Hi,
We are testing the CC1190 RF frontend for use in our application, we've conducted a burn in and have experienced some failures or unexplained changes we'd like to understand.
Following is a full description of our burn in conditions:
1. All testing was performed at Vcc = 3.4V
2. Regular burn in at 70 degrees centigrade.
a. the DUT was connected to CC1200 EVB through a 10dB attenuator so the maximum Pin in transmit was +4dBm.
b. The pair was controlled by the TI evaluation software through TI SmartRF TrxEB. The TX duty cycle was controlled by the evaluation software. The CC1190 configuration was done by toggle switches. High-Gain mode was used throughout.
c. Two units pairs were run one against the other in the oven, one serving as the transmitter and the other as the receiver. Total attenuation between the pairs (the path loss) was 70dB.
d. The functionality was reversed every 7 days.
e. Total test time was 40 days.
3. After the burn-in one of the two CC1190 units displayed decreased current consumption (approx. 10%) at room temperature as compared to the before burn-in tests. The RF gain of that unit decreased by about 0.3dB over the lower part of the frequency range only. The other unit's results were unchanged.
4. The suspect CC1190 unit was run again for 10 days at 70 degrees centigrade in CW TX mode with +8dBm Pin which is below the published AMR for this part. The unit was loaded by a 2W 50ohm dummy load.
5. After this test, the RF performance at room temperature deteriorated rapidly this could be seen changing within minutes during the test untill it settled on about 10dB gain.
6. Additional Burn-in test was performed as follows:
a. Five new units were burned-in at +5dBm 870MHz CW input, High Gain, PA mode and 2W 50ohm dummy loads at 70 degrees centigrade.
b. Two additional new units were similarly tested with +8dBm 870MHz CW input.
c. After 12 days, the 7 units were re-tested at room temperature.
d. one of the five +5dBm units displayed 5dB reduction in PA High Gain. It's 1dB compression point output power decreased from +25dBm to 22dBm approximately. The LNA gain changed acceptably by 0.2dB to 0.3dB.
e. after testing it a little further with 5dBm input power, it deteriorated to about 10dB gain exactly like the previous fauly PA.
f. The other six units displayed very small and acceptable changes in PA gain and 1Db compression and in LNA gain (0.2dB to 0.3dB).
What could cause this behaviour?
Thanks in advance,
Eli