If the application running in the KW41Z does not operate in low power mode, the device could work without the 32 kHz oscillator. However, for it to work correctly, the clock configuration must be changed. Figure 1.  KW41Z clock distribution By default, the software stack running on the KW41Z selects a clock source that requires the 32 kHz oscillator. However, if the 32 kHz oscillator is not available, the clock configuration must be changed. Fortunately, the option to change it is already implemented, it is only required to change the clock configuration to the desired one. To do this, change the value for the CLOCK_INIT_CONFIG macro located in the app_preinclude.h file. /* Define Clock Configuration */ #define CLOCK_INIT_CONFIG           CLOCK_RUN_32‍‍‍‍‍‍‍‍‍‍ In the selected mode in this example, CLOCK_RUN_32, the selected clock mode is BLPE (Bypassed Low Power External). In this mode, the FLL is bypassed entirely, and clock is derived from the external RF oscillator, in this case, the 32 MHz external clock. These macros are the default available options to change the clock configuration, they are located in the board.h file. It is up to the application and the developer to chose the most appropriate configuration. #define CLOCK_VLPR       1U #define CLOCK_RUN_16     2U #define CLOCK_RUN_32     3U #define CLOCK_RUN_13     4U #define CLOCK_RUN_26     5U #define CLOCK_RUN_48_24  6U #define CLOCK_RUN_48_16  7U #define CLOCK_RUN_20_FLL 8U‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍ More information regarding the different clock modes and their clock sources are available in the KW41Z reference manual, Chapter 5: Clock Distribution, section 5.4 Clock definitions, and Chapter 24: Multipurpose Clock Generator.

1 Introduction Two development boards transmit control information through ble. One development board connects to paj7620 and provides gesture information through IIC bus. The other development board uses ble and USB HID. Ble is used to receive data, and USB HID is used to simulate keyboard input and control ppt                  Figure  1 2 Preparation We need two development boards qn908x and gesture control device paj7620. We use IAR as development enviroment.The example we use is temperature_sensor, and temperature_ colloctor. The SDK version is 2.2.3   3 Code 3.1  temperature_sensor code We want to implement IIC to read gesture information from paj7620 and send data. The pins used by IIC are PA6 and PA7 Simply encapsulate the IIC reading and writing code in the code to create i2c_ operation.c and i2c_ operation.h. Realize IIC initialization and reading / writing register function in it                        Figure  2                        Figure  3   3.1.1 After having these functions, we begin to write gesture recognition code. First, we add two blank files paj7620.c and paj7620.h into our project.   Select bank register area                               Figure 4   Wake up paj7620 to read device state                    Figure 5   Initialize device                    Figure 6   Gesture test function                                   Figure 7   3.1.2 When you are ready to read the device information, You should initialize IIC and paj7620 in BleApp_Init function                                Figure 8 In principle, we need to create a custom service for the PAJ device, but we replace the temperature data as our gesture control data. If you want to create a custom service, refer to this link custom profile   3.1.3 Create a timer that sends gesture data regularly. In file temerature_sensor.c Define a timer，static tmrTimerID_t dataTimerId; Allocate a timer, dataTimerId = TMR_AllocateTimer(); Define the callback function of this timer                                           Figure 9 Start timer                                    Figure 10 Close the low power mode. #define cPWR_UsePowerDownMode 0 3.2 temperature_collector code The most important thing here is to port USB HID into our project. The USB  example we use is the USB keyboard and mouse. 3.2.1 Add the OSA and USB folder under the example to the project directory, and copy the file to the corresponding folder according to the file structure of the original example.                      Figure 11 3.2.2 Add header file directory after completion                                           Figure 12 At the same time, in this tab, add two macro definitions USB_STACK_FREERTOS_HEAP_SIZE=16384 USB_STACK_FREERTOS   3.2.3 Next, we need to modify the main function in usb example . Open composite.c file.                      Figure 13 It calls the APP_task. So this function also need to be modified.                                   Figure 14 3.2.4Find hid_mouse.c，Comment function USB_DeviceHidMouseAction Find hid_keyboard.h. Define the gesture information.                                  Figure 15 Find hid_keyboard.c. We need to modify the function USB_DeviceHidKeyboardAction as following figure.                                                  Figure 16   Among them, we also need to implement the following function. When the up hand gesture is detected, the previous ppt will be played. The down hand gesture will be the next PPT, the left hand gesture will exit PPT, and the forward hand gesture will play ppt                                                  Figure 17 It also refers to an external variable gesture_from_server. The variable definition is in file temperature_ collocation.c,.     3.2.5 After that, let's go to BleApp_Statemachinehandler function in temperature_colloctor.c. In case mApppRunning_c, we will call usb_main to initialize USB HID                                                  Figure 18 3.2.6 In BleApp_PrintTemperature, we will save the gesture data to gesture_from_server                                                         Figure 19 We finished the all steps.        

This Document describes the additional changes needed for JN-AN-1229 ZigBee PRO Application Template for ZigBee version 1002 to be able to compile correctly with the latest update SDK JN-SW-4170 Version 1745 and JN-SW-4270 Version 1746. Note:  These modifications can be also found in the JN516x ZigBee 3.0 SDK Release Notes v1745 (Chapter 4.3 Modifications Required) Tool modifications The .zpscfg file contains all the information available of the setup of the ZigBee network, this file it’s located in C:\...\...\workspace\ JN-AN-1229\Common The first step is to add the MAC Interface List to The ZigBee devices (Co-ordinator, Router, and Sleeping End Device) in the .zpscfg file. Then, add the MAC Interface selecting the New Child option. In the Properties tab, the Co-ordinator and the Router should have the Router Allowed option set to True. Stack modifications The new stack has support for better throughput and automatic buffering of data packets during route discovery. This requires the addition of a new queue in the application. Open the app_common.h file which it’s located in the Common folder. This queue should be tied to the stack definition: extern PUBLIC tszQueue zps_msgMcpsDcfm; After that, open the app_router.c and app_sleeping_enddevice.c. The ZPS_tsAplAib structure has been changed, so, the vStartup function should be modified Old version /* Set channel to scan and start stack */ ZPS_psAplAibGetAib()->apsChannelMask = 1 << u8Channel; Update needed /* Set channel to scan and start stack */ ZPS_psAplAibGetAib()->pau32ApsChannelMask[0] = 1 << u8Channel; Add the next code to the app_start.c(Coordinator and Router) and app_start_SED.c(Sleeping End Device) The buffer of the Router device should be modified. The size of the queue is defined as: #define MCPS_DCFM_QUEUE_SIZE 5 The storage of the queue must be defined: PRIVATE MAC_tsMcpsVsCfmData asMacMcpsDcfm[MCPS_DCFM_QUEUE_SIZE]; In the APP_vInitResources function an additional queue must be added: ZQ_vQueueCreate(&zps_msgMcpsDcfm, MCPS_DCFM_QUEUE_SIZE,  sizeof(MAC_tsMcpsVsCfmData),(uint8*)asMacMcpsDcfm);  

Many applications make use of 32 kHz clocks to keep tracking of real-time or to have a low power time reference for the system. Most of the systems might use a 32.768 kHz XTAL for this purpose. However, there might be some exceptions in which the application requires compensate the frequency of this clock due to temperature changes, aging, or just because the clock provides from a source which frequency is close to the ideal 32.768 kHz, but it presents some variations. QN908x devices require a 32 kHz clock source for some applications like running the BLE stack in low power. 32.768 kHz XTALs are more accurate so they are used to generate a 32 kHz source by compensating for the ppm difference. This provides us with tools to compensate for any external 32 kHz source by first obtaining the ppm difference from the ideal frequency. The solution consists in determining how off is the external clock input frequency from the ideal 32 kHz by making a comparison with a trusted clock in the system, typically the 32 MHz / 16 MHz XTAL. This process is executed in the background in an interrupt-based system so the application can initialize or run other processes in the meantime. Then, the results of the ppm calculation are reported to the main application to compensate for the changes and provide for a more accurate clock source. This example makes use of the following peripherals in the QN908x RTC Seconds Timer CTIMER DMA The RTC Seconds Timer uses the 32 kHz clock source as a reference. It contains an internal 32000 cycles counter that increases each 32 kHz clock period. On overflow, it rises the SEC_INT flag to indicate that one second has elapsed. The CTIMER makes use of the APB clock which derives from the 32 MHz / 16 MHz clock. This timer is used in free-running mode with a Prescaler value of 1 to count the number of APB pulses. The algorithm consists of counting the amount of APB pulses (trusted clock reference) elapsed by the time the RTC SEC_INT flag is set. Ideally, if both clocks are in the right frequency, the number of APB pulses must be equal to the reference clock frequency. For example, if the APB clock is 16 MHz, by the time the SEC_INT flag sets, the CTIMER counter must have a value of 16 x 10 6 counts. Assuming our reference clock is ideal, the difference between the CTIMER counter value and 16 x 10 6 can be used to obtain the ppm difference given the following formula. Where f 0 is the ideal APB frequency (16 MHz) and f 1 is the real measured frequency (CTIMER counter value). Since the pulses counted using the CTIMER correspond to the time it took to the RTC Seconds Timer to count one second, we can extrapolate the obtained ppm value as a difference in the 32 kHz clock source from the ideal 32 kHz. To prevent from any application or task servicing latency, the algorithm makes use of the DMA to automatically capture the CTIMER Counter value when the SEC_INT flag is set. The program flow is described in the diagram below. As a way of demonstrating this algorithm, two APIs were implemented to calculate the ppm value and apply the compensation to the system. Both APIs are included in the file fsl_osc_32k_cal .c and .h files. OSC_32K_CAL_GetPpm (osc_32k_cal_callback_t pCallbackFnc): Initializes the required timers and DMA and starts with the CTIMER capture. A callback is passed so it can be executed once the ppm calculation sequence completes. OSC_32_CAL_ApplyPpm (int32_t ppmMeasurement): Uses the previously calculated ppm passed as an input parameter to compensate the RTC and the BLE timer used during sleep mode. OSC_32K_CAL_GetPpm is called every time the ppm value of the 32 kHz source clock needs to be calculated. It takes around one second to complete (depending on how off the 32 kHz source clock is from the ideal frequency) and the application cannot enter into low power state during this time. The registered callback function is executed once the calculation is complete. The ppm calculation is performed into the DMA callback. It consists of obtaining the CTIMER counter difference and use it as f 1 in the formula shown before. The ppm values are calculated using floating point unit. /* Calculate PPMs */ ppmResult = (float)((float)1-((float)ApbClockFreq/(float)ApbCountDiff)); ppmResult *= (float)1048576;‍‍‍‍‍‍ Then OSC_32_CAL_ApplyPpm must be called using the ppm value obtained after calling OSC_32K_CAL_GetPpm. This API programs the necessary values in the RTCàCAL register and the BLE Sleep timer registers to compensate for the differences in the 32 kHz source clock. Finally, the user must account for all those other APIs that make use of the 32 kHz clock frequency and update the values accordingly. For the particular case of the NXP BLE Stack, there are two APIs that need to be updated to return the clock frequency after the calibration has been applied. uint32_t PWRLib_LPTMR_GetInputFrequency(void) { uint32_t result = 32000; int32_t ppm = 0; if ( RTC->CTRL & RTC_CTRL_CAL_EN_MASK) /* is calibration enabled ? */ { /* Get the current calibration value */ if (RTC->CAL & RTC_CAL_DIR_MASK) { /* Backward calibration */ ppm -= (int32_t) (RTC->CAL & RTC_CAL_PPM_MASK); } else { /* Forward calibration */ ppm += (int32_t) (RTC->CAL & RTC_CAL_PPM_MASK); } /* Obtain the uncalibrated clock frequency using the formula * fUncal = 32000 - (ppm*0.03125) where 0.03125 is the number * of Hz per PPM digit obtained from (768 Hz/0x6000 PPM) */ result -= (float) ppm * (float) 0.03125; } else { #if (defined(BOARD_XTAL1_CLK_HZ) && (BOARD_XTAL1_CLK_HZ == CLK_XTAL_32KHZ)) result = CLOCK_GetFreq(kCLOCK_32KClk); /* 32,768Khz crystal is used */ #else result = CLOCK_GetFreq(kCLOCK_32KClk); /* 32,000Khz internal RCO is used */ #endif } return result; }‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍ uint32_t StackTimer_GetInputFrequency(void) { uint32_t prescaller = 0; uint32_t refClk = 0; uint32_t result = 0; #if FSL_FEATURE_SOC_FTM_COUNT refClk = BOARD_GetFtmClock(gStackTimerInstance_c); prescaller = mFtmConfig.prescale; result = refClk / (1 << prescaller); #elif FSL_FEATURE_RTC_HAS_FRC int32_t ppm = 0; (void)prescaller; /* unused variables */ (void)refClk; /* suppress warnings */ result = 32000; if ( RTC->CTRL & RTC_CTRL_CAL_EN_MASK) /* is calibration enabled ? */ { /* Get the current calibration value */ if (RTC->CAL & RTC_CAL_DIR_MASK) { /* Backward calibration */ ppm -= (int32_t) (RTC->CAL & RTC_CAL_PPM_MASK); } else { /* Forward calibration */ ppm += (int32_t) (RTC->CAL & RTC_CAL_PPM_MASK); } /* Obtain the uncalibrated clock frequency using the formula * fUncal = 32000 - (ppm*0.03125) where 0.03125 is the number * of Hz per PPM digit obtained from (768 Hz/0x6000 PPM) */ result -= (float) ppm * (float) 0.03125; } else { #if (defined(BOARD_XTAL1_CLK_HZ) && (BOARD_XTAL1_CLK_HZ == CLK_XTAL_32KHZ)) result = CLOCK_GetFreq(kCLOCK_32KClk); /* 32,768Khz crystal is used */ #else result = CLOCK_GetFreq(kCLOCK_32KClk); /* 32,000Khz internal RCO is used */ #endif } #elif FSL_FEATURE_SOC_CTIMER_COUNT refClk = BOARD_GetCtimerClock(mCtimerBase[gStackTimerInstance_c]); prescaller = mCtimerConfig[gStackTimerInstance_c].prescale; result = refClk / (prescaller + 1); #else refClk = BOARD_GetTpmClock(gStackTimerInstance_c); prescaller = mTpmConfig.prescale; result = refClk / (1 << prescaller); #endif return result; }‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

Our customer is evaluating RF characteristics using FRDM-MKW24. Regarding Tx max power, they have one question. The spec of max tx power is +8dBm and I could verify the power using TWR-KW24d512 before. They observed tx power with tx un-modulated cnt transmission and informed that the power was about +2dBm. That is to say, "Power 31" in Connectivity_Test means to +2dBm. I feel that it is small... Would you comment regarding the spec of the max tx power on FRDM-MKW24? Regards, Koichi

Commissioner Authentication server for new Thread devices and the authorizer for providing the network credentials they require to join the network. A device capable of being elected as a Commissioner is called a Commissioner Candidate. Devices without Thread interfaces may perform this role, but those that have them may combine this role with all other roles except the Joiner. This device may be, for example, a cell phone or a server in the cloud, and typically provides the interface by which a human administrator manages joining a new device to the Thread Network. Commissioner Candidate A device that is capable of becoming the Commissioner, and either intends or is currently petitioning the Leader to become the Commissioner, but has not yet been formally assigned the role of Commissioner. Commissioner Credential A human-scaled passphrase for use in authenticating that a device may petition to become the commissioner of the network. This credential can be thought of as the network administrator password for a Thread Network. The first device in a network, typically the initial Leader, MUST be out-of-band commissioned to inject the correct user generated Commissioning Credential into the Thread Network, or provide a known default Commissioning Credential to be changed later. Joiner A device to be added by a human administrator to a commissioned Thread Network. This role requires a Thread interface to perform and cannot be combined with another role in one device. The Joiner does not have network credentials. Joiner Router An existing Thread router or REED (Router-Eligible End Device) on the secure Thread Network that is one radio hop away from the Joiner. The Joiner Router requires a Thread interface to operate, and may be combined in any device with other roles except the Joiner role. Information extracted from the Thread Whitepapers available at threadgroup.org

Introduction This document describes the hardware considerations for the schematic and layout of the MKW36A512VFT4 device. This MCU is packaged into a 48-pin HVQFN - 7x7 mm, dissimilar to MKW36Z512VHT4 which comes packaged into a 48-pin LQFN - 7x7 mm (the last one takes part of FRDM-KW36).   Pin Layout  The MKW36A512VFT4 MCU is pin to pin compatible with the MKW36Z512VHT4 (FRDM-KW36) MCU, except for the DCDC pins. The following figure shows the distribution of the pins in the MKW36A512VFT4 MCU (left), compared with the MKW36Z512VHT4 (FRDM-KW36 MCU, right). Surely, this is the most important consideration for MKW36A512VFT4, since you can not simply move the FRDM-KW36 layout on your design. Minimum BOM The following figures show the minimum BOM necessary for each DCDC mode in KW36. For more information about DCDC modes and hardware guidelines, please visit: MKW4xZ/3xZ/3xA/2xZ DC-DC Power Management Bypass Mode   Buck Auto-Start Mode   Buck Manual-Start Mode     Layout Recommendations The footprint and layout are critical for RF performance, hence if the recommended design is followed exactly in the RF region of the PCB, sensitivity, output power, harmonic and spurious radiation, and range, you will succeed. For more information of layout recommendations, please visit Hardware Design Considerations for MKW35A/36A/35Z/36Z Bluetooth Low Energy Devices. The footprint recommended for the MKW36A512VFT4 is shown in the figure below. NXP prefers to use a top layer thickness of no less than 8-10 mils. The use of a correct substrate like the FR4 with a dielectric constant of 4.3 will assist you in achieving a good RF design. Other recommendations during EMC certification stages are: - Specific attention must be taken on 4 pins PTC1, 2, 3 & 4 if they are used on the application. - 4 decoupling capacitors of 3pF are mandatory on those pins and be positioned as close as possible. - Wires from those 4 pins must be underlayer. - NXP recommends putting the vias under the package in case the customer HW design rules allow it. Some recommendations for a good Vdd_RF supply layout are: - Vdd_RF1 and Vdd_RF2 lines must have the same length as possible, linked to pointA (‘Y’ connection). - 12pF decoupling capacitor from Vdd_RF wire must be connected to the Ground Antenna. The purpose is to get the path as short as possible from Vdd_RF1/Vdd_RF2 to the ground antenna. - 12pF decoupling capacitor from the Vdd_RF3 pin must be as close as possible. Return to ground must be as short as possible. So vias (2 in this below picture) must be placed near to the decoupling capacitor to get close connection to the ground layer. The recommended RF stage is shown in the following figure. The MKW36A512VFT4 has a single-ended RF output with a 2 components matching network composed of a shunt capacitor and a series inductor. Both elements transform the device impedance to 50 ohms. The value of these components may vary depending on your board layout. Avoid routing traces near or parallel to RF transmission lines or crystal signals. Maintain a continuous ground under an RF trace is critical to keep unaltered the characteristic impedance of the transmission line. Avoid routing on the ground layer that will result in disrupting the ground under RF traces. For more information about RF considerations please visit: Freescale IEEE 802.15.4 / ZigBee Package and Hardware Layout Considerations.

During last week I receive a question about identify a way to notice when the 32kHz oscillator is ready to be able to use it as clock source. In other words, when the 32kHz oscillator is stable to be used as reference clock for the MCG module. Then, system can switch over it. Kinetis devices starts up using its internal clock which is called FEI mode, then, the applications change to a different mode which require an external clock source (i.e. FEE mode). Normally, we have two options which are: Main reference clock. Talking specifically to KW devices, there is the 32MHz external crystal for the main reference clock. The 32kHz external crystal for the 32kHz oscillator which is driven through the RTC registers.  For the first option, there is a register which could be monitored to know when oscillator is ready (RSIM_CONTROL_RF_OSC_READY). So, it can be polled to notice when the oscillator is up and ready.  About the second option, there is no “OSCINIT” bit for the 32 kHz oscillator. However, there is a way to know when the 32kHz Oscillator is running, it is to simply enable the RTC OSC, and configure the RTC to count. After doing that immediately poll (read) the RTC_TPR register and check it until it is greater than 4096. It will roll over once it reaches 32767 so, it is important to the poll this register in a loop doing nothing else or the register could potentially be read when it is less than 4096 but had already rolled over. Once it is greater than 4096, it can be determined that oscillator is running well. If the RTC is not required, the counter can be disabled. Now that the 32kHz clock is available, the application can switch to FEE mode. So, in order to perform the above description, I modified the “CLOCK_CONFIG_EnableRtcOsc()” as shown next: static void CLOCK_CONFIG_EnableRtcOsc(uint32_t capLoad) {        rtc_config_t rtc_basic_config;        uint32_t u32cTPR_counter=0;        /* RTC clock gate enable */     CLOCK_EnableClock(kCLOCK_Rtc0);     if ((RTC->CR & RTC_CR_OSCE_MASK) == 0u)     {       /* Only if the Rtc oscillator is not already enabled */       /* Set the specified capacitor configuration for the RTC oscillator, "capLoad" parameter shall be set          to the value specific to the customer board requirement*/       RTC_SetOscCapLoad(RTC, capLoad);          /*Init the RTC with default configuration*/       RTC_GetDefaultConfig(&rtc_basic_config);       RTC_Init(RTC, &rtc_basic_config);       /* Enable the RTC 32KHz oscillator */       RTC->CR |= RTC_CR_OSCE_MASK;       /* Start the RTC time counter */       RTC_StartTimer(RTC);             /* Verify TPR register reaches 4096 counts */       while(u32cTPR_counter < 4096)       {          u32cTPR_counter= RTC->TPR;       }       /* 32kHz Oscillator is ready. Based on the application requirements, it can let the RTC enabled or disabled.           In this case, we can disable RTC since it is not needed by this application */       RTC_Deinit(RTC);     }     /* RTC clock gate disable  since RTC is not needed anymore*/     CLOCK_DisableClock(kCLOCK_Rtc0); }‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍ Then, using the above function, it can be noticed when the 32kHz oscillator is ready to be used. Hope this helps....

/*** 2024 October 15th latest disclaimers:  - KW47, MCX W72 are direct derivative from KW45 and MCX W71  - please bookmark this page for future update - this article is for early enablement based on KW45, K32W148, MCX W71 and is pending updates up to wider releases of new KW47 and MCX W72 in 2025 -- Datasheet, Reference Manual and HW manufacturing files shared on request--  ***/ Please, find the important link to build a PCB using a KW47 or MCX W72 and all concerning the radio performances, low power and radio certification (CE/FCC/IC).   KW47 product NXP web page:  pending release of KW47/MCXW72  MCXW72 product NXP web page: https://www.nxp.com/products/processors-and-microcontrollers/arm-microcontrollers/general-purpose-mcus/mcx-arm-cortex-m/mcx-w-series-microcontrollers/mcx-w72x-secure-and-ultra-low-power-mcus-for-matter-thread-zigbee-and-bluetooth-le:MCX-W72X KW-MCXW-EVK getting started NXP web page pending release of KW47/MCXW72  KW47-LOC getting started NXP web page pending release of KW47/MCXW72  MCXW72-LOC getting started NXP web page pending release of KW47/MCXW72     HARDWARE   KW47 MCX W72 EVK board: attached     KW47 LOC Channel Sounding board - diagram: attached     KW47-MCXW72-EVK HW guideline: Available on request        HVQFN48 package specification: SOT619-17(D)   pending release of SOT619-17(DD)       KW47-MCXW72-EVK User Manual pending release of KW47/MCXW72        Minimum BoM (attached file) >> KW45 applicable for KW47 waiting release of KW47/MCXW72      DCDC management guide (AN13831) : KW45/K32W148 - Power Management Hardware (nxp.com) KW45 applicable for KW47 waiting release of KW47/MCXW72      Design-In check list: see attached file at the bottom of this article     RF matching: S parameters (attached file) pending release of KW47/MCXW72    Information: “As RF behavior are dependent of PCB layout & manufacturing; PCB prototypes (based on NXP recommendations) will have to be fine-tuned to insure the expected qualified in RF is reached on the final productized platform.”   RADIO     RF report: KW45 and K32W148 RF System Evaluation Report for Bluetooth LE and K32W148 for 802.15.4 Applications ...  pending release of KW47/MCXW72 - available ON DEMAND     Radio co-existence: Kinetis Wireless Family Products Bluetooth Low Energy Coexistence with Wi-Fi Application (nxp.com) pending release of KW47/MCXW72      Distance performances: refer to attached file pending release of KW47/MCXW72      Antenna:  Compact Planar Antennas for 2.4 GHz Communication Designs and Applications within NXP EVK Boards Antennas for Channel Sounding Applications     BLE connectivity test binary file:  available in SDK on demand      Return loss (S11) measurement: How to measure the return loss of your RF matching (S11) part of the RF report (AN13728)     Loadpull: pending release of KW47/MCXW72   SW tools for RF trials:     IoT Tool box (mobile application)     Connectivity test tool for connectivity products (part of the IoT toolbox)     DTM: How to use the HCI_bb on Kinetis family products a... - NXP Community https://community.nxp.com/t5/Wireless-Connectivity-Knowledge/BLE-HCI-Application-to-set-transmitter-...   CRYSTAL   Article : KW45/K32W1 32MHz & 32kHz Oscilllation margins - NXP Communitypending release of KW47/MCXW72     Recommended Crystal attached LowPower      Estimator Tool https://community.nxp.com/t5/Wireless-Connectivity/Kinetis-KW35-38-KW45-amp-K32W1-MCXW71-Power-Profi... pending release of KW47/MCXW72       Low Power Consumption: https://www.nxp.com/docs/en/application-note/AN13230.pdf pending release of KW47/MCXW72     CERTIFICATION pending release of KW47/MCXW72  

The TWR-KW2x board's OpenSDA is programmed with PE Micro's OpenSDA firmware which enables MSD, debugging and CDC Serial port. This firmware can be easily modified by putting the K20 part in bootloader mode and load another firmware to it with a simple drag and drop. Follow these steps to modify the OpenSDA firmware on the TWR-KW2x board. Segger's OpenSDA v2.1 will be used as an example of the new OpenSDA firmware (Instead of the default PE Micro's) 1. Unplug the board 2. Insert a Jumper in J30 to put the device in Bootloader mode 3. Plug in the board (Mini-USB) 4. Device will be enumerated as a "Drive Disk" But now with a "Bootloader" label 5. Drag and Drop the Segger's JLink_OpenSDA_V2_1.bin firmware (https://segger.com/opensda.html) into the Bootloader unit 6. Unplug the board 7. Remove Jumper 8. Plug in the board (Mini-USB) Now you should see the board being enumerated as "JLink CDC UART Port", allowing serial port communication. You should also be able to debug your application using J-Link debugging interface through the OpenSDA interface, no need of external hardware. Note1: Drivers can be found at Segger's website (https://segger.com/opensda.html) Note2: Jumper has to be in place in J29 for debugging Note3: IDE options must be set to use J-Link Driver

This patch fix the issue in hdmi dongle JB4.2.2_1.1.0-GA release that wifi((STA+P2P)/AP) cann't be enabled properly. In the root directory of Android Source Code, use the following command to apply the patch: $ git apply hdmi_dongle_wifi_jb4.2.2_1.1.0.patch

When having several ZigBee Networks in the same area, and therefore several potential parents, it may become necessary to join one of them and discard the rest. While having a mechanism to only accept joining devices when desired is the best method (like using a button to trigger the joining), it might not always be possible since the parent nodes could be commercial devices or another vendor’s product without this feature. Below are some mechanisms that could be used for this purpose. In general, when searching for suitable parents, the process is as follows: ZDO of device to join sends a MAC scan request. The MAC layer starts scan. For every beacon it receives, it sends a beacon notify indication that is processed in ParseBeaconNotifincaiton() function from AppStackImpl.c The ParseBeaconNotifincaiton() function will add the relevant information in the discovery table and for this it needs a free entry, so it calls GetFreeEntryInDiscoveryTable() function with reuse parameter as FALSE. If the table is full, it will call GetFreeEntryInDiscoveryTable() with reuse set to TRUE to literally re-use low priority entries. When the MAC scan has finished, it will send a MAC scan confirm. When this reaches ZDO, the SearchForSuitableParent() function is called. At this point, there are several approaches that could be used: Use a specific Extended PAN ID to search only for a specific parent node Use a specific PAN ID to prioritize the network’s ID Search in a specific Channel where network is supposed to be operating in All these parameters are configurable in ApplicationConf.h file of the project’s Configure Folder and used in SearchForSuitableParent() function to filter Discovery table entries. Nevertheless, those solutions are not always the best for all applications since it may require hard-coding the network’s parameters. Fortunately, BeeStack leaves all this open for any modification in case it is necessary. In brief, if the discovery table gets full with suitable parents that you DO NOT want to use, you should update the "if(reuse)" statement of the GetFreeEntryInDiscoveryTable() function to replace an entry. In other words, if you think that the desired parent is not present in the discovery table (due to its size limitation or other reason), you should update the GetFreeEntryInDiscoveryTable() function to make sure discovery table contains only devices that are of interest to your node. Please note that the criteria used to select the desired parents is totally application specific. As mentioned, it is always best having a way to trigger the joining such as a button so the rest of parents have permit join set to FALSE and therefore join only to the desired parent without having to implement custom code. Anyway, you may select the solution that meets your application’s requirements the most.

The homologation requirements in China (MIIT [2002]353) obviously are planned (end of December 2022) to be sharpened (MIIT publication from 2021-01-27: “Notice on Matters Related to Radio Management in the 2400MHz, 5100MHz and 5800MHz Bands”).   A modification register is need on the KW38 and KW36 to pass the new Chinese  requirement with acceptable margin: PA_RAMP_SEL value must be set to 0x02h (2us) instead of 0x01h (1us default value) Modification SW: XCVR_TX_DIG_PA_CTRL_PA_RAMP_SEL(2) in the nxp_xcvr_common_config.c All the details are in the attached file.   Note: This SW modification is for China country only.

The SMAC & IEEE 802.15.4 protocol stacks are a KSDK add-on, therefore you need the installation of the KSDK 1.2 before installing these connectivity stacks. Install the Kinetis SDK 1.2: Software Development Kit for Kinetis MCUs|Freescale After installing the KSDK 1.2, download the desired protocol stack. The connectivity software for this platform can be found in the board webpage, in the downloads tab: Modular Reference Boards for Kinetis KW0x|Freescale The installation window will guide you through an easy way to install the software. Best regards, Luis Burgos.

Document Purpose This post entry provides an example of a hybrid application (Wireless_UART + GFSK Advertising) by covering Bluetooth Low Energy multiple node connections in parallel with GFSK (Generic Frequency Shift Keying) communication.  This is an additional example for the SDK where we have defined a Hybrid application for Bluetooth LE advertising and scanning in parallel with GFSK communication. Audience The goal of this post is to serve as a guide for software developers who want to use, adapt and integrate GFSK functionality in a Bluetooth Low Energy application.    Setting up the development environment Toolchain:           - IAR Embedded Workbench 8.32 or newer;            https://www.iar.com/iar-embedded-workbench/   SDK:          - This version of firmware has been tested using SDK_2.2.1_FRDM-KW36, that can be downloaded using the following link: https://mcuxpresso.nxp.com/en/select            (please consider to select as Toolchain/IDE: All toolchains);             Hardware:       - 2 to 5 FRDM-KW36 development board:  FRDM-KW36 Development Kit KW36/35 MCUs | NXP  Implementation This demo application is design for the FRDM-KW36 platform and can be easily integrated into any board that is using KW35/36 MCU family. The functionality is based on the coexistence mechanism available on the SDK (Mobile Wireless System - MWS module). Based on the HW link-layer implementation, the Bluetooth Low Energy has a higher priority than the GFSK protocol and as the effect, the GFSK communication is executed during the Idle states (inactive periods) of the Bluetooth LE.  For more details related to the MWS module, please refer to connectivity framework documentation from SDK (Connectivity Framework Reference Manual.pdf). As for functionality on the Bluetooth low energy, both roles, central and peripheral, are supported.  Integration to the KW36 SDK - download the attached file and unzip to ...\SDK_2.2.1_FRDM-KW36\boards\frdmkw36\wireless_examples\hybrid folder: - open IAR project (SDK_2.2.1_FRDM-KW36_2019_07_19\boards\frdmkw36\wireless_examples\hybrid\ble_w_uart_gfsk\freertos\iar\ble_w_uart_gfsk_freertos.eww). - the project is organized like below: Functionality Switches functionality:     - functionality is defined in main.c file, BleApp_Handle Keys function;    - on the FRDM-KW36 we have:                 - SW2 - start scanning - Central device;                 - Long SW2 - start advertising - Peripheral device; (long SW2 - SW2 pressed for more than 3 seconds)                 - SW3 - start/stop GFSK TX operation (advertising);                 - Long SW3 - start/stop GFSK RX operation (long SW3 - SW3 pressed for more than 3 seconds) Logs:    - Serial events for different states of the board;    - BaudRate 115200; Validation The solution has been validated using 1 Master and 4 Slave devices as below: 1. Create the network:     a. Open serial communication of all devices. After reset you will see the following message:    b. On the Central device press SW2 to start scanning;    c. On the Peripheral device press Long SW2 to start advertising and wait for the confirmation on the serial port:   d. Repeat steps b. and c. for all of the slave devices.   e. When the network is completed on the Central device you will see something like below:   f. Check the over the air connections (connection interval = 312.5 ms): 2. Validate functionality on the Bluetooth LE: - from each slave (Peripheral) serial terminal write a message (e.g: testslaveX) and check that the message is printed on the master serial port. - do the same test from the master (Central) serial terminal. - Below is an example of this step:   - over the air log: 3. Initiate GFSK communication: - in one of the board's press SW3 to start GFSK TX operation (Advertising packet with AdvAddress = 0909090909); At every 1 second (gGenFskApp_TxInterval_c), an ADV packet will be sent over the air. - Select other board and press Long Sw3 to initiate GFSK RX operation (RX interval = 100ms = gGenFskApp_RxInterval_c); - Each time an ADV packet from address = 0909090909 is received this will be listed on the serial port as below: - over the air the GFSK TX packets will be listed as a ADV_NONCONN_IND: 4. Validate Bluetooth LE in parallel with GFSK: - write a message on the Master (Central) serial terminal and check the feedback on the slave(Peripheral) serial terminals: Attached is the source code for this application. Regards, Ovidiu

Hello community, This time I bring to you a document which explains what are the ZigBee Test Client commands and how to use it. Before to read this guide, I widely recommend to take a look into the document Running a demo with BeeKit (802.15.4 MAC, SMAC and ZigBee BeeStack). This guide requires the BeeKit Wireless Connectivity Toolkit​ and the Test Tool for Connectivity Products applications.     I hope you find this guide useful. Enjoy this guide! Any feedback is welcome. Best regards, Earl Orlando Ramírez-Sánchez Technical Support Engineer NXP Semiconductors

BeeStack solutions included in BeeKit contain several low level drivers that definitely ease customer’s development phase.  Ranging from UART, SPI, NVM, I2C, among many others, these drivers could be used to interface the MW2x devices with different devices or sensors. It is also true these drivers will not support all custom applications by default, but they are conveniently provided in source code so anyone can modify them to the application’s needs. One example would be the need to use an accelerometer such as FXOS8700 or MMA8451. In this case, the default functionality of the I2C drivers might not be well-suited to work with these devices out-of-the-box. Nevertheless, this could be achieved with simple modifications to the source code. This project implements the basic I2C functionality to interface a TWR-KW24D512 board with a FXOS8700 sensor using the drivers included in Kinetis BeeStack Codebase 4.0.x solutions. The demo uses a ZigBee Home Automation GenericApp template to initialize and periodically read the accelerometer data X, Y and Z. A change in the registers read and written would be enough to use MMA8451 instead.  Following images illustrate the I2C frames obtained from the analyzer: FXOS8700 Initialization: Accelerometer X-Axis Data: Accelerometer Y-Axis Data: Accelerometer Z-Axis Data: IMPORTANT NOTE: Support of the attached project is limited. Please use this project as reference only. If it does not fulfill your requirements, you could always modify its source code to meet you application’s needs.

This article describes how to compile the Linux BSP of the i.MX platform under ubuntu 18.04, 20.04 LTS and debian-10. This is a necessary step to integrate WIFI/BT to the I.MX platform. See the attachment for detailed steps.

Hello community, This time I bring to you a document which explains how to run a demo from BeeKit and how to sniff it. Before to start you need to install the BeeKit Wireless Connectivity Toolkit.     I hope you find this guide useful. Enjoy this guide! Any feedback is welcome. Best regards, Earl Orlando Ramírez-Sánchez Technical Support Engineer NXP Semiconductors