XBee mesh mode

Hi all,

has anybody over here got any experience with xbee in mesh mode? I am making a new design which will concist of multiple pcbs with an emx or USBizi144 on it as controller. Nothing fancy, it just has to drive a rgb led. Just for fun, so i hope a USBizi144 will be enough for this.

But i would like to use them in a mesh so each of them relays it signals and the range will be longer. For example i want to place 2 of them in my living room and then a 3th in the hall, etc etc.

xbee mesh mode would be just perfect for this i guess. But anybody has got any experience with the mesh mode and .net micro?

Regards,

Per

This is actually something I’ll be testing soon myself. Since the mesh is completely controlled by XBee I don’t believe there is anything special you have to do from the NETMF side. From what I’ve read so far all that’s really required is to set the PAN IDs for all the XBees in the mesh the same and it should relay the data on it’s on. From NETMF it should seem that you’re still only dealing with a single XBee.

Be sure and update this thread with your results. I’m using XBee to control a copter so I’ll be doing some experiments with XBee soon to discover it’s real range limits. Once I’m ready to exceed those, I’ll upgrade to the Pro modules but for now I’m using the standard modules I already have.

Hi ianlee74,

well i must say i have done just some small tests with xbee modules at this moment, just point to point. They where pro modules and got quite some range in the open field. We needed to cross some water with a data logging device. Worked fine, but it was done with an atmel module and in bascom avr.

Would be nice if you could also share your findings on this. I am curious to find out what the range from node to node will be then. This project is just in it’s beginning though. I is something that i build between the projects i design for my professional job. It will all need to run on a 12v battery so i am still looking for a good way to calculate what kind i need if it all needs to be able to run for let’s say 12 hours.

Regards,

Per

A couple of in-my-head calculations and wild guesses on standard 12V automotive batteries and pessimistic figures for power draw and SMPS efficiency gives me:

Over 1 week

Along with what ianlee74 said, unless you have to change the xbee configuration dynamically at runtime, all you have to do in NETMF is the serial data stuff. I’m also currently getting the hang of configuring xbees, and have found with the X-CTU tool digi provides, the learning curve almost disappears for the xbee command set.

Now it would be cool if somebody wrapped up those commands in a nicely commented NETMF class to share with the community…

[quote]Now it would be cool if somebody wrapped up those commands in a nicely commented NETMF class to share with the community…
[/quote]

I had started down that road until I also discovered X-CTU and decided it wasn’t worthwhile since I so rarely reconfigure the modems. Here’s a place to start…

Have you ever checked out the XBee classes in Michael Schwartz’ Toolkit? He by far has the most functional C# classes that I’ve run across for XBee. I used it in a previous project and if you need advanced XBee functionality it’s the way to go. Unfortunately, the documentation is almost non-existent. Have fun…

http://mftoolkit.codeplex.com/

@ ianlee74 - i did see what was on the wiki and had some kind of unexpected petit mal seizure when i saw the arduino code :slight_smile: thanks for the pointer to the mftoolkit. i have looked at that but missed the xbee part. ‘wrapping it up’ can be tedious development, but it would get someone good community karma…

I could definitely use the karma… But, I couldn’t convince myself that I would ever have a real use for it. If perhaps someone has a project where they need to dynamically change the PAN IDs or baud rates then it might be of use. Otherwise, it’s something I normally set once then forget. Maybe annually I pop out the radios to upgrade the firmware but that’s about it.

Per - also, I don’t know if you’re aware but the Control4 system (http://control4.com/) operates exactly like what you’re trying to do. Looking at their system might give you more ideas. I demoed it earlier this year when we were building our house and it’s very slick.

So morning again in the Netherlands :slight_smile:

I always think it is a good idea to be able to change baud rate settings etc by your program. It might take a bit more work, but it will make a better device (to my opinion). I saw the zigbee parts in the mftoolkit also, but i didn’t really take a good look at them yet. But for the beginning it is ok to do the setup with the external program of course.

@ ianlee74. I didn’t know about them, but they seem to have nice products. Most of them look a bit like the standard x10 products you can find everywhere (but then with a different communication interface of course). I am not going to design a home automation system though, i already have everything with x10 at this moment (running my own control software and touchscreen software). Still want to change it to something else because x10 isn’t such a good way to do it. But that is all for the future, since it is working ok at this moment :wink:

It is more of a test thing with meshes and how far they can range. Some kind of portebal rgb light i can take in and around the house. For the battery part, it is not going to be a car battery, those are to big. Something like this is more suitable [url]http://swingleydev.org/blog/p/1896/[/url] . I guess it is easy to just connect everything and measure the power usage (and then calculate the needed battery capacity to run for 8-10 hours). Maybe it would be a good idea to monitor the battery also then through processor.

Well, if 35 Ah gave you a week of runtime, then a 5 Ah (such as the one in the linked article) should give you 24 hours, right?

well in an ideal world that would be correct, but the power decrease of a battery aren’t linear sadly. But there are some online calculation programs for that online. And for cases like this simple trial and error methods are also good :slight_smile: Just hook up the design to a battery of 5amp and see how long it will run :wink:

I have got a power led rgb wall washer here that runs on a battery, as a demo from a supplier. It will run around 10 hours on a full load. But there are quite some 1w power leds in the unit. Maybe i should open it and see what kind of battery they use…