Doing a Zigbee project?

[homebrew]

Hi,

I'm currently developing small temperature sensors which I want to read out remotely. I would really like to use Zigbee for communication but as size does matter in this project I don't want to use an of the shelve module.

I'm planning to use one of these Zigbee transceivers:
http://ww1.microchip.com/downloads/en/DeviceDoc/39776C.pdf

And an antenna layout given here:
http://www.ti.com/lit/an/swra117d/swra117d.pdf (page 4)

So my question is:
Is this feasible on a normal two sided FR4 board?
I'm a pretty experienced hobbyist regarding microcontrollers but I'm absolutely new to any RF related designs.

So what you think? Is this doable?

Am I even allowed to do (and use) my own RF designs? Are there any regulations, or is it ok as I'm technically only attaching an antenna to the IC which does all the interesting stuff?

[tan98010]

Hi,
It will be better for you get one of the module that's available from microchip(the module is relatively small), it will save you lots of time for designing the PCB and finding the component,solder the tiny SMD component.
perhaps you could make a carrier PCB and your micro controller and temperature sensor mounted at the back of the module.

the another advantages the microchip module has the EMC certification, so you no need to worried too much for the EMC stuff.
Note: ZIGBEE is working on the ISM band so you are free to used but if you wanna commercial it, yes you need to obtain and EMC certification and yes it was costly.

and the microchip stack is hard to understand, it may takes you plenty of time to study the stack and get it working, or perhaps you can  make used of the sample project available.

or if just wanna simple wireless system, maybe you can try SI4463, it was fairly simple compare to zigbee, and the best thing is low cost.    

[LukeW]

Is this feasible on a normal two sided FR4 board?

Is this doable?

Yes.

Am I even allowed to do (and use) my own RF designs?

I would say generally yes, but it depends.
It depends on the legislation of your country, and it matters if you're selling a commercial product vs. doing your own thing in the lab.

In this case, you're working with very small amounts of power, for example +0dBm for the Microchip part you mentioned, you're within the 2.4 GHz ISM band, and you're likely to be using antennas with very small gain, so the EIRP will be small.

What microcontroller do you want to use? If you're not attached to any particular decision on your microcontroller, you may want to look at single-chip 802.15.4 solutions such as TI's CC2538 or the ATmega256RFR2 from Atmel, for example. This will reduce your bill of materials to one main chip - and if the accuracy is good enough and the chip is exposed to the atmosphere/environment you want to measure you could just take the temperature off that chip directly, no temperature sensor needed.

And a planar inverted-F antenna such as the one you linked to should be fine for a basic design like this.

[LukeW]

Is this feasible on a normal two sided FR4 board?

Is this doable?

Yes.

Am I even allowed to do (and use) my own RF designs?

I would say generally yes, but it depends.
It depends on the legislation of your country, and it matters if you're selling a commercial product vs. doing your own thing in the lab.

In this case, you're working with very small amounts of power, for example +0dBm for the Microchip part you mentioned, you're within the 2.4 GHz ISM band, and you're likely to be using antennas with very small gain, so the EIRP will be small.

Generally speaking, you need to look at what your country says about any kind of exemption, "general license" or "class license" that applies to RF devices transmitting very low amounts of power under certain conditions in certain frequency bands.

What microcontroller do you want to use? If you're not attached to any particular decision on your microcontroller, you may want to look at single-chip 802.15.4 solutions such as TI's CC2538 or the ATmega256RFR2 from Atmel, for example. This will reduce your bill of materials to one main chip - and if the accuracy is good enough and the chip is exposed to the atmosphere/environment you want to measure you could just take the temperature off that chip directly, no temperature sensor needed.

And a planar inverted-F antenna such as the one you linked to should be fine for a basic design like this.

[homebrew]

ATmega256RFR2 from Atmel, for example. This will reduce your bill of materials to one main chip

Tank you for this very good pointer. I wasn't aware of Atmel SOCs that integrate Zigbee on the chip itself. I would have gone for AVR anyway so this is a perfect match.

Regarding the PCB antenna, am I right, that there should not be any soldermask on top of it? Would standard gold plating be ok?