header.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 23:29 #7979 by PhracturedBlue
Replied by PhracturedBlue on topic Developing a universal module
That anytx module is pretty cool. That is exactly the kind of think I'd want to do. Looks like he's using my protocol stack (at least there's a mention of deviationTx in there). Add an RF switch, a couple more modules and and spi bypass, and that is what I described above.

Please Log in or Create an account to join the conversation.

More
21 Mar 2013 23:34 #7980 by PhracturedBlue
Replied by PhracturedBlue on topic Developing a universal module
The 7e has the proper pinout but is not pin compatible with the other devo radios.

The only way to make this compliant would be FCC testing, and even then that probably is not be since we could easily tromp on other Tx signal even if the power output is compliant.

All of the boards I have use either 0.1" headers or the 2mm headers and I have the pinouts for several of them. so designing a 'connecting board' would be pretty straight forward

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 00:55 #7981 by PhracturedBlue
Replied by PhracturedBlue on topic Developing a universal module
Rogue, what do you do as far as solder masks go for your boards. do you make them yorself, order them, or just apply by hand?

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 00:58 #7983 by andri65
Replied by andri65 on topic Developing a universal module
Hi, it's a wonderful project and the Devo really will be the best radio on the market, I think that the good way is an AVR or an MCU that can generate all protocols switchable from software. I will be very happy, if possible, to help you. Very good job, my best congratulations.
I think also that is possible to install on Devo 12/12s a card reader, maybe need a firmware update!?! :unsure: :blink:

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 01:17 - 22 Mar 2013 01:17 #7985 by RoGuE_StreaK
Replied by RoGuE_StreaK on topic Developing a universal module

PhracturedBlue wrote: Rogue, what do you do as far as solder masks go for your boards. do you make them yorself, order them, or just apply by hand?

Just to clarify, I assume you mean solder stencil? (solder mask is supplied on the board, which does help the soldering process)

Currently checking a few different options, I'd previously made my own from etched brass but the toner transfer method is causing some issues at the moment; toner has apparently changed in the cheapie laser printers, not really usable for this anymore, have just printed some stuff using work's Xerox beast and hope to test it over the weekend; apparently has lower fuse temps.

I'm contacting a few cheap laser-cut stencil companies, but not having too much luck getting replies. There are a few in the $30ish range for mylar, even a Chinese one on ebay doing laser-cut stainless steel for about that price. As these are unproven prototypes, I'm trying to do my testing on the cheap until I know the board actually does what it's supposed to.
There are a few cheap ones in the US, but they charge ridiculous international shipping.

I won't be doing hand-application as there are a couple of really fine pitch no-lead components there, and I want to minimise my chances of getting bridging under those suckers.
Last edit: 22 Mar 2013 01:17 by RoGuE_StreaK.

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 08:44 #7992 by kille
Replied by kille on topic Developing a universal module
I often order prototypes at www.pcb-pool.com/ . They are not cheap, but they offer free laser stencils and for this pagage is a laser stencil very usefull.
This is a very cool project! I also use Eagle, so I can help you by layouting or so on!

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 08:46 #7993 by kille
Replied by kille on topic Developing a universal module
Rogue is the round PCB a board for a Cree MTG?

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 11:17 - 22 Mar 2013 11:21 #7998 by RoGuE_StreaK
Replied by RoGuE_StreaK on topic Developing a universal module

kille wrote: Rogue is the round PCB a board for a Cree MTG?

Close, XM-L (or XM-L2). Just added it for the hell of it, experimenting to see if a via-riddled 0.6mm FR4 with max-copper planes works OK for heat transfer. But ideally going for direct to copper soldering. No existing MCPCB suits my purposes, so seeing what my options are for DIY.
Seeed will do weird shapes like this, including internal shape milling.

Personally, after some prototype DIY runs to check everything works properly, I say get one of the small pro companies to do the population, unless you are only making a couple a week.

PS. I don't want to give the impression that I'm a professional at this, it's very much a side hobby, just with a fair whack of googling behind it.
Last edit: 22 Mar 2013 11:21 by RoGuE_StreaK.

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 11:23 #7999 by vlad_vy
Replied by vlad_vy on topic Developing a universal module
What about ML2724 for Futaba FASST and CC2520 for JR DMSS?

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 14:49 #8003 by PhracturedBlue
Replied by PhracturedBlue on topic Developing a universal module
I don't know how to source the ml2724. If you know of anywhere that can source them in quantities of 10 or less, I'm all ears. Adding the CC2520 is a good idea. I really wish it were possible to find a 2.4GHz transceiver with a flexible modulator. All of these chips are just barely outside the specs of each other such that they are incompatible.

Please Log in or Create an account to join the conversation.

More
22 Mar 2013 19:31 #8008 by SadSack

Please Log in or Create an account to join the conversation.

Time to create page: 0.100 seconds
Powered by Kunena Forum