Has anyone ever tried a zigbee chip?
Anyone know if it would be a suitable interface to re-program a PIC wirelessly?
Or suggest an RF product that would do the job?
Thanks!
I have not tried any of the Zigbee solutions yet, I think TI has bought over Ember and they will be offering MSP430+Zigbee silicon soon.
Anyways, if you have serial bootloading firmware, you can use the Bluetooth modules on the PIC’s UART and USB Bluetooth Dongle on the PC side, its just a simple cable replacement.
HTH
Jay
I am planning on using the PIC18F252. It has USART.
Is this compatable with the BlueSMiRF chip? Are there any aditional components I would need to get the PIC and the bluetooth chip to work together?
Thanks for the help 
I’ve been looking into zigbee for use as a distributed control network for home control. My first impression is “wow! that’s complex”. However, the benefits seem pretty compelling if it really can be made as reliable as is claimed. I wouldn’t use it just for a point-to-point solution.
How reliable is the bluetooth stuff? I’d be concerned about interference though if you are willing to require close proximity it might be ok.
Since you are looking to wirelessly program a PIC, I suggest you think about the “programming protocol” before worring about the network transport. Being able to handle data loss situations is pretty important. Like I said in the other thread I’d define a protocol that uses smallish packets - say 16 bytes of payload (though there may be other sizes that are more natural). Each packet would have a command, address, sequence number, payload and checksum. The bootloader would read the packet, test the checksum and if good, program the bytes and then ACK or NACK the packet. The bootload sender would send packets and listen for ACK/NACKs. On NACK, it would resent the same packet. Then just about any transport could be used. Sequence numbers could be used to allow out of sequence packet delivery if your transport was attuned to that (like UDP). I think this makes for a somewhat bigger bootloader but it doesn’t sound grevious…
Just got a Mouser catalog. I notice Freescale has a chip with ZigBee plus an MC9S08 on-chip (for handling the protocol, but it looks like it can also contain user code). They’re also carrying a ZigBee chip from Chipcon, no on-chip MCU but cheaper. I’m surprised Nordic doesn’t have a ZigBee offering; seems like it would fit in with their existing 2.4GHz line.
I am working with LPC2138 and CC2420. Planing to port 802.15.4 first, then start with zigbee. I have 22 months to finish the project 
Why don’t we get projects like that!? 22mnths? That’s like a half-life at SFE!!
I still don’t like zigbee. Too complex, too regulated, too power hungery, and jeebus, too expensive! $12k for dev kit!!
I prefer a simple, quick Bluetooth drop in for research 1-offs like the WiTilt and BlueSMiRF. If you need long range RF, the UM96 is cheap and much better than zigbee. If you need true mesh-node functionality, the nRF line is easy to control, though it’ll take more firmware. The new nRF chips will have much better re-transmit functionality. In the end, all better, cheaper, faster, longer range, and easier to manipulate than Zigbee. My two cents.
Oh - but please do keep us updated! I have a feeling, there will be a very easy zigbee company on the horizon that will allow cheap and easy integration rather than these large conglomerate companies trying to push 1million pcs of silicon per contract (“Zigbee in every smoke detector!”).
-Nathan
I decided to go for the BlueSmiRF option in the end, it semed to be less complex than the other options!
The distance I am working at will be less than 5m usually, so hopefully not many bits will be lost!
No doubt I will need loads of help with the boot loader whan the time comes!
You’re not the minty MP3 ‘april’ by chance?
-Nathan
Nope - I don’t know what you are on about!
Never mind - must be another customer.
sparky:
Why don’t we get projects like that!? 22mnths? That’s like a half-life at SFE!!I still don’t like zigbee. Too complex, too regulated, too power hungery, and jeebus, too expensive! $12k for dev kit!!
I prefer a simple, quick Bluetooth drop in for research 1-offs like the WiTilt and BlueSMiRF. If you need long range RF, the UM96 is cheap and much better than zigbee. If you need true mesh-node functionality, the nRF line is easy to control, though it’ll take more firmware. The new nRF chips will have much better re-transmit functionality. In the end, all better, cheaper, faster, longer range, and easier to manipulate than Zigbee. My two cents.
Oh - but please do keep us updated! I have a feeling, there will be a very easy zigbee company on the horizon that will allow cheap and easy integration rather than these large conglomerate companies trying to push 1million pcs of silicon per contract (“Zigbee in every smoke detector!”).
-Nathan
Yes, you’re right! I’am from sensor company and 802.15.4 is ok for low duty cycle, fast wakeup… Chipcon has cc2420 which sells at 5€ for single part and 2€ for volume. Few caps and patch antenna and you’re in the air. I developed some apps with BT and for (near) permanent connection is OK, but it will never beat sub 100ms joining time as with 802.15.4. This was my main decision to go with zigbee hassle anyway. I could stay with BT. Another issue is power consumption. I took som low(er) power components and whole system in active mode sinks about 80mA (60mA is for graphic LCD baground )…
OK, anyway, I took 50€ dev. board from olimex, 10€ RF demo board from microchip and few components laying around… Less than 100€ for hardware. Software is 3000€ keil pkarm. First steps into 802.15.4 stack was with gnu tools.
I mentioned 22mths, it’s for writing complete zigbee stack from scratch, my own, without any “jurasic park” partner. Maybe we can offer some nice, cheap, clean and veeeery competitive zigbee stack in near future What about making it open source??? At the moment I have PHY and MAC layer. And I am not proud to say, I am only one on this project for software and 50% of hardware team