We have a network of about 25 nodes running in an indoor environment. the nodes are programmed to wake-up from hibernation every 30 mins and transmit data to a logging server.
The network has one Coordinator and 3 router nodes which are always powered. when the network is switched on it works for about two days with no problem whats so ever. but after that time about 6 nodes randomly stops communications. furthermore we found out that they have a problem to be associated at wake-up or subsequent wake-ups.
I’d be pleased if anyone can shed some light on this.
I m not sure about the sleeping parameters of the devices. Create an Online Support Request at the following link, http://www.digi.com/support/eservice/login.jsp , digi experts will help you to solve your problem.
If your issue is already solved please update the forum so that other users of this forum can use the solution to solve their issues.
I’ve just finished installing a bigger network (30 end-devices, 2 routers) with 12 devices working in sleep mode (sleep times range from 1 to 5 minutes). Except for one node that constantly losses connection all work fine for days (network is running for more than one month now).
Days ago I read in this same forum a post from lynnl talking about how important is setting, in both routers and coordinator, the max time an end-device remains in the router/coordinator child tables.
The essence is that, if this time is shorter than the end-device inter-wake periods, routers/coordinator delete from their child tables sleeping devices since they think that devices has roamed (or in general, that devices have left the network). When the end-device wakes and send data, routers/coordinator dont find a table entry for them and they drop the data.
The way of solving it is stating the right value for SN/SP in the ROUTERS and COORDINATOR. In these devices, the max. time a child is kept in the parent table is three times the value (SN*SP)/100 (in seconds).
So for instance, if your end-devices wake each 30 minutes, you must set a parent table timeout (I repeat, in routers and coordinator) higher than 30 minutes, ie, (SN*SP)/100 must be higher than 600 (since 600 seconds = 10 minutes and parent table timeout is 3 times that value, ie, 30 minutes).
So make sure you are using the latest 2xA7 firmware.
Also make sure you have the same Sn?SP in all the routers (ao any potential parent), plus use the JV=1 setting in routers so worst-case they ‘demand’ to find a coordinator after the router reboots. Without this, trying to ‘move’ the coordinator to a new channel is nearly impossible since the routers remain forever on the old channel.
Be aware that some older Xbee firmware have ‘bugs’. For example the 2x70 series in end-devices at times caused sleeping node to fail to reconnect AFTER the parent was rebooted. Also, the 2xA0 had join issues where even a ‘router’ might think it is associated to a coordinator, but the coordinator would never report it as joined.
- the details are on this Wiki page: