We are experiencing a series of communication problems when we have more than app. 20 Ant devices in the same room ( Ant+ Bike power devices used on indoor bikes ). The problems manifest themselves as the receiving devices can miss data for extended periods of time, up to several minutes. The problem escalates as more devices are active at the same time. ( tested with up to 40 devices in a 7m x 10 m room )
In order to track down the problems, a separate PC with a nRF24AP2-USB based USB dongle has been configured to listen to all communication by opening a channel in Rx Scan Mode with extended messages enabled.
A few informations which will be referred to later:
The Bike power message has a ever increasing "count" field ( well, one byte ) , which can be used to detect missing messages.
The time stamp in the extended message response will normally differ by 2560 or 2304 between two messages.
The messages received does occasionally exhibit some strange timing behavior:
One thing is that a message may be lost, that's expected. Receiving the same message twice with no delay
is not, but it happens.
We also receive messages with device id for one device and data from an other.
From time to time a message is delayed for up to 1.5 second, where after 1 to 6 messages are received within a few milliseconds. That the message is delayed and not the result of a series of lost messages can be seen from the "count" field.
Some time this delayed message is followed by a single message with the next "count" value, where after the next ~5 messages are lost, other times ~5 message are received within a few milliseconds with the same timestamp but wrong data.
I have included an Excel file with log dumps of some of these problems.
Some of these problems may be harmless, others definitely give problems, but I do not know whether they are the reason for our minute long dropouts.
I would like to know the reason behind the timing anomalies and what can be done to avoid receiving erroneous messages. Anyone?
Regards,
Klaus Roy-Poulsen
