Forums

I have two CC1100 nodes spaced some distance apart on my desk, transmitting at 315 MHz at -10 dBm, with 433 MHz antennas.  I don't believe saturation could be an issue because of the low output power and mismatched antennas.  I have also tried the same setup with CC1101 radios with the same results.

I am attempting to implement a manual duty cycling mechanism in software (long preamble transmissions for transmitters, energy based receive checks on the receivers).  I can get it to work flawlessly with CC2500 radios, but the CC1100 radios are a different story:  when a CC1100 radio wakes up in the presence of another transmitter, the carrier-sense line goes high and intermittently remains high (or toggles rapidly) even after nearby transmitters are gone.  This CS line remains high until the node receives a valid packet or about 5+ seconds of waiting around. Because the node thinks there is a carrier on the channel when there isn't, I cannot transmit any packets or even go back to sleep in my custom low power communications strategy - the node thinks there's something on the channel it should stay awake for.

I started reporting RSSI values to see what was going on.  The noise floor is about -45, and jumps to about 19 to 70 in the true presence of a transmitter.  In the case where the carrier sense line gets stuck high even when there are no transmitters nearby, the RSSI readout stays at around -1, which is well above the noise floor so would be interpreted as valid energy on the channel.  I observe with a spectrum analyzer there is nothing on the channel.

Ideally, I'd like the RSSI value to go back down to the noise floor when no transmitters are on the channel :)  I've just had no luck in doing so. What work-arounds can I attmept to get the radio to read the "correct" RSSI value and unstick that CS line?

My AGC registers are setup as follows:

  CC1100_AGCTRL2 = 0xC7,
  CC1100_AGCTRL1 = 0x20,
  CC1100_AGCTRL0 = 0xB0,