I have a problem with a setup using Xbee pro chips. The picture should show how:
the setup is and I will explain what is happening.
Due to some Line of Site limitations and building layout and distance, we
incorporated several “Repeaters” to send signals down the line.
The “repeaters” are just 2 chips on two boards, sending the signal down… you can
see what I mean on the diagram.
Now, each building has several more units inside with their own channels, again,
you can see this on the diagram.
ALL the chips are set at the factory default settings except for TWO. The channel
designation, and the Baud rate. which is set at 38400 instead of 9600.
As you can see in the picture, with the repeaters, I have line of site with all the buildings. Now this has worked in the past with no problems. All of the sudden I cant communicate with building 2 using the repeater on the maintenance building. i CAN still talk to buildings 3, 4, and 5 with no problems. I can only talk to building 2 with a mobile unit with the channel set to 13 to match the repeater on building two.
There is NO obstructions and there is clear line of site with the antennas from the maint building and building 2.
Based on the information you have provided, I would recommend upgrading to the XBee ZB product. After all, the 802.15.4 protocol was designed for Point to point or point to multi-point connections. The XBee ZB offers a Mesh capabilities that can resolve your repeater problem as well as determining the best possible rout on their own.
Thanks, and upgrading may be a possibility… in the future, but for now, due to costs of travel and upgrading, we are just trying to figure out WHY the problem is the way it is…
I still welcome any suggestions and or possible solutions for the current setup. Again, upgrading is a possibility.
a couple of more pictures here showing what we are using.
The USB is the main office “polling” board. All of the repeaters and units in the buildings are using the Serial board. These were installed about 2 years ago. Just used the basic “drop in and go” way of doing it. Only changes were the baud rate and channel designations.
If I were to upgrade, would I still be able to use the same boards?
Also, we are still using the 1083 firmware version. In the process now to upgrading firmware on all modules. Could this help?
Looking at the pciture diagram of the radio layout from my original post… you mentioned upgrading to the ZB chips…
um… would I still need to program the chips so they still have their own channels? or does this new upgrade take care of that? Meaning… is that what a “mesh” network is?
As you can probably tell… I am somewhat ignorant and the technical side of these things. we bought and used them because of the “drop-in and work” ease of use they are… and in a couple of others areas, they work great.
One thought additional to those already suggested: have you tried an energy scan (ATES)? Maybe a source of interference has been introduced affecting that channel. A suitably placed source of interference might conceivably give the symptoms you see. For a related discussion, see the very first post in this forum.
If you do upgrade the XBees, I think you could still use the same boards. Better to check with Digi when you know what you’re upgrading to, though.
As for a firmware upgrade, that’s always a good idea. I wouldn’t expect it to help with the immediate problem because as you say 1083 has been proven to work in the past. But an upgrade is helpful because it gives you access to newer features if you need them, and it’ll contain bug fixes that may improve the performance of older features. It also gives you extra street cred if in the future you should need to ask firmware-related questions
At the risk of becoming boring, I still have a fondness for my interference theory. From your original diagram, I make these deductions for the repeaters involved with channel 13:
The repeater on the maintenance building works fine with the repeater on building 5, so we assume both repeaters are in good order.
The mobile unit works with the building 2 repeater, so we assume those repeaters are in good order.
That means all the channel 13 repeaters are working fine - yet the maintenance building still can’t communicate with building 2.
So, had anyone installed any new equipment when the problem occurred? A likely candidate would be a wireless video camera because of its continuous transmission. The X10 units are often mentioned as problematic in the 2.4GHz band. One of those, situated near or on the maintenance building but shielded from the path between there and building 5, could be worth looking out for.
As for the mesh, there are plenty of people who know them better than I do. However, with a mesh you would do away with your ingenious repeater solution and set all devices to the same channel. Every node would be able to talk to any other node, the critical difference being that if a pair of nodes are out of range of each other then any other nodes can act as repeaters (routers) to carry the signal. The nodes work out for themselves what the best route for any transmission is.
Even with a mesh, though, an interfering device could disrupt its operation unless you have enough nodes to route the signal right around the interference source. That’s why I’m still harping on about interference.
One more thought does occur to me with the idea of the mesh upgrade. You don’t describe what the end XBees are doing or how your software is interrogating them, but whatever the arrangement is, would it work unchanged with a mesh or would you need new control software (eg on a host PC in the maintenance building) as well? I can’t tell one way or the other - I just raise the question.
What we do is install monitoring systems for fiberglass manufacturers. Our monitors measure the material and then the managers are able to “poll” the information the monitors generate for reporting purposes. Usually, in about 90% of our customers, these monitors are connected in a daisy chain fashion using 3 wire cable. From the office all the way to the last monitor in the line.
As you can see in the original attached diagram, they have 5 separate buildings housing these monitors. The farthest one away in 1/4 mile. This makes using cable not very feasable, hence the wireless.
The communications part of the monitor is just using 232 to 485 comms with a converter at the office.
Attached to this post is a picture which shows our wireless box that we connect to the same port on the back of the monitor as we would the cable. Inside you see the wireless xbee stuff connected to the 485 converter using a null modem setup. The converter does just that, makes the data from the monitor transmitable thru the wireless, then at the office it is converted back to make it readable by the computer.
It is a really basic setup. We have about 4 other manufacturers using this wireless setup with xbee products and have no problems at all. So, that makes it feasable that there is some sort of interference going on.
As for the mesh setup, if I put mesh capable chips in the same spots as I have the current stuff at, then that should work, right? As long as we can figure out where, if any possible interference is coming from.
The repeater pic attached shows it without the antennas or the pigtail whips attached… just put on together quickly to give you an idea. The repeaters for all buildings are installed inside, with one antenna going thru the wall to the outside and the antenna communicates with the monitors.
Also, sure glad you understand exactly how we got things routed with the repeaters.
Dave, thanks for that clarification. It really does help when trying to picture the scenario.
Your mention of the 3-wire daisy chain in other installations confirms what I was already thinking: the XBees are acting as a form of broadcast network. Does this set of deductions square with reality?
In the maintenance building there is a Box. I’ll assume for the sake of argument that it’s a PC running a monitoring program.
When the Box wants to know the reading from a remote monitor, it sends out a message that says something like “Monitor 42: report!”
All monitors receive this message, but only monitor 42 responds. So when the response comes back to the Box, the Box knows where it came from.
In other words, the XBee network is conmpletely transparent. Whatever the Box sends gets transmitted to all monitors, and whatever any monitor transmits gets forwarded back to the Box.
Also:
The site is not under your control, and is perhaps at some distance from you. So glib suggestions of “why not try this?” are maybe not immediately helpful.
If I’m wrong in the above, you may as well stop reading now
If I’m more or less right, then read on.
The problem we’re taking about began at some specific time. Are you able to quiz the site owners about what else might have happened at that time? My interference theory would look for the installation of a new piece of kit and especially a video camera, but the question in general would look for any clues at all.
And then returning to the mesh upgrade idea: I can’t immediately see how the mesh system would act as a drop-in replacement for the system you have. If you want to replace the existing system with a mesh system, then I think I now understand enough about what you’re doing to say “You need to talk directly to Digi support”.
John, you got it down good with what I have described. Except that the maint. building is a “repeater” building. The main Polling box is in the office.
Ok, I just talked to the guy and he has come up with something new lol…
There are pictures attached to give you a visual with what I will explain here.
When he uses the computer in his office, and hooks the wireless main box to the antenna that goes to the outside (Ch. D), that communicates with the repeater on building one (Ch. D & E) which will communicate with the four units inside. (Ch E’s) Doing it this way, he can only poll units 3 & 4.
Now, when he takes the computer unit and puts an antenna on it (Taking off the antenna that goes outside), he points it through his open door, which goes to another door (opened or closed), he can then poll all four units.
Again, hopefully my not-to-scale drawings should show this.
I have been to this plant several times so I know exactly where everything is at, what walls and other obstructions there are. I have asked him if anything new was placed or whatnot, and he says no. He is becoming adept at doing things with the wireless, and has already been really efficient with our monitors. So, when he says no or yes to something, he means no or yes.
Any ideas on this new problem? Also, we are thinking of buying around 10 of the new ZB modules with the mesh networking and getting him to set them up one building at a time. They have the same footprint and stuff so thats not a problem. Which leads me to one other question. I can go either 2.4 m or 900 m… since our current setup is 2.4, will the 900 m maybe work better? Only a five dollar cost difference per module.
Could it be possible that the power supply isnt providing enough… power to the units? maybe just a bad power supply… hmm but that doesnt makes sense either as we already tested the repeater from outside using the mobile unit…
Could it be possible that the power supply isnt providing enough… power to the units? maybe just a bad power supply… hmm but that doesnt makes sense either as we already tested the repeater from outside using the mobile unit…
Just posted that… meant to add that he is currently able to poll from the office… buildings 3 and 4… just now he cant poll 1 and 2…
At the very least… we know that the repeaters on the maint building and 3 and 4 are good along with the antennas and antenna cables.
This is beginning to raise more questions than answers…
When he uses the antenna in the office, I assume he’s still transmitting on channel D. That means his communications must still be going through the D-E repeater on building 1. This seems to raise the question of why the antenna inside the office is working better than the one outside it.
Yet if I understand right, the antenna outside the office still communicates fine via the maintenance building - a longer distance, seemingly. Are these antennae directional? Could there be an alignment problem? Any birds still nesting at this time of year? Could there be any damage to the antenna cable, thus weakening the signal?
It doesn’t sound like a power supply problem - if it were, things would either work or not.
Re the mesh solution, can you explain what he does when he polls a monitor? Is he typing a direct command into a terminal window which then gets transmitted straight to the monitor, or is there a piece of software with buttons and a fancy display? If the latter, are you prepared to rewrite the software if necessary?
You also ask whether 900MHz might give better results. I have a very simple answer to that - I have no idea! My own experience is only with the straight 802.15.4 devices. Maybe someone else can comment…
Finally, your comment about installing a mesh system one building at a time. If you’re picturing this as a piecemeal replacement exercise, it won’t work exactly like that because the 802.15.4 and ZB devices can’t intercommunicate. What you’ll be doing is replacing the whole network, initially with two separate polling systems in the office - one for the old network, one for the new. When all buildings have been converted, then the old network can be removed.
Mind you, I’m still not convinced that a mesh will be as easy a solution as I believe you think it is. I do think you should consult Digi on that.
Just a quick note before you buy any new modules to try mesh. You can use what you have now to do mesh. Since you are using the Series 1 Xbees you can load them with the XB24-DM/XBP24-DM. This type of Mesh is not Zigbee but it is a proprietary mesh developed by Digi.
[b][i]This is beginning to raise more questions than answers…
When he uses the antenna in the office, I assume he’s still transmitting on channel D. That means his communications must still be going through the D-E repeater on building 1. This seems to raise the question of why the antenna inside the office is working better than the one outside it.
Yet if I understand right, the antenna outside the office still communicates fine via the maintenance building - a longer distance, seemingly. Are these antennae directional? Could there be an alignment problem? Any birds still nesting at this time of year? Could there be any damage to the antenna cable, thus weakening the signal?[/i][/b]
I forgot to mention…
When he tried it from his office thru the doorways… he changed the channel to E to talk directly to the 4 units in building 1
The antennas are not directional.
Dont know about birds nesting… it IS possible, but unlikely… fumes chemicals… etc etc
The little antenna connectors that clip on the chip… Pigtails I think they are called…
Well… He was experimenting with the repeater in building one. He was watching it as someone else in the office was trying to poll a unit. When it didnt work and he noticed nothing out of the ordinary, he pressed down where the pigtail clips on, then he had the other person try again, this time it worked!
He then unclipped it and looked at it, and the little tiny piece of brass was BARELY visible. It seems that when he pressed down, he caused it to make good contact therefore able to communicate with the units 1 and 2 which were farther away.
The theory is (so far) and this is backed up by me, that there was enough of a signal to reach the close units just by the chip. Ive done that here at the shop. I was able to communicate with another unit with no antenna attached and no pigtail attached to the chip… It was a short range test as I recall, but just the chip by itself was able to work.
We ordered some new pigtails for him, so we will know in a couple of days if this was the problem. He will be putting them in the repeaters in the problem buildings… 1 and 2… and have a couple extra for other possible problem repeaters or units.
