What difference between Connectivty_test and HCI black box FW?

[Annie_he]

Hi NXP,

   We are testing SRRC for kw45 bluetooth boards and it designed by us.

   When we test RF test, is accour different result by using different FW, connectivity_test and HCI black box. 

   like below:

   When test Frenquency range, signal send by LE_transmitter_cmd is lower than connectivty test cmd

send by LE_transmitter_cmd：01 1E 20 03 00 25 00, result=-36.953dBm/MHz:

Annie_he_1-1714103099054.png

send by connectivty test cmd："Mode TX, Rate BLE 1Mbps, Whiten Fixed, Channel 42, Power 31, Payload 37, XtalTrim 26> Mod PN", result=-39.23dBm/MHz:

Annie_he_2-1714103661278.png

     It is higher 2dB in connectivity test.  Why and how it can occured? and we use the same sounding only different FW.

    Here is we use:

     SDK example: KW45B14ZEVK 2.12.7

     FW: connectivity test / HCI black box from, which only I change is Xtaltrim=26.

     MXUXpresso IDE: V11.8.0_1165

   If you have any ideas about the commands and how to pass certification, please let us know, thanks.

   Best Regards,

   Annie

[Christine_Li]

Hi, @Annie_he 

These two is built for different examples. For the details inside of these two FWs, I need to confirm with our internal team.

But before that, may I ask what kind of issue do you meet during your certification?

Which FW can pass the certification? and what is the standard of the certification?

Best regards,

Christine.

[Annie_he]

Hi Christine,

    Thanks for your reply.

    Now we are going to get pass the China Ble standard named SRRC. And the issue I meet in certification is 2M-Frequency range fail.

    When I use Connetivity_test FW, I get the 2M-Frequency range result as below, it is so close to the limit.The Frenquency of Power 30dBm/100Khz is 2.40000GHz=2.4GHz. We do not have enough value to pass the certification:

cmd: 

Mode TX, Rate      BLE 2Mbps, Whiten Fixed, Channel 42, Power 31, Payload 37, XtalTrim 26> Mod PN

Annie_he_3-1714350079258.png

When I use HCI black box FW, I get the result as below, Power 30dBm/100Khz is 2.40065GHz>2.4GHz, Pass :

cmd: 01 34 20 04 00 25 00 02

Annie_he_0-1714349866861.png

And also, The 2M-suprious emission of 2390-2400 is so close to the limit as below Picture:

Annie_he_5-1714350555441.png

For connectivity test FW, Power in 2.4GHZ is -30.68dBm, we have no more margin to pass:

cmd: Mode TX, Rate      BLE 2Mbps, Whiten Fixed, Channel 42, Power 31, Payload 37, XtalTrim 26> Mod PN

Annie_he_4-1714350440659.png

For HCI black box FW, Power in 2.4GHZ is -33.750dBm, we still have 3dB margin to pass:

cmd: 01 34 20 04 00 25 00 02

Annie_he_6-1714350734284.png

Our Surrounding:

SDK: KW45 V2.12.7

IDE: V11.8.0_1165

FW: connectivity_test &HCI black box

     thanks

     Best Regards,

     Annie

[Annie_he]

And, This Afternoon, I try to change Connectivity test cmd and it can get a better result.

Cmd- Burst Transmission using random-payload packets:

Mode RX, Rate BLE 2Mbps, Whiten Fixed, Channel 42, Power 31, Payload 37, XtalTrim 26>

Packet Mode PRBS started. Press [SPACE] to stop test

and 2M-Frenquency Range result as below, there is 70KHz margin:

Annie_he_0-1714362769766.png

and 2M-Suprous Emisson in 2390-2400Mhz result as below, there is 6dBm margin:

Annie_he_1-1714362841019.png

And if we can use Burst Transmission using random-payload packets cmd to test certification? What different in Burst Transmission using random-payload packets cmd and Continuous Modulated Transmission pseudo-random cmd and HCI_LE cmd? 

connectivity test FW cmd as below:

Annie_he_2-1714362996509.png

If you have any ideas about it, please let us know. Thanks.

[Christine_Li]

Hi, @Annie_he 

Thanks for your reply and your efforts on this issue.

Sorry, I am not good at this Power related to certification,. I need to check with our internal team.

But as you know, we will have a public holiday from May 1st to May 5th in China, totally 5 days, and I will be back to office 6th of May.

I could not reply you in time during this period. I sincerely apologize for the inconvenience.

Have a nice holiday~

Best regards,

Christine.

[Christine_Li]

Hi, @Annie_he 

I found this guide in our SDK:

Generic FSK Link Layer Quick Start Guide.pdf 

which is located in SDK_2_12_7_KW45B41Z-EVK\docs\wireless\GENFSK\ path.

Please see section:  4.3.1 Continuous tests 

So I think we just provide some examples and different modes to customer to choose according customer's preference. About SRRC certification, I think you can choose any modes from them. 

Do you know SSRC has any special rules to limit the test modes?

If there is no such kind of limitations, then you can pass it with Cmd- Burst Transmission using random-payload packets.

Best regards,

Christine.

[Annie_he]

Hi Christine,

   SRRC may request the duty cycle, but I did not get the specific limit about it. Duty cycle of an old SRRC report in 2008 is demanded more than 10%.

   I try to send cmd-Burst Transmission using random-payload packets to DUT, it pass. But its duty cycle is less than 10%.

   So if you have better Ideas about below:

1. HCI black box can support HCIExt_ModemTestTxCmd & HCIExt_EnhancedModemTestTxCmd
HCIExt_ModemTestTxCmd:

1Mbps 2402MHz UNMoudulation： Send->01 08 FC 02 01 00

recieve>>04 FF 05 08 04 00 08 FC

HCIExt_EnhancedModemTestTxCmd：

1Mbps 2402MHz UNMoudulation： Send->01 27 FC 03 01 00 00

>>04 FF 05 27 04 00 27 FC

those cmd we got From other chip platform, if you have similar solution about unmodulation signal in KW45?

2. How to get 2M-Phy Frenquency Range in 2400 lower down?

We really need your help, if you have any ideas, please share us, Thank you so much.

Best regards,

annie

[Christine_Li]

Hi, @Annie_he 

Thanks for your reply.

1.For the unmodulated signal in KW45, can you please help to try with mode [6]?

Christine_Li_0-1715150291522.png

2. How to get 2M-Phy Frenquency Range in 2400 lower down?

==>Sorry, can you give me more details for better understanding? I didn't get your point.

Best regards,

Christine.

[Christine_Li]

Hi, @Annie_he 

I haven't received your feedback in 5 days, do my above comments help you?

Please let me know whether still need our support on this thread.

Best regards,

Christine.

[Annie_he]

Hi Christine,

    Sorry, I'm late for your reponse.

    What I talk may not so clear.

1. I hope to use HCI black box image to send unmodulation signal, not connectivity test image to send.

2.For sencond question, you can know taht 2M-Frenquency range of KW45 fail in China SRRC test. Like below picture, I send 2M signal in 2402Mhz by connectivity test, and its Frenquency Range is more than 30dBm in 2400Mhz. SRRC request in 2400Mhz, power need less than 30dBm.

Annie_he_1-1715761908187.png

But HCI black box image to test 2M Freq Range, it is easily to pass this SRRC item. But it can not sopport Ummodulation signal.

We need a image which can pass SRRC and send unmodulation signal&modulatuion signal. And if you have better idea?

And sorry for late again and Thank you so much for your reply.

Best Regards,

Annie

[Christine_Li]

Hi, @Annie_he 

Sorry for my late reply.

Do you have our KW45-EVK? Have you tested with our KW45-EVK?

I am suspect whether it is a hardware issue.

I am not so familiar with hardware, if you have our KW45-EVK, please help to test with it and let me know the result. At the same time, I will create an internal ticket to our expertise team to ask help.

Best regards,

Christine.

[Annie_he]

Hi Christine_Li,

    We have KW45-EVK, and I will use KW45-EVK to test its 2M Freq Range with Connectivity test image this afternoon or tomorrow and give a feedback for you. 

    Thanks for your help and reply.

Best Regards,

Annie

[Christine_Li]

Hi, @Annie_he 

Thanks for your reply.

Sure, please take your time.

And also I have created an internal case to our expertise team. Hope we can resolve this issue ASAP.

If I need any other info from your side, I will leave message to you.

Best regards,

Christine.

[Annie_he]

Hi Christine_Li,

     We got the 2M-Freq Range from KW45- EVK. It also behaves less margin in 2400MHz, Like below pictrue, and its power is about 9.8dBm in 2402MHz:

Annie_he_0-1716281012907.png

The test Way we get from ETSI EN 300 328 V1.7.1 (2006-10) as the attach file.

Now we have tunned a matching and the result in 2M-Freq Range is similiar to KW45-EVK.

If you have better ideas about the test Way or how to get more margin, please talk to us. Thank you very much.

Best Regards,

Annie

[Christine_Li]

Hi, @Annie_he 

Thanks for your results.

I have forwarded the test result on both your board and our KW45-EVK to our expertise team.

Once receive any feedback, I will share with you.

I apologize again for my poor RF knowledge.

Best regards,

Christine.

[Annie_he]

Hi Christine_Li,

    Thanks you so much. You take much of you time to help us.

    You can check the test way of Freq Range and if you have any ideas about the 2M Freq Range, you can share with us.

    And if you have any updates, you can send mail to me or leave massege in this post. My Email is [email protected].

Best Regards,

Annie

[Christine_Li]

Hi, @Annie_he 

It's my pleasure!  You are welcome anytime.

I receive our internal team's suggestion, can you please have a try and let me know your result?

Could you please try the transmitter test command with the pattern of all 0 bits or 1 bits? Thank you.

Christine_Li_0-1716369100990.png

Best regards,

Christine.

[Christine_Li]

Hi, @Annie_he 

Did you get any chance to test with our internal team's suggestion?

Please let me know whether it works for you.

Best regards,

Christine.

[Annie_he]

Hi Christine_Li,

    I have tried your suggestion. But the signal is not what we want. it behaves like umodulation signal.

Continuous Modulated Transmission 1's: 

1.png

Continuous Modulated Transmission 0's:

2.png

Best Regards,

Annie

[Christine_Li]

Hi, @Annie_he 

Thanks for your reply.

I have forwarded your result to our internal team. Hope we can get more suggestions to support you.

Best regards,

Christine.

[Christine_Li]

Hi, @Annie_he 

Thanks for your patience.

I received our internal team's reply as below:

========

 Yes, the HCI commands can not generate the pure continuous wave. The frequency of transmitting 1 or 0 has a positive or negative deviation than the pure continuous wave. We can consider it as an unmodulated signal while ignoring the frequency.

=========

Please let me know anything else I can do for you. If you still have any other concerns, please do not hesitate to reply me.

Best regards,

Christine.