BOOSTXL-AOA: Questions about the RF Test DTM (Direct Test Mode) under BOOSTXL-AOA and LAUNCHXL-CC26X2R1

Thanks for your reply!

1. According your message, If performs the AoA Receiver test use this config (BOOSTXL-AOA and LAUNCHXL-CC26X2R1), The test needs to be split in several iterations.
(If during one CTE reception, only the antennas from array 1 or from array 2 will be sued, only need to test 3 antennas at the time.)
But the Bluetooth conformance test system(like.. R&S CMW270 with CMW RUN or Interlab RF Test System) is usually tested at auto mode. It means it cannot modify the connect during the test.

As above condition, if I need to perform all of AoA receiving test using batch (May be included AoA Transmitter test if possible)
What connect method are suggesting?

2. Now i have 3 solution for the AoA Transmitter test as below:
(A). Remove the BOOSTXL-AOA and connect directly the tester to the JSC connector of the CC2642 launchpad.
(B). Identify which antenna is used to send out the CTE (It should be the reference antenna). It depends on the state of the RF switches (Which depends on the state of the DIOs 27, 28, 29 and 30 - use truth table).
(C). Keep the splitter. The "addition" of the signals sent by all the antennas was arriving to the tester. (as only one antenna is sending out data, this is like having only one antenna connected)

I may select the (B) or (C) solution due to I cannot touch the setup between the tests.
But I have some questions about the solution 2and 3:

For Solution (B), I think the RF tester should not get the DIOs state automation?
Or I can use the HCI command to control the RF status?

For Solution (B) & (C) , If use the splitter and addition all of the signals sent to the tester,
It may have some path loss occur on the splitter and each output of antenna output,
How to get the each pathloss each and make up?

And...I have another question need to confirm:

3. Use The HCI Command "HCI_LE_Read_Antenna_Information" tried to get the DUT information as below:

HCI_LE_Read_Antenna_Information:
Send Command: 0x01582000
EUT Request Event: 0x040E0801582000044B4B14

From Event and comparison the Bluetooth Core SPEC ,
It can find the EUT information as below,
I have some question about the information from the HCI event:

Supported_Switching_Sampling_Rates: 0x04
Num_Antennae: 0x4B
Max_Switching_Pattern_Length: 0x4B
Max_CTE_Length: 0x14

(A) Supported_Switching_Sampling_Rates: 0x04 => Bit2
The EUT that support sampling rates is:
1 μs switching and sampling supported for AoA reception
Is correct?

(B) The number of antennae supported by the Controller is 0x4B = 75,
And the Maximum length of antenna switching pattern is 0x4B = 75,
It means the EUT support 75 antenna and the Max switching pattern length is 75?
But the BOOSTXL-AOA is only supporting the 2 antenna,
Have something misunderstood?

4. Following the above question, When i perform the AoA Receiver Test use this config (BOOSTXL-AOA and LAUNCHXL-CC26X2R1) ,
What command should I enter?

Eg: 0x014F200B0001001400020400010000
(Above command may have some issue, just a sample)

(A) What value is suitable should fill in the field of "Expected CTE Length"?
(B) According to the Bluetooth Core SPEC, If the DUT have 2 antennae,
The test switching pattern should be "A0,A1,A0,A0",
And Mandatory supported lengths of antenna switch pattern are "1 to 4"
What value is suitable should fill in the both fields(Switching_Pattern_Length, Antenna_IDs[i])?

Hi,

yeh William said:

As above condition, if I need to perform all of AoA receiving test using batch (May be included AoA Transmitter test if possible)
What connect method are suggesting?

yeh William said:

For Solution (B), I think the RF tester should not get the DIOs state automation?
Or I can use the HCI command to control the RF status?

yeh William said:

For Solution (B) & (C) , If use the splitter and addition all of the signals sent to the tester,
It may have some path loss occur on the splitter and each output of antenna output,
How to get the each pathloss each and make up?

yeh William said:

1 μs switching and sampling supported for AoA reception. Is correct?

Yes and no. The RF switches used on the BOOSTXL-AOA maybe a bit too slow to support 1 us sampling slot. However, the BLE stack and the CC2642's radio do support 1 us sampling slot.

yeh William said:

(B) The number of antennae supported by the Controller is 0x4B = 75,
And the Maximum length of antenna switching pattern is 0x4B = 75,
It means the EUT support 75 antenna and the Max switching pattern length is 75?
But the BOOSTXL-AOA is only supporting the 2 antenna,
Have something misunderstood?

The result you get is a bit unexpected. I would rather expect to see the controller answering it support 4 antennae.
For the rest, I think you made a confusion between "antenna" and "antenna array". The BOOSTXL-AOA has two antenna arrays made of 3 antennas each (i.e. 6 antennas in total). The concept of "antenna array" is not supported by the RF testers. Here nothing prevents you to use 4 antennas at the time for the tests (A1.1, A1.2, A1.3 and A2.1).

yeh William said:

4. Following the above question, When i perform the AoA Receiver Test use this config (BOOSTXL-AOA and LAUNCHXL-CC26X2R1) ,What command should I enter?

It's a bit unclear to me why you are asking this question (the RF tester should send the HCI commands itself).
Anyway, if you are using HCI commands, you don't need to specify the whole switching sequence. The sequence will be directly chosen by the device according to the number of antennae enabled.
So if you want to use 4 antennas, you can use "00:01:02:03" as switching sequence parameter.

I hope this will help,

Best regards,

Thanks for your reply!

Clément said:

I assume you could use an other RF tester able to handle more antennas at the time. An other idea is to use a spliter that can be controlled by the computer running the tests.

At the TI, How to verify the AoA Tx & Rx Test and connect with BOOSTXL-AOA? Which type of the
RF-TESTer be used? Could you provide related information?

Clément said:

As you cannot change the set-up between tests per the constraints you mentioned), I don't think you will use this solution (this solution need to remove the spliter). That being said, when in Tx mode, the CC2642 will always use the same antenna. There is no need for the tester to read the state of the DIOs. (Only you should do this once to know which antenna is used in Tx mode)

I tried to connect every RF output form A1.1~A1.3 and A2.1~A2.3 after entering the HCI Command
"LE Transmitter Test command [v1] and LE Transmitter Test command [V3]" to watch the output
signal, I can watch about -13 dbm at every output port (A1.1~A1.3 and A2.1~A2.3).

This seems to be different from what you mentioned(or i misunderstood?), I cannot find the that
signal only output at one antenna, All of output have a signal.
It seem to the antenna switch is still switching?

Clément said:

Yes and no. The RF switches used on the BOOSTXL-AOA maybe a bit too slow to support 1 us sampling slot. However, the BLE stack and the CC2642's radio do support 1 us sampling slot.

So At this time, This configure is only support 2 us sampling slot?

Clément said:

The result you get is a bit unexpected. I would rather expect to see the controller answering it support 4 antennae.
For the rest, I think you made a confusion between "antenna" and "antenna array". The BOOSTXL-AOA has two antenna arrays made of 3 antennas each (i.e. 6 antennas in total). The concept of "antenna array" is not supported by the RF testers. Here nothing prevents you to use 4 antennas at the time for the tests (A1.1, A1.2, A1.3 and A2.1).

This result of the response (0x4B = 75) is got from the EUT base on latest F/W (*. hex) this
should be a bug? Or The TI have planned to publish more model of the antenna array board in the
future?

I need to clarify something information:
The BOOSTXL-AOA has two antenna arrays made of 3 antennas, each (6 antennas in total).
So, From the point of view of RF testers, The BOOSTXL-AOA has two antennas is right?

If the DUT have 2 antennas, why we need to use 4 antennas at the time for the tests (A1.1, A1.2, A1.3 and A2.1)?

Clément said:

It's a bit unclear to me why you are asking this question (the RF tester should send the HCI commands itself).
Anyway, if you are using HCI commands, you don't need to specify the whole switching sequence. The sequence will be directly chosen by the device according to the number of antennae enabled.
So if you want to use 4 antennas, you can use "00:01:02:03" as switching sequence parameter.

Due to having got different result (HCI Command) when my recent try 2 brand of Bluetooth RF tester (R&S and Anritsu), So I want to confirm the correctly HCI command and the suitable HCI command for the BOOSTXL-AOA and LAUNCHXL-CC26X2R1 use for the CTE TX/RX test and AoA Test.

And, The LPRF BLE HCITester have plans to support the CTE or AoA Test command?

processors.wiki.ti.com/.../LPRF_BLE_HCITester

Hi,

yeh William said:

At the TI, How to verify the AoA Tx & Rx Test and connect with BOOSTXL-AOA? Which type of the RF-TESTer be used? Could you provide related information?

We use a tester able to drive only four antennae. Please discuss this point further with your local FAE, he may have more details for you

yeh William said:

I tried to connect every RF output form A1.1~A1.3 and A2.1~A2.3 after entering the HCI Command
"LE Transmitter Test command [v1] and LE Transmitter Test command [V3]" to watch the output
signal, I can watch about -13 dbm at every output port (A1.1~A1.3 and A2.1~A2.3).

This is kind of unexpected. Could you please verify if the antennas are switching while you send the CTE? To do so, you need to attach a logic analyzer to the pins DIO27, DIO28, DIO29 an DIO30. I expect to not see any activity on these pins. (Note: this is for debug only, latter you won't need this logic analyzer)

yeh William said:

So At this time, This configure is only support 2 us sampling slot?

Correct. If you want to support 1us, you need a different antenna board.

yeh William said:

This result of the response (0x4B = 75) is got from the EUT base on latest F/W (*. hex) this
should be a bug? Or The TI have planned to publish more model of the antenna array board in the
future?

This is a bug, I have opened an internal ticket to solve this. Thank you for noticing it.

yeh William said:

If the DUT have 2 antennas, why we need to use 4 antennas at the time for the tests (A1.1, A1.2, A1.3 and A2.1)?

The DUT has 6 antennas. (the fact the software then consider it is two antenna arrays has no effect on the RF tester - the 6 antennas have to tested as 6 antennas)

yeh William said:

Due to having got different result (HCI Command) when my recent try 2 brand of Bluetooth RF tester (R&S and Anritsu), So I want to confirm the correctly HCI command and the suitable HCI command for the BOOSTXL-AOA and LAUNCHXL-CC26X2R1 use for the CTE TX/RX test and AoA Test.

I confirm my previous answer, the antenna ids have to be specified only once. The device will then chose the proper switching sequence. That being said, the device should still use the proper sequence if you specify the full sequence.

yeh William said:

And, The LPRF BLE HCITester have plans to support the CTE or AoA Test command?

HCI Tester already support the AoA test commands. If needed, here is the updated XML file: Tester_HCILib_ble5stack_2_01_04_00.xml

If it helps, here is a the small script I wrote while I answered your questions: 

Send_HCI_Reset
Wait_HCI_Command_Complete_Reset_Event 5000, any, HCI_Reset, 0x00

Send_HCI_LE_Set_Event_Mask 0xffffffffffffffff
Wait_HCI_Command_Complete_LE_Set_Event_Mask_Event 5000, any, HCI_LE_Set_Event_Mask, 0x00

#Read Antenna information 
Send_HCI_LE_ReadAntennaInformationCmd
Wait_HCI_Command_Complete_LE_ReadAntennaInformationCmd_Event 5000, any, HCI_LE_ReadAntennaInformationCmd, 0x00, 0x04, Any, Any, Any

# RX test - Need to receive a CTE to have switching occuring
Send_HCI_LE_Enhanced_Cte_Receiver_Test 0x13, 0x01, 0x01, 20, 0x00, 0x02, "00:01:02:03"
Wait_HCI_Command_Complete_LE_Enhanced_Cte_Receiver_Test_Event , any, HCI_LE_Enhanced_Cte_Receiver_Test, 0x00

Best regards,

Clément said:

This is kind of unexpected. Could you please verify if the antennas are switching while you send the CTE? To do so, you need to attach a logic analyzer to the pins DIO27, DIO28, DIO29 an DIO30. I expect to not see any activity on these pins. (Note: this is for debug only, latter you won't need this logic analyzer)

We have not a logic analyzer, But we use the Oscilloscope to monitor the DIO27, DIO28, DIO29 an DIO30. Have only found the High signal on the DIO27 and there are not found the signal at DIO28, DIO29 and DIO30 after entering the HCI Command "LE Transmitter Test command [v1] and LE Transmitter Test command [V3]".

What action of the switch on the antenna array board (BOOSTXL-AOA) when perform the CTE TX Test HCI command or Standby state (connect power, but not perform any command) in this Host test firmware version(*.Hex)?

Clément said:

I confirm my previous answer, the antenna ids have to be specified only once. The device will then chose the proper switching sequence. That being said, the device should still use the proper sequence if you specify the full sequence.

I am not very familiar about the antenna ids and switching sequence of mode operation (eg. Test Switch Pattern), The TI have any document could be refereed?

Clément said:

HCI Tester already support the AoA test commands. If needed, here is the updated XML file: (Please visit the site to view this file)

If it helps, here is a the small script I wrote while I answered your questions: (Please visit the site to view this file) Please let me know if you encounter some difficulties with HCI tester.

Thanks for your provide, The both XML and TxT file are working normally on the HCI Tester.

If I want to get the latest version of the XML file in the future, Where can I find on the TI web site?

Hi,

Based on the state of the DIOs 27, 28, 29 and 30 and the truth table here, you can know which antenna is used. Once again, no antenna switching happens except during the reception of the CTE. In the out-of-the box software, you should have DIOs 27 and 28 set to high and 29 and 30 set to low (i.e. the antenna A1.1 is used when no receiving a CTE).

The switching pattern for test mode is given in the documents you mentioned in your previous message. Here are some details about the HCI_LE_EnhancedCteRxTestCmd() function.

yeh William said:

If I want to get the latest version of the XML file in the future, Where can I find on the TI web site?

Here :)

Bet regards,