Jump to content
nathancrum

New Dev Platform for MSP430FR5969

Recommended Posts

Hi Guys,

 

My name is Nathan.  I've lurked around here on 43oh a bit, but this is my first official post.  I've been working with the MSP430FR5969 for several months now and I've quickly grown to really like it.  It seems to me TI's FRAM processors should really be getting more exposure for battery powered development in both professional and maker communities.

 

To that end - I was hoping to get some feedback on a product idea.  We've packaged the FR5969 into a tiny coin-cell powered module.  The module can operate as a stand-alone processor or it can plug into an UNO-form-factor breakout board with an eZ-FET lite programmer, allowing you to program it with CCS or Energia.

 

We're waiting on the first PCB's to debug the base product right now, but the goal would eventually be to offer with with a handful of matching small daughter-boards to add WiFi, BTLE and either the Sharp LS013B4DN04 hybrid screen or an e-ink display - all except the WiFi would be powered directly by the coin cell (WiFi would require a couple AAA's).  It might make sense to have a few sensor boards too depending on interest level (humidity/temp, compass, GPS, accel, gyro, prox, etc.)

 

Does that make sense?  Any thoughts?  Is this a product you think people would be interested in?  It doesn't seem like anyone is offering really well-packaged small, ULP, battery powered MCU modules, especially with Arduino code compatibility and this would fill a good niche (certainly one I have some uses for anyway).

 

Anyway - I'd appreciate any feedback or suggestions.  Thanks.

 

NC

Share this post


Link to post
Share on other sites

@@spirilis - Right now the processor module has standard 0.1" headers so you can pop it on a breadboard.  We could use smaller, but the limiting factor for size is the CR2032 anyway.  IO includes 14 digital I/O's, 6 analog, UART, I2C, SPI and SBW (26 pins in all, one NC).

 

@@Rei Vilo - Yes similar idea.  Although I don't know why, but I haven't seen any micro-form-factor Arduinos like that based on the AVR equivalent to the MSP430FRxxxx - the XMega processors:

http://www.atmel.com/products/microcontrollers/avr/avr_xmega.aspx

 

There are also a few low-power libraries for Arduino, but they're fragmented and don't seem to provide the ease of use of LMP3 on the MSP430's.  Plus none of them have FRAM to allow you to leverage the equivalent of LMP3.5 and 4.5.  The biggest problem though is none of them are really aimed at being low power - so even if all they have is a regulator or some other secondary IC on them, usually it has a quiescent current of several mA - making sleep irrelevant.

 

@@yosh - basic renderings attached.  The processor module is currently 1.2" x 1.42".  It includes two low mA LED's, an int/ext power jumper, a CR2032, a 32kHz crystal and of course the FR5969.

post-38251-0-81870400-1410792095_thumb.png

Share this post


Link to post
Share on other sites

Layout for the programmer attached.  As mentioned this is in the form factor of the UNO, but it's actually an eZ-FET lite programmer using the MSP430F5528.  It includes the F5528, micro-USB, 4 LEDs (2 for the ezFET, 2 for UART), a regulator and a breakout for standard JFET.

 

...and before you ask, yes the product name is Storm Ninja.  For those unaware, it's a joke from one of our favorite shows - Archer.

post-38251-0-34777200-1410793069_thumb.png

Share this post


Link to post
Share on other sites

Why is the programmer board in UNO form, rather than Launchpad form?  (Sure, Arduino is popular but people who use MSP430 may already have boosterpacks).   Might make sense to have a programmer with boosterpack connector, and/or to have an adapter so can connect expansion boards for this to regular launchpads as boosterpacks.

 

(Facilitate testing/debugging, etc.)

 

Since there are already several small platforms out there, TeensyMicroduinoMaple Mini, Arduino Pro Mini, etc., etc., ... , it would be nice if there was some compatibility between them. Rather than defining yet another socket width, with yet another pinout, and producing a few boards which will only work with this socket, it would be nice if you could adapt a pinout compatible with some existing project.  Then you could leverage existing boards for that pinout, and add to it with additional expansion boards.  http://xkcd.com/927/

 

If need be to facilitate the above goal, might consider making the battery a separate booster pack - allowing making the boards smaller (narrower) so they would be more breadboard friendly.  (Would also give more flexibility if one was incorporating the board into a project and wanted to make the device thinner by putting the battery next to the board, rather than stacked).

 

Have you considered how that pin spacing will do for creating an xBee adapter?  (Looks like width may be close enough to width of an xBee socket that could make things more challenging)?  (I assume pin spacing is selected to work well with various other breakout boards.)

 

Solar power board?

Share this post


Link to post
Share on other sites

@@igor I get you're annoyed there are multiple connector layout standards, but you think we should match our programmer to a less common one than Arduino?  Your reasoning is exactly why we put it in UNO format instead of boosterpack - to support the most existing developers and hardware.

 

We'd love to support both, but you can blame that impossibility on TI - why on earth did they set up the pin spacing on the launchpads to be exactly 0.10" shorter than Arduino (gah! so close!).  If it was at least 0.20" we could just add a second set of headers inboard.  I'd have to look at the pin assignments, but I don't think we could get away with only one side double-row and one side shared with Arduino.  Maybe we can just add an off-center double row to support both, but it's not worth it if it means we need 4-layers.  Jumper wires works fine for development.  ;)

 

Anyway - it's not in stone, but I think you understand why we started with the UNO format.  It's possible we can support both - we'll look at it some more.  At a minimum, a breakout to connect just the processor module to boosterpacks (like the Sharp LCD) makes good sense.

 

Regarding the micro-board - that one is totally up for grabs.  We started with "small", 0.1" spacing and CR2032 for a prototype, but all of the above could change.  I agree - it would be great to leverage another micro-format and the CR2032 is cramping our style there.  That said - most of the other modules out there aren't low power enough to run on a CR2032.  There is a sea of products out there that can easily run off a CR2032 - I'd like to leverage them and build an ultra-low-power toolbox, for the same reason as using the TI FRAM processors.  Maybe I'm just not seeing the existing products, but I think there's a big gap in this area in general.  Do you perceive any of the products you mentioned as having enough traction to support and help expand as a standard format?

 

The XBee compatibility is definitely worth looking into.  Again - no way even the 1mW radio could run off a CR2032 (50mA TX), but it's pretty well established and I'm sure would have good crossover in potential applications.  The size is also probably the closest of any existing format.

 

Solar is something we'd like to play with too.  You could run a datalogger off a tiny panel and a supercap.  Add a radio and you start to need a panel bigger than the board and probably a small rechargable, but yes - certainly within the realm of the possible.  If you only need daylight operation, I'm sure a tiny panel would be enough for many applications.  Is there an application you had in mind for Solar?

 

What do you guys think about adding a low power accelerometer to the processor board?  RTC or pin wake-ups cover a lot of ground, but having an accelerometer that can wake it up would be useful in a lot of portable applications I would think.  You could power it directly off an IO pin so you could kill it completely if you don't want it.  Freescale's MMA8652FC is pretty slick and not too pricey (~$0.80 @ 100).  Seems like that might be a worthy addition for a low power focused board.

Share this post


Link to post
Share on other sites

For RF daughterboards I'd like to see something using the ChipCon-based sub 1GHz radios.  The only way you'd get low-power WiFi is using some proprietary chips (the CC3000 absolutely is not, and I don't think the CC3100 makes it either); my understanding is BT is pretty hungry too.  If you use the Anaren modules it could be pretty easy to put together.  Both 433MHz and 915MHz would be nice.

 

Have per-unit end-user costs been mentioned anywhere?

 

[addendum: I don't think the Anaren would run off a coin cell either, but even with AAs it'll run much longer than any WiFi solution.]

Share this post


Link to post
Share on other sites

@Rei We're not necessarily DOA without 5V, but the thought had occurred to us.  The processor board will be 3.3V, but we were considering using something like this on the UNO breakout to provide both 5V compatibility and higher current capacity (MSP430 is only 6mA per pin max vs. the UNO's 40mA):

http://www.ti.com/lit/ds/symlink/txb0108.pdf

 

@@pabigot Our tentative plan was to use the Microchip RN4020 for Bluetooth LE.  16mA is skirting with the limit of what you could get out of a coin cell, but as long as it's used in bursts it should be ok.  For continuous use or WiFi you'd need AAA's, but the RN171 from Microchip is the best module we've found for WiFi.  Configured for 0dBm TX, you can get the power draw <100mA.  Used in burst mode, it can wake up, connect to an AP, open a TCP socket, transmit data and go back to sleep inside 50ms.  That should allow a couple AAA's to last several months.  Again - continuous mode will obviously kill it much faster.

 

Regarding ChipCon - I hadn't seen those before.  The CC1000 isn't bad - only 10.4mA TX for 77kbps.  That's a little less hungry than BTLE, but the data rate is about the same (unverified data) and it's limited to a proprietary radio so...probably not worth the power trade-off.  The CC1101 is perhaps more interesting @ ~15mA TX & 500kpbs.  That's about half the power and similar throughput to BT2.1.  That could be a good setup for a low power star network with a WiFi hub or maybe mesh?  Thanks - we'll definitely check those out some more.

 

We haven't talked about costs yet.  The approximate goal would be to keep the processor module below $20 and the programmer below $30.  That may change depending on how many extras like the level shifter and accelerometer we might throw in, but ballpark - that's the plan.  It seems like people are more sensitive to cost than features though so our priority is price.

Share this post


Link to post
Share on other sites

@@pabigot Our tentative plan was to use the Microchip RN4020 for Bluetooth LE.  16mA is skirting with the limit of what you could get out of a coin cell, but as long as it's used in bursts it should be ok.  For continuous use or WiFi you'd need AAA's, but the RN171 from Microchip is the best module we've found for WiFi.  Configured for 0dBm TX, you can get the power draw <100mA.  Used in burst mode, it can wake up, connect to an AP, open a TCP socket, transmit data and go back to sleep inside 50ms.  That should allow a couple AAA's to last several months.  Again - continuous mode will obviously kill it much faster.

Is that 50ms verified personal experience or anecdote? Power-up, associate, transmit, disassociate, and power-down in 50ms for 802.11 is pretty damned fast. I haven't used RN modules for over three years, but with some competitors I've evaluated under contract since then I was seeing on-times closer to a full second best-case, longer if I actually wanted confirmation that the packet had been received.

 

Regarding ChipCon - I hadn't seen those before.  The CC1000 isn't bad - only 10.4mA TX for 77kbps.  That's a little less hungry than BTLE, but the data rate is about the same (unverified data) and it's limited to a proprietary radio so...probably not worth the power trade-off.  The CC1101 is perhaps more interesting @ ~15mA TX & 500kpbs.  That's about half the power and similar throughput to BT2.1.  That could be a good setup for a low power star network with a WiFi hub or maybe mesh?  Thanks - we'll definitely check those out some more.

ChipCon was and perhaps still is the RF solution for much of the wireless sensor community; it's the basis of the CC430. The CC line has been around for years (the CC1000 is way obsolete) and even though most configurations are proprietary radio, unlike the WiFi SOCs you do get full control of the radio and open documentation so you can actually do interesting things. Using Anaren's AIR modules you could get 433, 915, 2500 MHz proprietary radio as well as 802.15.4 options with little rework. I suspect it's also best-in-class for power and range, at least until Silicon Labs/Energy Micro finally turns their EFRA/EFR4D Draco pipe dreams into purchasable items.

Share this post


Link to post
Share on other sites

@@pabigot I have to apologize - I thought for sure someone at Microchip had said 50ms and I had that number floating in my head for weeks.  It probably came from the attached (convincing) power graph, but but I just checked the presentations and they also included an actual time trace.  *cough* I was only off by an order of magnitude *cough*.  Later they also claim typical on time of ~250ms so they're all over the map.  I haven't personally verified any of this yet, but I have both an RN171 and RN131 and plan to play with them soon.

 

Time Trace:

Total wake-to-sleep time = 463 ms

Associate to Sleep = 304 ms

Time socket is open = 107 ms

Packet transmission = 27 ms (bulk of tx power usage - maybe this is what got me confused)

 

They claim this includes a successful transfer confirmation response.  Even at a conservative 0.5s @ 100mA, we should be able to get several tens of thousand of transfers out of a couple AAA's.  Obviously depending on application, YMMV.  

post-38251-0-55561200-1410891513_thumb.png

post-38251-0-62621300-1410891520_thumb.png

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...