NCR: Transceiver Task Halted UHF Signal Transmission

[AndronikosKostas]

No description provided.

[AndronikosKostas]

UBaseType_t uxHighWaterMark;
uxHighWaterMark = uxTaskGetStackHighWaterMark( NULL );

while(true) {
    uxHighWaterMark = uxTaskGetStackHighWaterMark( NULL );
    vTaskDelay(200);
    LOG_DEBUG << uxHighWaterMark ;

[AndronikosKostas]

I am currently working on the c9c4167 . This variable is printing 1270. This means that are remaining 1270 * 4 bytes = 5080 bytes.

[AndronikosKostas]

4 bytes, since the controller is 32-bit.

[AndronikosKostas]

1st cycle hot FFT

Not proper logging...

1st cycle cold FFT 23/07 11:33

• branch : campaign_eqm_v2
• Mode Sequence : TX then RX
• Findings : TX stopped after 1min
• Notes : There was no delay in transceiverTask...Then we run another FFT (15s RX and then TX) which was successful

2nd cycle hot FFT 23/07 13 : 20 (approximately)

no logging ( I guess that tx was not working properly...)

2nd cycle cold FFT 23/07 16:17

• branch : campaign_eqm_v2
• Mode Sequence : TX then RX
• Findings : TX stopped after 1min

3rd cycle hot FFT 23/07 17:54

• branch : campaign_eqm_v2
• Mode Sequence : TX then RX
• Change : increase of the delay : 300 ms
• Result : Successful with lower data rate

3rd cycle hot FFT 23/07 18:09 v2

• Mode Sequence : TX then RX
• Change :
  • branch : campaign_eqm_v2_debug_for_tx
  • commit "same config with Tartu (without the INA) "
• Findings :
• Notes : successful with low data rate

3rd cycle hot FFT 23/07 18:14 v3

• branch : campaign_eqm_v2_debug_for_tx
• Mode Sequence : TX then RX
• Change :
  • commit : "stack increase and add of the INA task (delay in transceiverTask)"
• Findings : PACKETS SENT : 400
• Notes : successful with low data rate

3rd cycle cold FFT 24/07 11:08

• Mode Sequence : RX->TX
• Change :
  • different branch (campaign_eqm_v2_sw_1)
  • commit : beacon and fft is working only by changing the timings (11:27) , also increased the stack to be sure.
• Findings
  • PACKETS SENT : 400
• Notes : successful with low data rate

4th cycle hot FFT 24/07 12:27

• branch : campaign_eqm_v2_sw_1
• Mode Sequence : RX->TX
• Change : probably nothing
• Findings : Packet that were sent : 400
• Notes : successful with low data rate
  
  WITHOUT DELAY FROM NOW ON

4th cycle cold FFT 24/07 14:55

• branch : campaign_eqm_v2_sw_1
• Mode Sequence : RX->TX
• Change :
  • commit : changed the priorities and remove the delay of the tranceiverTask (commit 2b4ab3c) : TIME 13:27
• Findings : Packet that was sent : 545 packets stopped at 250000 (around 1.5 min)
• Notes : UNSUCCESSFUL with High Data Rate

Differences between 4th cycle cold and 5th cycle hot ... INA was not working properly...

5th cycle hot FFT 24/07 16:55

• branch : campaign_eqm_v2_sw_1
• Mode Sequence : RX->TX
• Change : without delay
• Findings : Packets that were sent : 860
• Notes : SUCCESSFUL with High Data Rate

5th cycle cold FFT 25/07 10:51

• branch : campaign_eqm_v2_sw_1
• Mode Sequence : RX->TX
• Change : commit : go back to the previous priorities and also change the duty cycle of tasks to be more representative
• Findings : Packet that were : 890
• Notes : SUCCESSFUL with High Data Rate.

6th h/c 7th h/c were successful with high data rate , 7th cycle was successful with 5min TX and 5min RX

Why we had issues with 4th cycle cold FFT ?

Priorities on 4th cycle cold FFT (13 : 27)

CANGatekeeper 1
MCUTemperature 1
UARTGateKeeperTask 1
Watchdog 2

Priorities on 5th cycle hot FFT (16:55) "change on priorities but still have crashes" 16:57

CANGatekeeper 2
CanTest 2
Temperature Sensors : 2
MCUTemperature 1
UARTGateKeeperTask 2
Watchdog 3

Conclusion of Analysis and Causes

The issue came on the first cycle where we had changed the transceiverTask on high data rate mode without testing it in the lab. We solved it by halting the transceiverTask for 200ms before sending the next packet with a resulting data rate of 21.84kbps. With the previous change and an increase in the stack of the transceiverTask, we completed three full functional tests with success (1h, 3h, 4h). Also, we successfully applied the higher data rate from the 5th cycle at the hot case after modifying the priorities of the tasks. After that, every FFT till the 7th cycle of the cold case was successful. So total there were another 6 FFTs successful.

[AndronikosKostas]

Issue resolved