CC1101 receiver threshold testing

Anybody run the canned packet error tests (PER) for the chipcon cc1101 eval boards? I am getting very poor receiver thresholds (the power level where the transceiver chip starts to have 1% packet error rate). Data sheet says I should have thresholds better than -90 dBm, while I am measuring more like -70 to -75 dBm.

I am trying to move 250 kbps at 915 MHz, and have tried the chip in GFSK, FSK, and 4FSK modes–all with poor test results.

Any tricks I should know? Maybe someone has a better chip to recommend that has a better RX threshold value.

Have you asked Chipcon/TI?

They directed me to their TI forum, so I figured I should post here too and maybe get an answer.

Long story short, the eval board has a “canned” software packet error rate test, where it sends packets between two boards. If you run that test, you get very poor receiver thresholds. Some sort of software glitch.

If instead you hook up one Chipcon eval board as TX only, and the 2nd board as RX only, and control them via their USB cables and a laptop, you can get decent receiver threshold values (like -95 dBm at 250 kbps gfsk).

Drove me nutz for a while trying to figure out where the 15 db discrepancy was.