RF Modules (Zigbee)

[Chris Barron]

Although mentioned in another thread, I thought it might save a few people some money if, when they're looking for a simple RF data link, they consider the Telegesis ETRX2 modules.

These use a self healing mesh networking topology so you can introduce as many nodes as you wish to your network and once created by you the network manages itself.

These particular modules have inbuilt digital I/O and A/D, so,
using a simple remote temperature logger as an example, we can do away with the remote host processor completely and replace the RF Module+PIC combination with a single ETRX2.

Sending the following command to the distant ETRX2 from the main coordinator (and data sink) makes the remote send back data relevant to the I/O pins and the A/D

ATREMS:<address>,XX?

Where <address> can be the remote node’s EUI64, NodeID or address table index and <XX> is the register number relating to the data.

All commands are in the AT format so communicating via PICBasic and a PIC is made easy.

My units cost £11.90 each and I suppose if you were making a simple 2-node network then you might be able to beat that price (especially if you are operating in simplex mode), but I feel the ETRX2 are more reliable and less noise prone than other RF modules I have used before. The units have a feature where they can actively search for the quietest channel, and data transmission can be monitored and the creation of ACK and NACK messages is automatic, which means that we can have some reassurance that a message was delivered.

I don't work for Telegesis, but recently have tried several Zigbee modules and development kits and this was by far the easiest way to go
http://www.telegesis.com/products/test_page_1.htm

The full list of commands is on the document download page in a document called "R3xx command list"

[Normnet]

These use a self healing mesh networking topology so you can introduce as many nodes as you wish to your network and once created by you the network manages itself.
Is the mesh in hardware or is it code with a C or Basic example?

Norm

[Chris Barron]

Is the mesh in hardware or is it code with a C or Basic example?

Norm


Norm,
these particular modules are based on the Ember EM250 chips, which have the Zigbee stack built in, written in C code.

you could, if yoiu wanted, blow away all signs of the Zigbeee software and use the devices as ordinary IEEE 802.15.4 compatible radio modules, IE just use the hardware and lay your own software on top. Ember provide the required development tools to do that, as do (I think) Telegesis, who will help you out with any special versions of firmware you might require.

Once I started to play I realised that it might be fun to try to implement something similar in PicBasic for use with the vanilla 433MHz (etc) modules. The software stack needn't be too large. Microchip offer two stacks for their hardware, one if them is Zigbee 2006, the other is their own protocol called 'MiWi', which is a 'star and cluster' topology capable of sending messages between nodes which are a couple of hops away from each other.
The Microchip stacks are written in C and are free to download from the Microchip website, and may provide some ideas. MiWi is a significantly smaller protocol than Zigbee, and can be implemented in a smaller pic than is required for Zigbee

Chris

[Chris Barron]

I've just been reading an online newsletter (Free to subscribe to if anyone wants to stay up to date)
http://www.embedded.com/217400339?cid=NL_embedded
....which poiints out that the Zigbee Alliance is planning further integration of internet standards into the Zigbee protocol. This would mean that you ought to be able to connect anything to the internet using wireless (plus a gateway too, of course)

[Ioannis]

Nice one. Thanks Chris.

Ioannis

[lester]

If your building a ZIGBEE or any other wireless network at 2.4ghz, a major consideration is through put of data and integrity of the wireless link.

Its a good idea to use a spectrum analyser to ensure that your locality is clear of potential killer interferance such as baby monitoring devices, leaky microwave ovens, wire less video camera's and wireless alarm modules, all of which can cripple zigbee or 802.11 network performance

Take a look at www.wi-spy.co.uk

Here is a little video that explains the project range

for home users there is a low cost option: http://www.crownhill.co.uk/product.php?prod=2139

Seriously worth the money, even we cleared up a rougue 2.4ghz signal radiating intermittantly accross the band, without wi-spy we'd probably never have found nit......even though we have a 30K GBP Agilent spectrum analyser sitting in the workshop, wi-spy is just so convienient.

[Chris Barron]

Thanks for that Lester.
In the end my chosen modules, the ETRX2 by Telegesis turned out to be based on the Ember250 chip.
In normal operation the unit has a function which finds the quietest channel before forming the network. Operating any 2.4GHz equipment while it runs this 1 minute long routine helps it to select the best channel. Obviously , it may be a different story at the other end of the wireless llink.

I think I read somewhere that the replacement for the EM250, known as the EM260, may have the ability to run split channel operation for that reason.

I can't complain about the EM250, the automatic quiet channel function got me out of a problem I had with interference from a badly made video sender. The EM260 would be my first choice these dayd just because they seem even easier to use !


Chris










think I read that the replacemebt

[lester]

Yes, the Telegesis solutions are nice, well engineered products.

Notice though, they promote Wi-Spy on their web site http://www.telegesis.com/telegesis_zigbee_technology_-_technical_support_/wi-spy.htm for the very reasons that i mentioned.

There is nothing better than visualising the environment that you are working in, to better understand what your wireless link is 'seeing'.

Rememeber the zigbee prodcuts and wifi products only look at the spectrum on a per channel basis....its what they cannot see thats important!

So its worth remebering that there is a low cost solution for Spectrum analysis should you need it.

I also recall that in Circuit celler there was a 2.4ghz analyser project, so its quite feasable to build one if you are working in that area a lot.

The Circuitcellar hardwrae, PBPro and a 128x64 GLCD make a great portable spectrum analyser with better than per channel resolution.

[Chris Barron]

Yes, the Telegesis solutions are nice, well engineered products.

Notice though, they promote Wi-Spy on their web site http://www.telegesis.com/telegesis_zigbee_technology_-_technical_support_/wi-spy.htm for the very reasons that i mentioned.

I wasn't trying to make an argument against it, and if someone did have an EMI problem which was hampering their efforts, and which could not be solved by using the inbuilt channel search functions, then they might out of curiosity decide to try to use another technique for finding the cause of interference which is over and above what a well trained engineer should do first of all, such as turn off all other known 2.4GHz sources and then reactivate them one at a time until the problem reoccurs.

When I initially dealt with Telegesis for support they promoted the Wi-Spy as a form of surveying tool, preferably to be used prior to deployment of a wireless networking solution. In that respect it would be ideal if you can buy a cheap device and use it to discover that your chosen band is too noisy to begin, rather than pay out £100's on radios which can't break through the noise.

I see that Zigbee has now been fully certified for other frequency bands, and also that several manufacturers offer meshing radios on sub 2.4GHz bands which do the same as the 2.4GHz devices but in a band which has fewer devices assigned to it. This might be the better option when you are in a noisy environment on the 2.4GHz option.

Personally, I wouldn't pay for a Wi-Spy if I am in a typical domestic environment and look to introduce a small 2.4GHz network, when it is much cheaper to buy two £12 Zigbee modules and test them for yourself. It seems probable that if you discover a problem with Zigbee radios which you can relate to the environment being too noisy then at that point you should have already noticed poor wireless path reliability on your other 2.4GHz devices too, and that should lead you to choose another frequency band anyway.


I agree that it is sometimes useful to 'see' what is going on, but if you see a strong signal close by which does not cause interference, or if you see that the band is quiet, the radio units will still work noticeably well, but if there is a signal which causes interference you will definately know about it, unless it is so intermittent that to catch it you need data logging tools. Even then you might never track it down.

I used a Microchip 'Zena' Zigbee analyser which I would recommend buying as a useful tool for analysing data paths, packet data and wireless node settings. You can see at a glance where transmissions are being interfered with just by analysing the data being produced and if you're a newcomer to Zigbee and it's seemingly baffling modes of operation then the Zena will help you make sense of any problem which has to do with data transmission.

But, Zigbee is a bloated concept in terms of software, and my preferred devices (when they become as cheap as Zigbee that is) are going to be the 6LoWPAN radio devices, because they promise full IPV6 compatibility right down to node level, and in effect that means that any device can be given an IP address on the internet, very cheaply. These will also eventually be multiband devices so again there will be no need to stick to 2.4GHz

I don't know if such a lengthy comment was needed, and I hope it doesn't deter anyone from using Zigbee because if you find a vendor with a product as easy to interface with as the ETRX2 is then you can't really fail to be successful. Creating longer range data paths is as easy as adding another cheap Zigbee node in the datapath to act as a repeater and therefore radiated power levels never need to become so high as to cause interference to neighbouring devices. I think it is Stockholm in Sweden which has introduced smart power meters in every home, all working on a Zigbee mesh, and the reports are that interference issues have been kept to a minimum by careful channel selection on the Zigbee side, and no interference to other 2.4GHz devices has been detected due to the overall low power nature of the devices being used.

Chris

[rsocor01]

I found this Zigbee 2.4 GHz IEEE 802.15.4 radio transceiver module at the very low price of $9.95 each one.

http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en535967

Now, with all the debate that is going on in this thread about interference issues, I guess I would have to research this product even further before buying it. Any comments on this transmitter and the interference issues that have been discussed here so far?

Robert

[Chris Barron]

Robert, I have a trio of those modules working together running the MChip MiWi protocol. The host processor is an 18f4620. Because those modules are able to report on noise levels it is simple enough to write code for them which causes them to find an empty channel.

I am able to run Zigbee and MiWi on the same channel due tot he fact that the 'return receipt' can be relied on to provide an accurate inidcation that a message arrived.

The reason those modules are so cheap is because the user has to write a protocol stack for use on the hist processor where as rhw Telegesis modules already have that part figured out. Although the MChip MiWi protocol is good enough for many ordinary uses, and that is free.

Tge Ember 260 chip is like a condensed ETRX2, but with more functions thrown in, and also has a very easy cgoing command set.

Chris