CC256XB-BT-SP: CC2564B Power Control Step Size and Maximum Power

Here is the log file. It contains the activities for the continuous transmission test - starting from the power level 1, up to power level 8

Let me know if you require more information.

BT_CONT_TX_TEST.rar

Thanks,
Chetan

Hi Chetan,

I second Brad, the logs do not cover the part where the power levels were actually written to the CC256x.

Capturing the log starting before the device initializes will capture all the info we need. With that info, we can look into what is going wrong.

Best regards,
Vihang

Hi all,

Thank you for your responses. Here is the full log. The last part of the log should be the continuous transmission test.

Brad, at this moment, it is unclear that if the chip is taking the adjusted power levels or not but EDR2 ( and EDR3) schemes are following the maximum power level of ~2dBm so  I wouldn't say for sure that the new levels are not in effect. 

We are going to perform tests with all 8 power levels increased by 3dBm and 6dBm. The purpose of  these tests would be to see if the chip took the new levels or not and if it behaves the same or not. I will let you guys know the results soon.

BT_CONT_TX_TEST_FULL_LOG.rar

Thanks,
Chetan 

Please see the results for the new tests performed with the 3dBm and 6dBm increased power levels.

PREVIOUS TEST POWERS: (TEST1)
Schemes:                              : { GFSK,         EDR2,      EDR3   }
Maximum power levels      : { 3dBm,         2dBm,      2dBm    }
Minimum power levels for : { -32dBm,  -33dBm,    -33dBm  }

NEW TEST (TEST2):

- All 8 powers incremented by 3dBm (with respect to TEST1)
Schemes:                             : { GFSK,         EDR2,      EDR3   }
Maximum power levels     : { 6dBm,         5dBm,      5dBm    }
Minimum power levels      : { -29dBm,  -30dBm,    -30dBm  }

NEW TEST (TEST3):

- All 8 powers incremented by 6dBm (with respect to TEST1)
Schemes:                             : { GFSK,         EDR2,       EDR3   }
Maximum power levels     : { 9dBm,         8dBm,       8dBm    }
Minimum power levels      : { -26dBm,   -27dBm,    -27dBm  }

RESULTS

Alrighty. Thank you for the response. I checked the RF calibration command and we are currently using calibration procedure value (CPV for short) of 0x00000001. The example at the link you sent me uses CPV of 0x00000800. I changed the CPV  to new value and the results no longer violate the step size specifications. I do see all 8 power levels values as expected except that observed average power levels are decremented by 3dBm ( but it could be our equipment calibration and losses calculations issues ) which is consistent throughout the tests for 3dB and 6dB incremented power levels.

Now, coming back to those CPVs, TI wIkipage 'CC256x VS HCI Commands' does not contain any information on what how this value affect the calibration.

In 'CC256x VS HCI Commands' wikipage revision as of 10:56, 11 December 2015 , there are two calibration commands used after updating the power vectors in the example shown and the first RF calibration command uses CPV of 0x00000001. Here are the commands for the reference.

#1: Run TPC (Transmission Power Control) calibration

Send_HCI_VS_DRPb_Enable_RF_Calibration 0xFD80, 0x00, 0x00000001, 0x01
Wait_HCI_Command_Complete_VS_DRPb_Enable_RF_Calibration_Event 5000, 0x00, 0xfd80, 0x00

#2: Run TPC (Transmission Power Control) calibration

Send_HCI_VS_DRPb_Enable_RF_Calibration 0xFD80, 0x3c, 0x00005ff0, 0x00
Wait_HCI_Command_Complete_VS_DRPb_Enable_RF_Calibration_Event 5000, 0x00, 0xfd80, 0x00

Later, Revision as of 13:21, 22 December 2015, the example contains only one calibration command with CPV of 0x00000800. Here is the command for the reference.

#Run internal calibration

Send_HCI_VS_DRPb_Enable_RF_Calibration 0xFD80, 0x00, 0x00000800, 0x01  

Wait_HCI_Command_Complete_VS_DRPb_Enable_RF_Calibration_Event 5000, 0x00, 0xfd80, 0x00

Questions:

1. I'd like to know that how these two different CPVs affect the calibration procedure ? If the chip gets an incorrect CPV, does it perfrom incorrect calibration or  no calibration at all ?

2. The tests with the invalid step size results utilized CPV of 0x00000001 with the stack version 1.4. Using the same CPV with stack version 1.6 yielded the same results. Does this mean that CPV of 0x00000001 was for older stakc version and is no longer supported in service pack 1.4 or 1.6 ?

Please let me know !

I really appreciate your help in tackling his issue.

Thanks,
Chetan

This comment is just a place holder.

Chetan Patel46 said:

For now, as per our observations, I can infer that CPV does depend on the BT stack version (1.4 and 1.6 do not work with 0x00000001 but 0x00000800).

No, this is not related to the firmware revision.  These have a fixed meaning.

<Values name="RF_Calibration_DRPb_Bitmap" type="mask">
<Value val="0x00000001">Pre RF init</Value>
<Value val="0x00000008">Clock Dependant Calc</Value>
<Value val="0x00000010">Temperature recognition</Value>
<Value val="0x00000020">DC</Value>
<Value val="0x00000040">LDO</Value>
<Value val="0x00000080">DCO Current Optimization</Value>
<Value val="0x00000100">DCO Coarse Open Loop</Value>
<Value val="0x00000200">KDCO Nominal Current</Value>
<Value val="0x00000400">PPA LDO Current</Value>
<Value val="0x00000800">TPC</Value>
<Value val="0x00001000">IFA Pole Location</Value>
<Value val="0x00002000">LPS</Value>
<Value val="0x00004000">Wide Band RSSI TH</Value>
<Value val="0x00008000">IQ MM</Value>
<Value val="0x80000000">Don't Change</Value>
</Values>

The value 0x800 relates to TPC (Transmission Power Control).  The comment in the wiki page above that command has been updated accordingly.  FYI, the initscript is already sending the following commands and it is not necessary to send them again:

Send_HCI_VS_DRPb_Enable_RF_Calibration 0xFD80, 0x00, 0x00000001, 0x01

Wait_HCI_Command_Complete_VS_DRPb_Enable_RF_Calibration_Event 5000, 0x00, 0xfd80, 0x00

Send_HCI_VS_DRPb_Enable_RF_Calibration 0xFD80, 0x3c, 0x00005ff0, 0x00

Wait_HCI_Command_Complete_VS_DRPb_Enable_RF_Calibration_Event 5000, 0x00, 0xfd80, 0x00

That helps. Thank you for your support.

Chetan