Bluetooth MCU

[adamlewerin]

Hello!
I am working with a school project where I need some kind of bluetooth communication and MCU as cheap as possible.
The MCU is required to have at least 2 digital ports and 1 analog, but will do very limited amount of work.
At first I settled for HC-06 and ATtiny13A, but after reading a bit I realized that the HC-06 has an MCU with flash that is way better than a ATtiny13A.
Unfortunately I have no possibility to buy the SDK from CSR due to it being way to expensive, du you know of any alternative?
At the moment I am looking into using this [ https://github.com/pfalcon/blutunode ] firmware in order to read and control IO, do any of you see anything horribly wrong with that?

(The MCU will be reading battery voltage and switching an LED + vibrators in an alarm clock)

[andyturk]

http://www.nordicsemi.com/eng/Products/Bluetooth-Smart-Bluetooth-low-energy/nRF51822

[autobot]

Look at silicon labs, "blue gecko". $0.99

[westfw]

Ready to go module, or bare chip?
I'm experimenting with Bluetooth Low Energy (BLE) now using some Simblee (https://www.simblee.com/ ) (ARM microcontroller with BT, programmable via Arduino IDE/etc) and Arduino 101 (Intel "Curie" module)
They're cheap by the standard of previous BT programmable boards (<$30, compared to  something like $60 for the Arduino BT, or even a $30 Arduino board + a cheap $6 BT module), but of course expensive compared to bare chips. (The simblee module itself is about $20 (0.5mm smt package), and the Intel Curie module by itself is supposed to be around $10 "soon" (16Q1))  (OTOH, FCC certified/etc!)

Note that BLE is not the same as bluetooth. :-(

[matseng]

Look at silicon labs, "blue gecko". $0.99

How many of them do you need to purchase before hitting $0.99?  The EFR32BG (just the chip on the blue gecko module) seems to be at $2.60 upwards at 2500/reel quantity.

[andyturk]

Note that BLE is not the same as bluetooth. :-(

True, but for non-audio purposes, BLE is way easier to deal with. Sending data over classic Bluetooth to iOS requires Apple's authentication coprocessor (another chip on your board) and to get that, your design has to be pre-approved by Apple.

[adamlewerin]

Module or chip does not matter, both can be incorporated in the project - we are gonna roll our own PCB anyways.
Bluetooth 2.x or 4.x does not matter either, since nobody in the group wants to develop for iOS.
The nRF51822 seems interesting, but with a relatively big price tag compared to the BC417 in HC-06. We might actually go for this anyways, but anything cheaper is appreciated.

[deksnis]

What Bluetooth profiles are you planning to use?
If you need only low data rate data transmission, then BLE is direction where you probably should look.
Lately I am fan of Cypress PSoC, so I can suggest to try it: http://www.cypress.com/products/psoc-4-ble-bluetooth-smart
Of course Blue gecko also should be ok.

If you need Bluetooth audio (A2DP) or other classic profiles, then it will be more difficult.
Even then I do not suggest to start CSR development, but already pre-programmed modules. (BC127, LAIRD BTM511, WT32i, etc.)

When you say "school project", I understand that that your goal is not to develop best product for mass-production, but to get concept working ASAP.
In that case use of some Bluetooth module can save time and money. (Even in small manufacturing batches (up to 1000pcs), because lot of them comes FCC, IC, CE certified)

[hamdi.tn]

Just use a bluegiga BLE112 or 113 it has an embedded mcu that you can program easily and your application with run on few hours, a lot of I/O and analog input, PWM output. could be pricey but it save a lot of dev time

[hli]

If BLE is OK with you (its a little bit more complicated to program with than a simple SPP serial protocol like in the HC-05 modules), then I can only second the Cypress PSoC 4 BLE recommendation.  They are very easy to program, and being able to program the hardware parts makes low-power consumption quite easy even with more complex stuff. And 49$ buy you a complete development system with two BLE modules and a base-board containing the programmer.

[adamlewerin]

Low data rate transmission will do just fine, the module won't transfer anything other than commands to set the two digital pins high/low and read the ADC voltage. I just ordered a nRF51822 module to test, since none of the other chips proposed can fit our very tight budget.
The hole product that we develop is about 8 USD so for example the BLE112 at 10 USD is a bit steep, likewise the Cypress PSoC 4

We are only required to make the product on paper but my group decided that we want to develop it fully to test it out and maybe even selling it in small amounts.