CC2640 CC1310 high and low temperature test

Aguilera

CC2640 CC1310 high and low temperature test [Copy link]

      CC13/26XX is TI's new generation of ultra-low power multi-protocol SOC processors that support Sub1G, 2.4G private protocols, BLE, Zigbee, RF4CE and 6LowPan. CC2640 is a BLE low-power Bluetooth chip, and CC1310 is a SOC that supports wireless products below 1GHz. The datasheet indicates that the supported temperature range is -40 to 85℃, but when doing high and low temperature tests in a laboratory high and low temperature box, running a CW carrier, the frequency deviation seems to be out of range in this temperature range. Can the terminal products actually developed still work stably at -40 to 85℃?

Figure 1

Below we actually test CC2640, run CW carrier, set the center frequency to 2.402GHz, and transmit power to 5dBm; CC2640 is powered on at room temperature of 20℃, heated to 85℃ for 15 minutes, and then cooled to -40℃ for 15 minutes. The whole process is normal; There seems to be no problem with the full temperature range test. Let's further do the following two extreme tests. CC2640 is powered on at room temperature of 20℃, and then heated. At the highest limit of +85℃, the chip is powered off and reset, and then the temperature is gradually reduced to -35℃. The spectrum measured at -35℃ is as follows. We found that the center frequency is offset at this time.

Figure 2

Then the above experiment continues to be reduced to the lowest -40℃. At this moment, the chip is powered off and reset, and then the temperature is gradually increased to 85℃. At 85℃, we found that the center frequency is offset to 2.40072GHz. Through the above two extreme experiments, we found that after resetting at any extreme temperature point, the frequency deviation increases significantly when running to the other end of the temperature, while there is no reset operation in the entire temperature range after powering on at room temperature, and the test is normal. What is the reason? This is because the RF signal clock of CC13/CC26XX comes from the 24M crystal source outside the chip, and the real-time operating system RTOS performs calibration on the external 24M crystal at: Power-on startup, reset or wake-up from Standby, perform calibration PLL calibration After calibration, enter the Tx or Rx state In the process of sending CW continuous carrier, only one frequency calibration is performed at the beginning of power-on, and then the chip is always in the Tx continuous transmission state, and no calibration calibration is performed. Therefore, it shifts with temperature changes, resulting in an increase in frequency deviation. Once there is a reset or exit from the Standby state, calibration will be performed and the frequency will immediately return to normal. Therefore, when the above frequency deviation is large, the reset operation is performed, and the CW frequency point sent will immediately return to normal. Below we also conduct high and low temperature experiments to verify CC1310: non-modulated CW wave, center frequency 433M, maximum power transmission.

Figure 3

1. -20 to 50℃ frequency deviation is relatively normal; 2. The frequency deviation increases from -25 to -40℃. As shown in the table above, at -25℃ the frequency is 433.184Mhz, and the maximum frequency deviation reaches 184KHz. However, after reset, the frequency returns to normal, 432.9994Mhz, 14.8dBm. 3. The frequency deviation increases from 55℃ to 85℃, but it returns to normal after reset, and the frequency is normal. Conclusion: In the laboratory, we can reflect the relationship between frequency deviation and temperature through some extreme tests. However, once the reset operation is performed, the frequency immediately returns to normal, which verifies the importance of calibration execution. In actual products, CC2640 turns on POWER_SAVING. Regardless of whether it is CC2640 or CC1310, the RTOS will automatically enter the Standby mode between two transmissions and receptions. Therefore, calibration will be performed in advance before Tx and Rx. Therefore, there is no frequency deviation problem when the actual product works at -40 to 85℃.

lyfly_away

Thanks for sharing the experimental data

wzc86354120

Thank you for sharing such good experimental data. Thank you

bqgup

Awesome, the test equipment is also awesome

懒猫爱飞

Thank you for sharing. This experiment is very well done and worth learning from.

Ederdan

Thanks for sharing. Is this the signal from the module itself or from the feeder?

2011559106

good