Z-Stack modification to enable CC2592 front-end with CC2530ZNP

Since you use ZNP, do you rebuild ZNP hex file with compile option HAL_PA_LNA_CC2592 defined?

Yes sir, I add HAL_PA_LNA_CC2592 define
you can see picture of my screen in this link: onedrive.live.com/redir

What HW do you use?

Dear Sir,

I built hardware exactly following your reference design: www.ti.com/.../tidra40a.pdf , using external antenna.

CC2530 connect to STM32F103 using SPI. I think this connection is OK because I configured and the client and coordination can communicate. But very short distance. In my sense, CC2592 not working and the RF only transmission at the wire between CC2530 and CC2592.

Can you send me your compiled hex file in your side? I can test it for sure that the error does not come from my compiled process.

BR,
Kien.

Try to call "ZMacSetTransmitPower(TX_PWR_PLUS_16);" in znpInit() to set TX power to 16dbm first. By the way, you can also check signal on P1.1, P1.0, and P0.7 to see if the communication between CC2530 and CC2592 is good.

Dir Sir,

I am using ZNP following this API: e2e.ti.com/.../6758.Z_2D00_Stack-Monitor-and-Test-API.pdf . so that I do not know how to make this ZMacSetTransmitPower(TX_PWR_PLUS_16); function.

Can I use SYS_SET_TX_POWER API in page 53 of the document about to set TX transmission power to 16?

By the way, how can I check P1.1, P1.0, and P0.7 in details?

Thank you,
Kien.

1. No, I don't think you can use SYS_SET_TX_POWER API. You just do copy "ZMacSetTransmitPower(TX_PWR_PLUS_16);" into your znpInit() of ZNP project and define compile option HAL_PA_LNA_CC2592 to rebuild hex file.
2. You can use scope to oscilloscope to check those 3 pins to see if they work as expect.

Dear Sir,

OK. I see your comment now. I will do it today.
Tomorrow at this time, I will discuss with you more about the results.

Thank you very much,
Kien.

You are welcome.

Dear Sir,

I did as you suggest with set TX_POWER to 16dBm, but the result is the same with very short distance for communication.

When I change TX_POWER to 4dBm and compile ZNP project without HAL_PA_LNA_CC2592 flag, the communication length is better, it is about 2m.

My testing hardware using coordination is a CC2530 module like this g03.a.alicdn.com/.../-Midas-Free-Shipping-1-PCS-LOT-CC2530F256RHAR-CC2530-zigbee-font-b-wireless-b-font-module.jpg and STM3F103 evaluation board; the end-device is the board I design with stm32f103, cc2530 and cc2592 following your reference design: www.ti.com/.../tidra40a.pdf

Do you think the problem comes from my cc2592 hardware?

I can borrow a spectrum analyzer, so can you tell me which API function in e2e.ti.com/.../6758.Z_2D00_Stack-Monitor-and-Test-API.pdf can I use to make cc2530 in transmission mode to measure it TX power using that spectrum analyzer?
It means that If i set it to be a coordinator, does it always transmission power to the air for measuring?

Thank you,
Kien.

I would suggest you using SmartRF Studio 7 to do RF test using your CC2530+CC2592 design first. This will help you to know if it is HW issue.

Dear Sir,

Can I use SmartRF04EB instead of CC-Debugger for using with SmartRF Studio 7?

Kien.

I am not sure if latest SmartRF Studio 7 support SmartRF04EB. You can try to know it.

OK, thank you sir,

I have one more question about it. Do I need to connect SPI interface from CC2530 to the SmartRF04EB or just need debug interface (DD and DC) to work with smartrf studio?

Kien.

I think DD and DC are enough.

Thank you sir,

I will do testing. Then we will discuss more with new results.

Have a nice day.
Kien.

You are welcome.

Dear Sir,

I do some testing and I summary results as below:

1. Hardware using: - a CC2350 module bought from the market, a CC2530-CC2592 board that I made following TI reference design, and two SmartRF04EB.

2. Using SmartRF04EB to control CC2530-CC2592 board, then using a spectrum analyzer to monitor the output. The input of the spectrum analyzer is connected directly to the antenna of CC2530-CC2592 board. The result is noted in this file: www.dropbox.com/.../Test CC2530_CC2592 with spectrum analyzer.docx
As the result, I think the CC2530-CC2592 hardware is OK. To double confirm this, I do more test as below:

3. Using a SmartRF04EB to control CC2530 module to transmit with TX=4.5dbm by smart RF studio in a PC, using another SmartRF04EB to control CC2530-CC2592 board to receive by smart RF studio in another PC, result is very good with 0% error rate:
- image of Transmit side: www.dropbox.com/.../cc2530 module transmit with 4.5dbm.png
- image of Receive side: www.dropbox.com/.../CC2530-CC2592 recv from TX = 4.5dbm.jpg

4. Using CC2530-CC2592 board to transmit with TX=5.4dbm, CC2530 module to receive, I also have good result:
- image of transmit side: www.dropbox.com/.../CC2530-CC2592 transmit with 5.4dbm.jpg
- image of receive side: www.dropbox.com/.../cc2530 module recv from TX = 5.4dbm.png

5. Using CC2530-CC2592 board to transmit with TX=15.7dbm, CC2530 module to receive, I also have good result:
- image of transmit side: www.dropbox.com/.../CC2530-CC2592 transmit with 15.7dbm.jpg
- image of receive side: www.dropbox.com/.../cc2530 nodule recv from TX= 15.7dbm.png

6. Finally, I compile ZNP project as you suggest: add HAL_PA_LNA_CC2592 flag and set TX power to 16. You can see my configuration in this picture: www.dropbox.com/.../znp modify.jpg
Then I use STM32F103 to control this CC2530-CC2592 module using SPI interface, I set this one to be a coordinator and use the CC2530 module to receive its transmittion. THe result is very bad even though I put two circuits closer and they are very close. You can see the result here: www.dropbox.com/.../CC2530 module - ZNP result with very short distance.jpg

All of my testing and configuration for STM32F103, I set frequency to 2425Mhz (channel 15).

Do you think what should I do next?
Did you check the result of ZNP project for zstack version 1.2.2a and got better result?

Regards,
Kien.

When you do this TX/RX tests, how far is it between CC2530 and CC2530+CC2592?

Dear Sir,

For experiment 3, 4, and 5, the distance is about 1 meter. If you need, I can do test again with longer distance. However, you can see in the RX side, even the transmit power is 4.5dbm, average RSSI is very good at about -50dbm to -60dbm, so i think it is also good with much longer distance.

For experiment 6 with MCU and CC2530ZNP-CC2592, with 1 meter distance, there is no receive, the result is the picture is about 20cm. The RSSI is very bad (more than -100dbm) and sometimes we have CRC error, even though I set TX power to +16 as you suggested.

Kien.

If you have a look at RX sensitivity when you use CC2530 TX 4.5 dbm and CC2530+CC2592 as RX, the RX sensitivity is only -52.6 dbm and it is too low for 1 meter test. I suspect the problem is still on your HW.

Dear Sir,

In your experience, what is good RX sensitivity for TX=4.5dbm for 1m distance?

I will check the hardware again. We will discuss more with new testing result.

Kien.

When I test on two CC2530DK with TX=4.5 dbm for 1 meter distance, the average RX sensitivity is no more than 40 dbm. However, I don't think your result ( -52.6 dbm) will make ZNP work only smaller than 5 cm. Can you show me your schematic between CC2530 and CC2592?

Dear Sir,

with your test, you mean maximum RX sensitivity is 40dbm or -40dbm for TX=4.5dbm?

You can see my schematic in this link: onedrive.live.com/redir

by the way, can I connect directly antena of two boards with rf coaxial cable for testing?

Kien,

RX sensitivity is -40dbm for TX=4.5dbm in my test. Your schematic looks OK.

Dear sir,

About STM32 MCU firmware which is control CC2530ZNP, do you think it is OK?

Kien.

I don't see why it would have problem to control CC2530ZNP. If you have concerns on it, I would suggest you to test your CC2530-CC2592 ZNP with Ztool first.

Thank you sir,

I will check hardware first. If I find anything wrong, I will tell you.

Kien.

You are welcome. Another suggestion is that maybe you can try to run SampleLight or SampleSwitch example on it to do range test and see if your HW is good.

Hello,

   There are 3 pins that are used to control the PA/LNA:

Name	Signal Name	A2530 Pin	CC2530 Pin Name
EN	LNA On	7	P1_3
HGM	PA/LNA High Gain Mode	12	P0_7
PAEN	PA On	9	P1_1

The A2530 is the Anaren A2530 module which is based on the CC2530ZNP but with some bugs fixed by yours truly. :)

Here's where the fun begins: P1.3 was not properly configured in the stack for PA/LNA control. They may have fixed it, but you should check. I had to manually fix this for the Anaren module.

//OLD
/* P1_4 -> EN (LNA control) */
//    RFC_OBS_CTRL1 = RFC_OBS_CTRL_LNAMIX_PD_INV;
//    OBSSEL4       = OBSSEL_OBS_CTRL1;

//NEW
/* P1_3 -> EN (LNA control) */
    RFC_OBS_CTRL1 = RFC_OBS_CTRL_LNAMIX_PD_INV;
    OBSSEL3       = OBSSEL_OBS_CTRL1;

Dear Derek,

Thank you for you suggestion. In my case, I using CC2592 with EN - P1.0 | PA_EN - P1.1 | HGM - P0.7

In the ZNP project:

else if(paLnaChip == PA_LNA_CC2592)
{
/* P1_1 -> PAEN */
RFC_OBS_CTRL0 = RFC_OBS_CTRL_PA_PD_INV;
OBSSEL1 = OBSSEL_OBS_CTRL0;

/* P1_0 -> EN (LNA control) */
RFC_OBS_CTRL1 = RFC_OBS_CTRL_LNAMIX_PD_INV;
OBSSEL0 = OBSSEL_OBS_CTRL1;
}

So, everything is correct. As sir YK Chen suggestion, I think the problem comes from my hardware, I am finding way to check it.

Thank you so much.
Kien.

The other thing, please remember to setting TX power for your demand

FYI.

Ethan Chen

Dear Sir Chen,

Thank you so much for your help. I will modify and do testing with your suggestions.

I will tell you results tomorrow.

Regards,
Kien.

Dear Sir Chen,

When I follow your suggestion, everything works correct as expected. All problems are solved now.

You and sir YK Chen helped me so much with your supports.

Thank you so much.

Best regards,
Kien.

You are welcome and it's good to know it works now.

Hi Kien,

It's great. Actually TI this version stack totally project have define for CC2592 beside of ZNP.

I can recommend TI to add this part with next version.

Thanks your verify.