4
\$\begingroup\$

First: I'm not a professional. I just want to know what is possible and if i'm on the right track.

Second: I want a little microcontroller which just measures (for example) the temperature and send it to my server (can send/receive bluetooth). It should be a SoC. I googled a lot and found an ARM Cortex M0 at nordic semiconductor which has built-in bluetooth (2,60 $) which I want to use. Before I continue and look for a professional to help me I want to ask you ...

The question: Is it possible to put this ARM cortex M0 with built-in bluetooth on a little board with some pins and connect it to a small power source? Would this work, or would I need a lot of components (resistors to pins, capacitors etc.) on that board?

The solution should be as simple as possible to keep it small.

I also googled for existing solutions, but they where 10 times more expensive than a simple ARM Cortex M0 would cost. I found a little iBeacon/Beacon which would cost 4$ with an ARM Cortex M0 and also some pins to connect some sensors, but it was to large (lot of capacitors, resistors, ...).

EDIT:

this is the ARM I found: nRF51822

this is the data sheet

|improve this question
\$\endgroup\$
  • \$\begingroup\$ It depends on the specific arm controller, can you tell us what chip you want to use or link the datasheet? \$\endgroup\$ – Funkyguy Jul 14 '14 at 13:52
  • \$\begingroup\$ I added the information in my initial post :-) \$\endgroup\$ – Laokoon Jul 14 '14 at 13:54
  • 1
    \$\begingroup\$ Do you want to make one of these (in which case, use a prebuilt board!) or thousands? \$\endgroup\$ – pjc50 Jul 14 '14 at 14:05
  • \$\begingroup\$ If it is possible. i want to make thousands. \$\endgroup\$ – Laokoon Jul 14 '14 at 14:30
  • 1
    \$\begingroup\$ Link the "too large" prebuilt module. Some of those resistors and capacitors are not optional; we can give you some insight into how much smaller even a custom design can go. \$\endgroup\$ – Ben Voigt Jul 14 '14 at 16:48
4
\$\begingroup\$

It looks very feasible although you will need to mount the chip onto a breakout board yourself and considering it is only offered in a QFN or BGA package, you will probably need hot air rework or some sort of reflow oven.

QFN to DIP breakout boards are very simple and can readily available on eBay or some electronics store like Sparkfun.

You will need a couple external components.

  1. It looks like the chip has specific pins to place a decoupling capacitor on (which is really really nice).
  2. You will probably need a pull-up resistor on the reset pin so your device doesn't reset randomly should you leave it floating.
  3. And you'll also either need some form of antenna, it could be a trace antenna or a chip antenna, or even a hefty SMA connector for the screw on antennas.
|improve this answer
\$\endgroup\$
  • 2
    \$\begingroup\$ You'll also need to wire up some means of programming it (JTAG, SWD, etc) \$\endgroup\$ – pjc50 Jul 14 '14 at 14:04
  • \$\begingroup\$ Oh yes! that too (how did I forget that!?) \$\endgroup\$ – Funkyguy Jul 14 '14 at 14:07
  • \$\begingroup\$ Thanks for that answer. I cannot upvote it, because I new to this stackExchange site. \$\endgroup\$ – Laokoon Jul 14 '14 at 14:36
  • \$\begingroup\$ What do you think would be the design-costs range (just design, not the manufacturing)? Well I think it's hard to tell me an exact number, but maybe something like "5,000$ to 10,000$"? \$\endgroup\$ – Laokoon Jul 14 '14 at 14:41
  • \$\begingroup\$ That's a reasonable range for just the electronics design. To answer your other question, electronics engineering and PCB layout would be reasonable terms. Wouter is right that you need another $1000-$10,000 to do RF regulatory approval. \$\endgroup\$ – pjc50 Jul 14 '14 at 14:58
4
\$\begingroup\$

I don't want to sound negative, but this is a bare chip, that must be interfaced to a (possible on-PCB) antenna. This is highly specialized work, I would not touch it myself. Since you ask here about the much much easier task of connecting the uC part of that chip, I think this is way over your head.

If you want to make large quantities of your product, contact a specialist in this field and be prepared for a costly development traject (including certification)

If you want to make a few, get a ready-made module. The manufacturer's site shows a reference design. Or get a bluetooth-only module, adding a cheap Cortex M0 chip is not that difficult.

I see that your intention is to build thousands. You should do some estimates, but my gut feeling is that this is NOT enough to develop, debug, and qualify (!!!) your own HF board. I would go for an existing HF-board, and add a uC if it does not already contain one.

Are you stuck on bluetooth? NRF24L01 modules are cheap and available all over the net.

|improve this answer
\$\endgroup\$
  • \$\begingroup\$ thanks for the hint for reference designs. I don't want to do it on myself. I know that I cannot do that on myself. \$\endgroup\$ – Laokoon Jul 14 '14 at 14:36
  • \$\begingroup\$ What would be the key word to find somebody who can do stuff like that. "electronic engineer" is to ... unspecific. Are there job-keywords describing people like "ARM board designers"? \$\endgroup\$ – Laokoon Jul 14 '14 at 14:38
  • \$\begingroup\$ IMO the ARM aspect is the easy part, can be done by any half-competent electronic engineer or even an advanced hobbyist. (Even I can do that!) The RF part is a different story, for that you'll need an EE with RF experience. \$\endgroup\$ – Wouter van Ooijen Jul 14 '14 at 19:43
1
\$\begingroup\$

The chip is fine. Nordic makes a number of evaluation boards for it, one of which, like the 51822 evaluation kit, https://www.nordicsemi.com/eng/Products/Bluetooth-R-low-energy/nRF51822-Evaluation-Kit, might serve your purpose. Be very careful. I just purchased a slightly different product, the 51822 development kit, and it really wants the $400 51822 Starter Kit to be useful. Nordic is not great at making all the relationships understandable.

|improve this answer
\$\endgroup\$
1
\$\begingroup\$

Instead of using a dedicated microcontroller and a bluetooth transceiver I'd like to recommend that you use Bluegiga's BLE112/113 modules.

I have used it in a commercial product with success. It's a great product.

It has a microcontroller from TI with common hardware (ADC, I/Os, Timers, Interconnectivity...) that you can program using a high level script language called BGScript. It's very easy to understand and quick to develop simple applications with it. In you case, since you need to read the temperature sensor, it's very easy to build with this module. They have example applications that reads temperature sensors too. Also they sell a nice development kit that helped me to build my first application.

The bluetooth module is very space saving, low power (designed to run using 3V coin cells) and trustable.

Although the hardware will be very simple you will have to put your efforts on learning the Bluetooth Smart Specs to understand the GATT (Generic Access Profile), since all Bluetooth 4.0 applications must adhere to it. Also, if you choose to use the BGScript you will have to study it a bit. But it's not difficult and it is well documented.

|improve this answer
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.