Hi

I have a basic radio configuration array(const uint32_t generated[]) , how can I turn it into a radio configuration file (.isc) and then change some of the parameters in the configuration file.

or Can Simplicity Studio import radio configuration array?

Thank you in advance

I setup the hal com port to be able to get reset, assertion and my own debug information using uart2.

On startup the serial port works fine and I get reset information and my debug information.

I am using the Flex v2.5.5.0 stack as a wireless sensor.

In the idle-sleep plugin, okToIdleOrSleep() calls emberSerialWriteUsed() which calls COM_WriteUsed() which prevents the board from going to sleep since it returns 5 bytes (comhandle->txQueue->used is 5). If I manually in the debugger clear comhandle->txQueue->used = 0, the board goes to sleep.

When I write debug information I called COM_WriteData() but I also tried COM_ForceWriteData() with the same result.

//-----------------------------------------------------------------------------
// BOOT_RXD3            - PF4    US2_RX Location16: PF4
// BOOT_TXD3            - PF6    US2_TX Location19: PF6
//-----------------------------------------------------------------------------
#define BSP_SERIAL_APP_PORT                      HAL_SERIAL_PORT_USART2
#define BSP_SERIAL_APP_RX_LOC                    USART_ROUTELOC0_RXLOC_LOC16
#define BSP_SERIAL_APP_TX_LOC                    _USART_ROUTELOC0_TXLOC_LOC19
#define HAL_SERIAL_APP_BAUD_RATE                 115200
#define HAL_SERIAL_APP_FLOW_CONTROL              HAL_USART_FLOW_CONTROL_NONE
#define HAL_SERIAL_APP_RXSTART                   16
#define HAL_SERIAL_APP_RXSTOP                    16

#define HAL_USART2_ENABLE                        (1)

#define BSP_USART2_RX_PIN                        (4U)
#define BSP_USART2_RX_PORT                       (gpioPortF)
#define BSP_USART2_RX_LOC                        (16U)

#define BSP_USART2_TX_PIN                        (6U)
#define BSP_USART2_TX_PORT                       (gpioPortF)
#define BSP_USART2_TX_LOC                        (19U)

#define HAL_USART2_BAUD_RATE                     (115200UL)
#define HAL_USART2_RXSTOP                        (16UL)
#define HAL_USART2_RXSTART                       (16UL)
#define HAL_USART2_FLOW_CONTROL                  (HAL_USART_FLOW_CONTROL_NONE)

#define HAL_SERIAL_APP_RX_QUEUE_SIZE             64
#define HAL_USART2_RX_QUEUE_SIZE                 (64)
#define HAL_SERIAL_APP_TX_QUEUE_SIZE             128
#define HAL_USART2_TX_QUEUE_SIZE                 (128)

I tried increasing the Tx queue sizes but I always get a stall with 5 bytes in the tx queue but nothing is being transmitted.

What am I doing wrong?

Hello, 

I am attempting to limit power consumption on my SI radio since it will be running on battery. My idea was to use LDC mode and configure it so that the LDC/WUT periods are such that I will never miss a preamble. Current configuration:

Preamble of 40 bits, 57600bps, Preamble RX Thresh 20 bits. 

Since my preamble is 694us long, I cannot sleep for more than ~300 us at a time.

I was a little confused on the WUT/LDC configuration since the datasheet and API doc have the same calculations, but the API units are seconds, while the datasheet has ms. Which one is correct?

I assumed ms, and calculated a WUT_R of 4, WUT_M of 153, giving a WUT of around 0.3ms. I did the same with the LDC so that LDC+WUT = around .6ms. This should allow me to never miss a packet.

However the response time from SLEEP to RX is 440us, so is this even possible? Or will I need to lengthen my preamble if I want to do this scheme?

I need to evaluate how the EFR32FG14 chipset performs when paired with one of our legacy radios to see if it is worth using for our new radio board, which must be backwards compatible. Thus I have to set the radios to work a certain way.

I am trying to follow the RAIL tutorial at

https://www.silabs.com/community/wireless/proprietary/knowledge-base.entry.html/2018/05/17/rail_tutorial_1_int-ZBDY

and I am having many issues, mainly with missing .h and .c files. I am following this because of the following line:

If you have an existing protocol, and you must be compatible with it, you'd probably have to use RAIL (as all of our stacks use a standardized frame format). If you want to use sub-GHz communication, you can use RAIL or Connect (and Z-Wave, in the near future), but for simple point to point communication, it's probably simpler to use RAIL.

This is apparently the RAIL tutorial part 1, but it seems to start somewhere in the middle. From the tutorial:

To link the RAIL library, you'll need the following emlibs:

• em_cmu.c
• em_core.c
• em_sytem.c
• em_emu.c (not a direct requirement but a requirement for em_cmu.c)

I was able to find these on the website, but not as files that could be downloaded. Instead, they had to be copy/pasted into notepad. (and ctrl+A grabs the whole page). They should be included in the folder already, or downloadable directly.

Before initializing RAIL, you have to initialize a few things first:

• DCDC converter should be initialized for the hardware
• HFXO clock should be running and should be used as the main oscillator

There are no instructions on how to do this.

From main.c in this project:

#include "rail.h"
#include "hal_common.h"
#include "rail_config.h"

These files can be found nowhere. I can't find them on the SI website, or in the files downloaded with the dev kit. Besides these errors, there are many, many variables that cannot be resolved in the main.c file, probably because of the missing headers..

Is this tutorial outdated? Should I be following a different one? Is there a guide for the tutorial?

All I need is to be able to configure the radio in a certain way, and then test how well it performs under different conditions. If the performance is good enough, then it will go to a different team for development.

I've tried using the "wire replacement" example, but when I change the configuration from the defaults, I can no longer transmit. It just tells me "Failed" with no other explanation when sending "start-master 0".

Regards,

Jeff

Hello,

I have a custom design that I am having a software issue in. Occasionally when I receive an INVALID_SYNC interrupt from the SI chip, the sequence of commands I use for recovery causes the CTS line on the SI chip to lock high (I'm reading it via GPIO). In my code, if it is high for > 5 seconds I assume the SI chip is locked up and reset it. However, I'd like to prevent it from locking up in the first place.

Here is my code structure for an invalid sync:

/* State change to */
SiChangeState(SI446X_CMD_CHANGE_STATE_ARG_NEXT_STATE1_NEW_STATE_ENUM_SLEEP);

/* Reset FIFO */
if(!SiResetFIFO(SI446X_CMD_FIFO_INFO_ARG_FIFO_RX_BIT)){
    SendDebug("At InvalidSync\r\n");
}
            
/* Restart RX Process */
SiStartRX();

Here is my SiChangeState() Code:

bool SiChangeState(SiStates State)
{
    Data[0]=SI446X_CMD_ID_CHANGE_STATE;                           //Command to Change State
    Data[1]=State;                                                //Desired state
    if(!WriteRadioSPI(Data, SI446X_CMD_ARG_COUNT_CHANGE_STATE, true)){
        SendDebug("SiChangeStateFailed\r\n");
        return false;
    }
    return true;    
}

Here is my SiFIFOReset() Code:

bool SiResetFIFO(uint8_t FIFO)
{
    Data[0]=SI446X_CMD_ID_FIFO_INFO;  //Command to reset FIFO buffers
    Data[1]=FIFO;                     //Reset given FIFO
    if(!WriteRadioSPI(Data,2, true)){
        SendDebug("SiFIFOResetFailed\r\n");
        return false;
    }
    return true;                      //Return success
}

Both of them use WriteRadioSPI() to communicate with the SI chip:

bool WriteRadioSPI(uint8_t * data, uint16_t len, bool RelCS)
{
    uint16_t i;
    uint8_t * pdata;
    uint8_t temp;

    //Check Radio CTS Line
    if(!check_si_cts_line()) {
        SendDebug("CTS Failed WRS at CMD: ");
        SendDebugHex(data, len);
        return false;
    } 

    pdata=data;
    SetRadioCS();

    SPI2STATbits.SPIROV=0;                  //Clear overflow errors
    IFS1bits.SPI2RXIF=0;                    //Clear the RX interrupt
    while (SPI2STATbits.SPIRBF)
    {
        temp=SPI2BUF;                       //Purge RX Buffer
    }
    IFS1bits.SPI2RXIF=0;                    //Clear the RX interrupt

    for (i=0;i<len;i++)
    {
        temp=*pdata++;
        SPI2BUF=temp;   //Send the data
        while (!SPI2STATbits.SPITBE){};     //Wait for the transmit buffer to empty
        while (!IFS1bits.SPI2RXIF){};       //Wait for the receive buffer to fill
        temp=SPI2BUF;                       //Read and discard
    }
    while (SPI2STATbits.SPIBUSY){};         //What for the SPI to shift out all the data

    if (RelCS)
    {
        ClearRadioCS();
    }
    return true;
}

During that FIFO reset for an INVALIDSYNC word, it occasionally locks up at the check_si_cts_line() inside WriteRadioSPI(). However, that function is utilized many other places, as is WriteRadioSPI(). This occurs at different frequencies, and not every time an invalid sync occurs... Does anyone have any idea what could cause such behavior?

EDIT:

Also here is check_si_cts_line():

bool check_si_cts_line()
{
    uint64_t cts_timeout = SysTime + 5000; //5 Second Timeout
    bool printed = false;
    while(!RadioCTS()) {
        ClearWDT();
        if(SysTime > cts_timeout && !printed) {
            printed = true;
            si_restart_si446x();
            return false;
        }
    };
    return true;
}

Hello, I have a weird issue with the EFR32MG12P432F1024IM48. I configure it for Ghz-Mode and still the range for setting the tx power (RAIL_SetTxPower) is from 0 to 248, which should be the case for Sub-Ghz-Mode.

static RAIL_TxPowerConfig_t txPowerConfig = {
    .mode = RAIL_TX_POWER_MODE_2P4_HP,
    .voltage = BSP_PA_VOLTAGE,  // 3300
    .rampTime = HAL_PA_RAMP,
  };

What I'm doing wrong?

Hello

Si4468 A2A proprietery board with RFswitch controled from GPIO3 that is configured as TX_STATE output pin.

When execute for first time the START_TX command with TXCOMPLETE_STATE state parameter equal to RX_TUNE, the GPIO3 does not go to LO state after finish transmission.
Even the chip goes to RX state or READY state  by send the appropriate command , the pin remains HI.
In case that transmits with TXCOMPLETE_STATE = STATE_READY, everything goes fine as expected.
I changed the chip with Si4463B1B and i observed the same behaviour.
The WDS script is attached

Hi,

we are evaluating EFR32MG12 chip with (BRD4001A rev A01)/(BRD4172B Rev A00) @ 169MHz for very big project ( more than 100000 units) . We have also our own custom hardware for 250 MHz operation. After spending a lot of time on this RFSoC we have few observations.

1.) RFSoC is working fine on low bitrates (<34 kbps) and low bandwith (<50 kHz) modes. We have nice LOS range, sensitivity, PER stability.

2.) We are not satisfied with wideband operation results(sensitivity, PER, link stability with boards movement ) in real enviroment
    -please send to us your best configured rail_config.c setup for your board @169MHz  for 1 Mbit and 2 Mbit operation
    -please send to us your best configured rail_config.c setup for your EFR32MG12 chip @250MHz  for 1 Mbit and 2 Mbit operation
    -you can use any modulation to achieve best sensitivity, PER, link stability with boards movement

3.) We can't get OQPSK modulation to work @169Mhz @ 1 and 2 Mbit operation
 
3.) Very big PER when boards are moving at lowe speed (<10km/h). We think that AGC GUI setup can't be optimal set in wideband mode?  
     -please refer to your answer on Propriety Wireless forum topic "EFR32MG12 and fast wireless channel changes" from 12.10.2019.
     -you made modificiation outside of the possibilities present within .isc/GUI to put RFSoC in operation !!!!!
     -If we can't setup AGC in GUI that is BIG problem for us ???????
      
5.) Are auto generated parameters outside ISM bands optimal? There is big difference from frequency to frequency when we measure system on table using coax cable and attenuators.

6.) There is deviation drift when we put FSK2 1000kbps 250kHz deviation!! Please check attached file. How can we get exact GMSK modulation and be shure that simbols are changing on continous phase even on carrier frequency change?

 All tests are done by high expirienced RF engineers in SDR. This evaluation is for project in licensed Sub-GHz bands (200-600 MHz) with mixed narrowband/wideband mode.

Best regards,

John

https://www.silabs.com/community/wireless/proprietary/forum.topic.html/about_flashing_with-9ECB

I have a same issue in that forum that I linked

He say that TPJ11 instead of TPJ10 should provide +3.3v.

But I don't recognize what TPJ11 & TPJ10 mean.

Can someone know that TPJ11 & TPJ10 mean?

你好，

si4010通弄过IO口唤醒之后，程序指针是在主循环接着运行，还是从nvm从新加载程序运行？

通过sleeptim进入休眠后，计时到0 程序唤醒后是从主循环接着运行还是从nvm从新加载程序运行？

急急急！！！

新入手了一份SI4010的开发套件，但是没有软件例程无从下手，我想问下官方有SI4010的软件开发例程吗？要是有一份完整的工程文件就最好了！

手上有一套SI4010的开发套件型号是： 4010-KFOBDEV-434，但是从官网下的PDF资料中找不到对应的模块板原理图，请告诉我哪里可以找到，感激不尽！

你好，

我们需要设计一个工程，上电之后定时周期性发送数据包，有以下几个问题需要您帮忙回答：

1、设置vSys_Setup( 10 )，后面使用vSys_FirstPowerUp()------这个的400ms的延时是电池首次上电才有？wakeup之后再次进入是不是没有这400ms延时？

2、之前看到这样一句话“Please note also that vSys_FirstPowerUp function clears the sleeptimer and the HVRAM, so don't use it in your code in a buttonless application. You don't need to use it, if you never set the vSys_BandGapLdo to zero“，如果我的工程需要上电之后周期性唤醒程序（不靠按钮唤醒），那么是不是需要屏蔽掉vSys_FirstPowerUp（）函数。

3、针对上面第2点，如果不使用vSys_FirstPowerUp（）函数，是否会存在异常。且SYSGEN & M_POWER_1ST_TIME会在sleeptim定时器唤醒之后置0？

以上请帮忙