CC2642R：Output test using host test

Hi Hiroki,

What kind of BT tester are you using?

Are you able to test with the latest SDK, 3.30, as well?

Best regards,

Vetle

Hi Vetle

I use Anritsu MT8852B when RF testing.

I did it in SDK3.3, but it was the same as SDK3.2.

Hiroki

Hi Hiroki,

SDK 3.10 used incorrect power settings for the BLE Stack which is why the output is 1 dB lower than expected. This was fixed in SDK 3.20 (and later).

Regards,
Fredrik

Hi Fredrik

Thank you for your reply

I understand that the power setting has been changed since SDK3.2.

When I run the RF test, the result is that SDK3.1 has higher output than SDK3.2 and later.

Is this result correct?

Is the expected result the opposite?

The difference is 1dB as I mentioned in the first question.

Regards,

Hiroki

Hi Hiroki,

I would expect SDK 3.10 to give about 1 dB lower Tx power than 3.20 and onwards. Which power setting are you measuring at, and how do you perform the measurement?

Regards,
Fredrik

You should also compare with latest version of SmartRF Studio.

Regards,
Fredrik

Hi Fredrik

I test output power according to Bluetooth RF Test Specification RF-PHY. TS. 5.1.0 using Anritu MT8852B.

The result measured with Anritu MT8852B is that the output power of sdk3.1 is higher than sdk3.2.

Regards,

Hiroki

HI Fredrik

I used RFstudio and measured the output (Continouous TX Unmodulate) of sdk3.1 and sdk3.2.

But it was the same output.

Regard,

Hiroki

Hi Fredrik

I am measuring the output power (TP / TRM-LE / CA / BV-01-C) of sdk3.1, sdk3.2, and sdk3.3 with an Anritsu tester.

As a result, an output difference of 1dB appears between sdk3.1 and sdk3.2.

Is it different for each SDK because the power setting is done in the host test project?

In the future, I will use SDK 3.2 or later, but I understand why this difference appears.

Regards,

Hiroki

Hi Hiroki,

My question was: What output power do you actually measure, what is the level?

The power settings are found in the ble_user_config.c file. 

Fredrik

Hi Fredrik

This is the output power result I measured with an Anritsu tester.

And DEFAULT_TX_POWER found in ble_user_config is changed, the result don't change.

DEFAULT_TX_POWER is the initial state (0dBm).

Why is the value high even though it is set to 0dBm?

That is interesting, I was pretty sure it was going to be the other way around.

When calling the HCI Tx Power test command used by the Bluetooth tester, the device will always output the highest available output power, no matter what the default level is set to.

Can you show me the Tx power tables you have in ble_user_config.c for both SDKs?

This is the Tx power table for each SDK's ble_user_config.c.

ble_user_config.c.pdf

I use ble_user_config.c in the initial state

Hi Fredrik

I would appreciate it if you would update me on this matter.

Is it under investigation?

Regard Hiroki

Hi Fredrik

Are you still investigating?

I need some information on this issue.

Regards

Hiroki

Hi Hiroki. 

What HCI Command are you using to measure the output power? Unless you specify the output power with the HCI_EXT_SetTxPowerCmd command, the HCI will output at 0 dBm. As Fredrik said, the 3.10.00.53 SDK got the wrong PA table, but the 3.20 and out should have the correct value. 

Regards,
Vegard

Vegard,

BT Testers will use HCI_EXT_EnhancedModemTestTxCmd. This command will transmit at the highest available power setting in the device, irrespective of what the default level is defined as. 

From the reported results, it seems likely that the device is at the 5 dBm power setting. What is strange is that 3.10 is transmitting at a higher level than 3.20. I would expect the opposite.

Fredrik

Hi Fredrik.

The PA table value of 0x941E translates to a PA table entry of RF_TxPowerTable_DefaultPAEntry(30, 0, 0,  74). This is higher than the PA setting value used in the 3.20 and later SDKs and should roughly translate to around 6 dBm. This explains the 1 dBm difference Hiroki is seeing.

Regards, Vegard

Aha, I thought 3.10 (and earlier) settings were lower than 3.20 (and later). It is anyway 3.20 settings that are correct and should be used. 

Hi Vegar

I checked the settings in the code.
In sdk3.1, it is set by RF_TxPowerTable_DefaultPAEntry (30, 0, 0, 74) as you reported. In sdk3.2, it is set with RF_TxPowerTable_DefaultPAEntry (23, 0, 0, 57), which means that this is a difference of 1dB. Does this mean that TI confirmed the problem with sdk3.1 and fixed it with sdk3.2?

Please answer this question and my problem will be solved.

Regards,Hiroki

Yes, that is correct.

Regards,  
Vegard