S2-LP Custom Board - Need help

[Silas Valera]

Posted on January 07, 2018 at 03:35

Hi

I have my custom RF board using S2-LP and balunBALF-SPI2-01D3 (

/external-link.jspa?url=http%3A%2F%2Fwww.st.com%2Fcontent%2Fccc%2Fresource%2Ftechnical%2Fdocument%2Fdatasheet%2Fgroup3%2Fe9%2Fff%2Fcd%2F1e%2F4f%2Fdd%2F49%2Fb4%2FDM00417488%2Ffiles%2FDM004174pdf%2Fjcr%3Acontent%2Ftranslations%2Fen.DM004174pdf

) and my custom board follows the evaluation design given in the balun data sheet. I am having problem with it:

The host controller communicates fine with S2-LP over SPI (I can see the expected signals on oscilloscope for SCL, SDI, SDI and CS). Also SMPS 1 and 2 (pins 2 and 3 of S2-LP) show expected waveforms. But there is not signal on TX pin. Can anyone help me please?

Thanks

Silas Valera

#868-mhz #custom-board #s2-lp #balf-spi2-01d3 Note: this post was migrated and contained many threaded conversations, some content may be missing.

[Igor Orlovich]

Posted on February 12, 2018 at 17:38

Did you actually configure it to go to CW mode? You need to configure radio settings, data source (PN9), modulation (or not) and go to TX mode before you'll see anything out of it.

[Silas Valera]

Posted on February 13, 2018 at 04:39

Hi Igor, 

thanks for your suggestions:) 

I am actually using the same code that is used in S2-LP DK 1.2.0 - SDK_Basic_Generic_A.c but only adopted for my microcontroller. I had actually tested my code using my controller and the S2-LP test board 

STEVAL-FKI868V1 - the eval board comes with STM32L1 Nucleo. I had tested this S2-LP board with my controller and it works fine. So I believe the code is fine.  I am trying to see if there is any hardware issue. 

Thanks again!

Silas

[Winfred LU]

Posted on February 13, 2018 at 06:46

Hi Silas,

If it is a HW issue, it will be worth to check 

http://www.st.com/resource/en/schematic_pack/steval-fki868v1_schematic.pdf

and please doubly check if everything is connected or soldered properly.

Best Regards,

Winfred

[Silas Valera]

Posted on March 13, 2018 at 02:39

Hi 

I did some redesign of the custom RF board, all capacitors in 0402 package, also using the 

BALF-SPI2-01D3 balun for 868Mhz. 

Now I have a different issue. With my custom microcontroller board, when I connect the 

STEVAL-FKI868V1 board using SCL, SDI, SDO, SDN, CS and GPIO3, the board transmitts fine (I am using Basic_Generic_A TX code) and on my other STEVAL-FKI868V1 kit (with nucleo) it receives fine and decodes the payload perfectly. But when I just replace STEVAL-FKI868V1 board with my custom RF board with my microcontroller board, I see that it transmits the signal but on the receiver STEVAL-FKI868V1 kit, it does not decode. I have attached the pic of RSSI with this. All other conditions remain the same (same modulation, data rate, bandwidth etc).

Where should I look? The balun could be the reason? Seems all the components are soldered as expected. Any help or suggestion?

Thanks

Silas

 

[Silas Valera]

Posted on March 18, 2018 at 01:47

Hi Igor,

Here is some more findings after I did some more test using CW mode

Coded for CW mode (no modulation) : Running the code for CW with S2-LP kit board with my controller board - and looking at the spectrum analyzer - it shows nice peak of -30dBm at 868Mhz, I have included the pic below. And then replacing the kit board with my custom RF board, I see no such peak for 868Mhz, but kind of spread, seems lot of harmonics, also included the pic below

Can this behavior be because of the balun 

BALF-SPI2-01D3 mismatch? Seems like there is no selectivity for the 868Mhz.  Can anyone help on this please?  Should I update my design to use the discrete components instead of the balun? The only benefit of using the balun is reduced component count and smaller PCB. I would really appreciate any of your feedback.. 

Thanks

Silas

[Winfred LU]

Posted on March 19, 2018 at 03:23

Hi Silas,

BALF-SPI2-01D3 is designed to match S2-LP 868MHz and 915MHz.

Some thoughts but i am not sure how much i can help:

Could you dump all the registers to doubly check if the custom RF board is configured exactly the same as EVAL board?

Is it possible to replace the receiver device (instead of the transmitting one) to your custom RF board?

Just to see if the custom RF board can receive well then the TX circuit will need to be checked.

Maybe it will be better to provide schematics and the board for some local FAE's reviewing.

Best Regards,

Winfred

[Silas Valera]

Posted on March 21, 2018 at 04:01

Hi Winfred,

Thanks for your replay and thoughts!

I have two S2LP eval kits - both for 868Mhz, So for one kit I replace the S2LP board with my custom board.Both configured the same as 2GFSK1/868Mhz freq/40Kbps datarate/20Khz freq dev/100Khz filter - I found the following:

1. My custom board as TX and kit as RX - no reception on rx side

2. My custom board as RX and kit as TX - no reception on my board side

3. My custom board configured for CW 10dbM - on kit side RSSI shows about -60dbM line

4. My custom board configured for CW / Max power - on kit side RSSI shows about -40dbM line

I have also attached the dump of registers from both my custom board and kit board. Seems only one value is different which is SYNT0 register, my board has 0x1F and kit board has 0x2F. I tried writing 0x2F to my custom board and running test, but no reception. Not sure why SYNT0 is different for my board, my board also has 50Mhz crystal. Let me know what you think..

I have also redesigned my custom board with discrete balun and sent out for fabrication. Should come back shortly. Meanwhile I will try to work on this version with BALF balun.

Thanks again for your valuable help!!

Silas

________________ Attachments : Register_Setting_Kit.xml.zip : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006Hxr0&d=%2Fa%2F0X0000000b1d%2FeZPfyZVSdc0u1hwkCJ_mQnXXP6MMvcCQXmqZVvoYQqQ&asPdf=falseRegister_Setting_Custom.xml.zip : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006Hxs7&d=%2Fa%2F0X0000000b1c%2FlgF_U7LUh.zUBHuUJRp1Z_7Rc7gDSteoNuXhgf3zihk&asPdf=false

[Winfred LU]

Posted on March 21, 2018 at 06:39

Hi Silas,

Both STEVAL-FKI868V1 and STEVAL-FKI868V2 use 50 MHz crystal, so all the registers need to be configured the same for the comparison tests.

As far as i know, SYNT0 needs to be the same too since SYNT matters for base frequency setting,

When you configure an EVB and the custom board with the same register settings, both S2-LPs should work the same, however, in your tests they didn't.

This seems to make it a hardware defeat. Is it possible to find someone help to check the circuit?

Using baluns helps reduce number of components along with the board space reduction.

The RF circuit becomes simpler. And it should be easier to debug if there is something wrong with the circuit. 

So maybe it is a good idea to try balun version.

Sorry for not helping too much.

Best Regards,

Winfred

[Silas Valera]

Posted on March 23, 2018 at 03:18

Hi Winfred

Thank you so much for your reply!I further tried the following

1. I put the S2LP eval board back on the STM conotroller board which was previously having my custom board, this kit read 0x1F for SYNT0 on COM13. The other kit is on COM12. And then when I re-initialized both kits, I found that the one on COM13 with S2LP eval board still read the 0x1F for SYNT0. Also with that all the tests between two kits ran perfectly fine. And referring to the S2LP RM on page 76, the SYNT is 28bit integer (PLL programmable divider), and SYNT0 register value contributes to bits [7:0], so values either 0x2F or 0x1F (which are 47 (decimal) or  31 (decimal) do not make significant difference for the actual value of SYNT and hence final channel frequency. 

2. For

BALF-SPI2-01D3, I 

did the open component resistance measurement which gave following results:

RxP-RxN = 1.0ohm

Tx-4GND = 2.6ohm

Tx-6GND = 4.8ohm

ANT-4GND = 4.9ohm

AND-6GND = 2.8ohm

Tx-ANT = 2.1ohm 

And with balun already soldered on my custom board, I got similar values, of course I cannot access the terminals of the balun when soldered, but measuring the values on connected component lead, this also confirms that the balun is properly soldered. Also checked that other components (capacitors and inductors and S2LP) are soldered perfectly. 

I also want to mention that my PCB is two layered, I want to make another version of PCB with 4 layers and follow exact guide lines that are given here (

http://www.st.com/content/ccc/resource/technical/document/application_note/group0/dc/41/6b/68/62/9f/41/d9/DM00341356/files/DM00341356.pdf/jcr:content/translations/en.DM00341356.pdf

) . I am sure it would work out fine. 

Meanwhile, I have another question: does ST offer S2LP evaluation board that come with this balun  

BALF-SPI2-01D3?

Thank you so much for your help and valuable inputs      

Silas