You're right, it looks like TI only integrated the CC1101 in this series of parts. Ignore my previous comment about picking the right frequency (RF RX and TX) parts.

I'm throwing together a possible BOM (because I had nothing better to do), and the TPS61120 appears to be a very good regulator for the job, since we need a split supply. We can run the high side of the level shifter (at 5V) from the switching side, and put the micro on the LDO side.

I had previously expressed a concern about core supply voltages, but we can safely operate the micro down to around 2.2V when the radio's running, and I don't think the battery will really get down around 2.5 (minimum drop on the LDO at 200 mA draw (whereas we'll probably take about 20) is 300 mV). Indeed, glancing at the datasheet for one of Sparkfun's Li-Po batteries, it's rated to discharge to 2.75V, giving us some headroom.

..and browsing the regulator's manual, I realize we can just shove the switching side's output into the LDO's input and not worry about battery voltage at all.

I'm filling out a sample cart on TI to put together a prototype of this; thus far I have 2x TPS61120PW and 2x CC430F5135IRGZR. Anything else I should grab before I check out? It looks like TI's BQ24026 is almost a direct substitute for the MAX1555, so I added 2x of that.

Edit: finally found a cheap Li-Po battery source. 200mAh for $2.00 seems nice to me!

my only question is can you throw together a prototype for me? Ill send you the money for parts and time. I just have no means for assembling one here even if I get all the parts >.<

also could i use a nds battery in place of that one in the protoype as i have like 6 of them kicking around lol

also we need to think of how we are programming these things for the prototypes i have a ti usb debugger already for my msp just wondering how you are going to program and debug it.

Well, I went ahead and ordered those two sets of 3 chips, but as you know, they're microscopic chips. I'm going to try dead-bugging 30-gauge wire onto them for breadboarded testing, which will be a fun adventure indeed, and once I have a working point-to-point test running, I'll use DorkBot to design and order a PCB.

Hmm, that's something I didn't think about at all, I'm sad to say. I hope I can take advantage of one of my Stellaris boards or my Arduino boards to somehow program it.

dead-bugging? and fair enough. i should be able to write up a point to point program for you all that needs doing is editing the environment for your chip. BUT there should be examples out there already. ill look into it later tonight.

As for programming a standard jtag interface can be used as well as a few other methods they are in the datasheet

http://www.ti.com/lit/ds/symlink/cc430f5135.pdf <-- thats the data sheet see the bottom of page 8 the method i have uses spybywire

As far as dead-bugging, here's what I'm going to need to do; I'll have to get new thinner solder, probably: http://www.mechatronicblog.com/2012/03/04/tutorial-1-dead-bug-soldering-a-qfndfn/

I see that I can make a JTAG interface out of several things, including an Arduino, which is nice to know.

after some googling its looking like my debugger for my devkit wont work with the cc430 with out a upgrade which stinks. gonna see how much of a pita it is to update and if its possible with out special tools :/

I trust you'll find good stuff for us to use to program the CC430. I'm impressed with the speed with which TI shipped me my samples; now I need to look at the sample implementations that they include in their datasheets to see what ancillary components I'll need in order to make all of this work well together.

Programming it is honestly easier then debugging it debugging it is the problem. As the standard debugger interface is 100$. The cheapest one is 15 and requires a parallel port unless i can find the updated usb debugger for cheap thats the only real solution :/


On second thought it looks like it may work. Ill just have to order a sample and test it my self unless u want to solder one up and ship it to me Kerm. If it does work then this will work to program and debug them but yeah no way to know with out trying....ti has crappy documentation.

If you has a MSP430 launchpad, you can use that for programming/debugging. Just break out Spy-Bi-Wire interface out to a header and plug the launchpad's programming lines into it.

Another option would be full-blown JTAG on a Bus Pirate (needs more
pins though), and either of those devices are very useful to have in your toolkit as well as being fairly inexpensive.

Kerm: I've been creating parts in EAGLE and working on some board layout, so I can be your layout lackey if you like.

Tari: That would be awesome! If you can come up with a good compact layout, I'd be happy to get it fabbed. What if I have a pair of Stellaris boards but no Launchpad; can I still pull a Spy-Bi-Wire out of that? Or should I get myself the MSP430 Launchpad?

getting a launchpad from a friend they are a fully features SBW dev board so grab one


Kerm the suggested setup is on page 88 of this pdf http://www.ti.com/lit/ds/symlink/cc430f5135.pdf

Kerm out of a programmer do you want to have debugging capabilities or just programming? Cause im sure your stellaris will easily program the chip it just wont have a debugging interface that a SBW programmer will give you.

And I see that that page says "For a complete reference design including layout see the CC430 Wireless Development Tools and related
documentation (MSP430 Hardware Tools User's Guide, literature number SLAU278)." It looks like we should decide which frequency band to use, as that dictates the capacitor and inductor and resistor values (see page 89 of the document you linked) that we'll need.

Having both would be ideal. I'm thinking that my Stellaris might offer that SBW interface too, as it include in-circuit debugging.

Im leaning towards the 800+mhz range. If I remember correctly it needs a smaller antenna among other things.


It might I dont know tho I would look at its documentations (your just itching to put that to use arent you lol)

That's true; the higher the frequency, the smaller the wavelength, the smaller the antenna. There are disadvantages as well, though.

I see I have the pins for it broken out; I just need to figure out how to temporarily disconnect the Stellaris's own MCU while I use the interface.

Please fill me in ive been sporadically googling things



good luck!

Ok to recap what we have so far

Major parts

- CC430F5135 - our current choice of RF chip we should spend sometime to look over the others and see what we like most
- bq24026 - replacement for a max155
- TPS61120PW - power regulator for our battery
- Li-Po battery - using this unless we find a way to power the device with out it. Note from Kerm: actually, the 250mAh battery is $1.90. Shipping is like $5, so it's best to get a few at a time to amortize the shipping overhead.

DataSheets

- CC430f5135
- bq24026
- tps61120

Important Links

- Simpliciti Rf stack
- Calcnet White Paper
-Grace GUI for configuring hardware

Comments and Notes in no particular order

- Who ever is coding for this will need a Spy-By-Wire enabled programmer like the msp430 launchpad
- I'm leaning towards the 800+Mhz frequency range. need your guys thoughts.
- Whoever gets designated as the main developer should probably invest in a Packet Sniffer if we end up needing it
- Notes on the timers and how to use them for more then one interval

To Do:

- Get a working chip that we can load with code - Kerm
- Setup SimpliciTI and basic functions for the various development enviroments used - Geekboy
- Work on schematics and PCB design - Kerm / Tari (Sorry I'm not gonna be very helpful here stil learning )
- Code, lots and lots of code - Anyone
- Brainstorm and draw out our network mesh design - Everyone
- Decide how the Memory structure will be so we can allow control over what the device acts like in the mesh network. - Everyone

Im sure there is more than this its just what i have typed up as of now ill edit it as we go and occasionally repost it.

I'd recommend a pre-build chip antenna to make things easy.

http://www.digikey.com/product-detail/en/W3013/553-1677-1-ND/2543339

is an example for 800 MHz. 400 MHz might be worth considering, as lower frequencies tend to penetrate buildings better, while 800 MHz and up are somewhat more line of sight.

Just a thought I had on software: the adapter could inject itself in the calcnet network with some sort of reserved address so its status could be queried by the host. Thus you could do things like track signal strength, battery state, or control the channel you're on.