My system has two XBEE Pro 802.15.4 modules, each connected to a PIC microprocessor.
One XB module acts as the master (M), sending periodic data packets. The other is the slave (S): when the slave receives a packet form M, it responds with a response packet to M.
This system has worked well, but with the latest modules that I purchased, it behaves in very strange way: the slave only transmits when my finger is close to the chip antenna of the slave module.
Can someone help me?
Exactly which chips did you buy in the latest set? There is one version which has a socket into which you plug an external aerial. If you have that version, without the external aerial, then it might explain the symptoms.
Thank you for your response.
The XBEE Pro 802.15.4 i’m using have chip antenna integrated.
The story is like this:
The system works well in my house. The master sends a packet, the slave responds.
I install the slave on a boat, which has several electronic devices with a power source connected to the 12V of the boat. Now the master transmits, and packets are received by the slave, but the slave does not transmit any packet. (The microprocessor that is connected to the slave module sends data to the XBee module 3 pin, but it does not transmit).
I Start the engine of the boat, and the slave begins to transmit Ok, but better if the distance between master and slave increases.
Engine Off, wrong again. I Check the 12V. voltage with an oscilloscope in DC and AC, it seems normal.
Homecoming. At home, now behaves like in the boat. If a finger close to the antenna of the slave, the slave transmits well. If you remove your finger, stops transmitting.
After an firmware update for both modules, to the latest version available on DIGI (Version 10E2), the system works well. (Like in 1).
Very puzzling.
When you checked the power supply, did you also look at the 3.3V supply to the XBee module itself? If there’s noise on that, it could cause all sorts of strange problems. The fact that it behaved differently according to whether the engine was running seems to suggest something of the sort. Is the engine petrol or diesel? A petrol engine could be generating interference from its ignition circuitry perhaps.
As to why the firmware upgrade should make a difference in the case of an intermittent problem, I really have no idea. Unfortunately Digi don’t seem to publish comprehensive lists of changes between versions, so it’s impossible to speculate here.
If it works now, maybe the answer is just to be grateful 
But I’d agree it would be nicer to nail the problem.
I suppose there is one possibility: if there is an interference problem, and if that causes the packet sent to the slave by its microprocessor to be corrupted, that would explain the absence of transmission. And then, the presence of the finger could alter the effects of the interference. This, however, is only wild speculation. In its favour is the observation that increasing the distance between master and slave helped, because that would move one or the other to a different position relative to the interference source. Against it is the observation that the firmware upgrade made a difference, because I wouldn’t expect that to have any effect on an interference problem.
I fixed the problem. My application circuit measures the voltage of 12V, and if it is too low, inhibiting the transmission. There was a problem with my firmware: I was using a version inadequate to actual PCB. That explains that inhibits the transmission.
The supply voltage of 3.3V seemed correct. Without noise. Well filtered.
But does not explain the issue of the finger.
I checked with an oscilloscope that the serial data were coming from the microprocessor to Digi module, but not transmitting.
But by bringing a finger to the antenna, the module starts transmitting!
I do not believe in microelectronics ghosts, but they exist!!
(In fact I think everything has an explanation, although it is sometimes hard to find)
Regards,
Manuel