怎么切换SPI速度

应该是好的。



      以下为原文

    Should be fine.
 

您可能希望在您的初始硬件设置中设置用于SPI引脚的LAT和TRIS，这样当禁用/启用SPI模块时不会出现故障。这真的不应该是一个问题，因为奴隶芯片的选择应该是高的，但人们永远不会知道。



      以下为原文

    you may want to set up the LAT & TRIS for the SPI pins in your initial hardware setup so that there are no glitches when you disable/enable the SPI module.  It really should not be a problem since the slaves Chip Selects should be high, but one can never tell.

嗨，不知道为什么你做…..（0）一个宏将只与括号一起工作。这是标准C。如果你想让它更具可读性，可以在回车之前使用反斜线来跨越很多行。你还没有识别出你的芯片。张贴在PIC24中，值得一提的是，很多PIC24s不能在SPI上实现10MHz。除数可以设置为这个，但数据表有一个警告，通知你限制到8MHz（由于PPS）。通常只有专用PIN SPI端口可以达到8～9MHz以上。只要检查一下你的数据表就行了。Yorky。



      以下为原文

    Hi,
Not sure why you have do..while(0)  a macro will work with just the brackets. This is standard C.
Also if you want to make it a bit more readable the back slash can be used prior to Carriage Return to spread across numerous lines.
You haven't identified your chip. Posted in the PIC24, just worth mentioning that many PIC24s can not achieve 10MHz on the SPI. The divisors can be set to this but the data sheets have a warning that informs you to limit to 8MHz (due to PPS). Normally only dedicated pin SPI ports can achieve above 8-9MHz. Just worth checking your data sheet.
 
T Yorky.

我认为DO { }（0）是原始微芯片宏的一部分。这是一个宏技术，确保它被当作一个语句，所以可以遵循和（）等，并允许一个半冒号跟踪宏。直{}通常不遵循A；



      以下为原文

    I think the do {} while(0) was part of the original microchip macros.  It's a macro technique that ensures its treated as one statement, so can follow and if() etc and allows a semi-colon to follow the macro. Straight {} is not normally followed by a ;

嗨，作为一个例子，下面是从一个工作项目中定义的一些代码……定义…//-------------------------------------------------------------------------------//^。为了保证类似的模拟/ /或销钉不会干扰引导加载程序使用的PIN／LED接口。＝0B000 000 000亿；\ANSELG= 0B000 000亿，}，并在应用程序中使用…//+------------------------------------------------------------+//~。EADER用于建立CPU和外围设备。类似的PrSimaCurrPuthCuuluxA类似物；//用于串行TX/RX UMeCyActhOngOnTracxIDLeTeX的设备特定配置引脚；MaxActoNo.StimxPotoWrand；MaultActuurPixFordTX；MaulActhOngSeRtxPrtoWrand；宏配置PixFrrx；/ /设备特定宏，用于配置状态DeLeMeCalm的PINIONIOSSEDLID.IDLeTalm；MaulActhOngStEnLeDig.PotoWrand；返回；}尝试它。它有效。T Yorky。



      以下为原文

    Hi,
As an example, below is some code cut from a working project...
definition...
//+---------------------------------------------------------------------------------------+
//| Device action to configure the analogue pins to digitals      |
//| Note that the Boot Loader does not use analogues and is provided to ensure analogue |
//| pins do not interfere with pins used by the boot loader comms/LED interface.  |
//+---------------------------------------------------------------------------------------+
#define macroCPU_Configure_Analogues { ANSELB = 0b0000000000001100;
           ANSELC = 0b0000000000000000;
           ANSELD = 0b0000000000000000;
           ANSELE = 0b0000000000000000;
           ANSELG = 0b0000000000000000; }
 
And use in the application...
 
//+---------------------------------------------------------------------------------------+
//| Initialise the device to operate as the Boot Loader       |
//| Macros defined in the DeviceMap header are used to set up the CPU and peripherals. |
//|                  |
void InitialiseChipAndPeripherals(void)
{
  // Device specific macro to disable all Interrupts
 macroCPU_Disable_All_Interrupts;
  // Device specific macro to configure the clock operation
 macroCPU_Configure_Clock;
  // Device specific macro to configure the analogue pins
 macroCPU_Configure_Analogues;
  // Device specific configure pins for Serial Tx/Rx use
 macroAction_SetTxIdleState;
 macroAction_SetTxPortDirection;
 macroConfigurePinForTx;
 macroAction_SetRxPortDirection;
 macroConfigurePinForRx;
  // Device specific macro to configure pin for an status LED
 macroAction_SetLED_IdleState;
 macroAction_SetLED_PortDirection;
return;
}
 
Try it. It works.
T Yorky.

关于SPI速度不工作。你应该用一个范围来查看SPI时钟和数据线。将额外的设备添加到总线将增加电容，并减缓上升时间。许多PIC芯片没有好的驱动程序。我有一个PIC18F26K22项目，它驱动SPI总线上的MPC22S17 I/O扩展器。它在4 MHz驱动9个I/O扩展器上工作良好。我需要把扩张器的数量增加到26个，而且太多了。我在SPI线路中添加了A7HC125缓冲器，大大提高了上升时间。我尝试了7VHC125芯片，它们太快了，我在公共汽车线路上打过电话。幸好我登上了面包板！我的新板卡是今天从FAB交货的（4月1日是的！）把这个范围拿出来！我有一个安捷伦/KiVIEW DSO-2014A，它解码SPI，I2C，CAN和串行。爱它！-克雷格



      以下为原文

    Regarding the SPI speeds not working.  You should look at the SPI clock and data lines with a scope.  Adding additional devices to the bus will add capacitance and slow down the rise times.  Many of the PIC chips do not have good drivers.  I have a PIC18F26K22 project that drives MPC22S17 I/O expanders on the SPI bus.  It worked well at 4 MHz driving 9 I/O expanders.  I needed to increase the number of expanders to 26 and that loaded down the bus too much.  I have added a 74HC125 buffer to the SPI lines and they greatly improved the rise times.  I tried 74VHC125 chips and they were too fast, I had ringing on the bus lines.  Its a good thing I bread boarded that!  My new boards are due from fab today (yes on April 1st!).
 
So get the scope out!  I have a Agilent/Keysight DSO-X 2014A and it decodes SPI, I2C, CAN and serial.  Love it!
 
- Craig

不要忘记SPI时钟速度只与主站和主站之间的交换有关。单个位在上升或下降位上交换（由CKP和CKE位设置决定），它们可以按照您希望的那样发生。（当然，你必须遵守最大时钟频率规格）（这是为什么不使用从SS的奴隶信号可能是如此糟糕的一个原因-如果奴隶看到SCK线路上的噪声，那么它可以被解释为时钟脉冲，这意味着奴隶与主机失去同步。从属设备不关心时钟脉冲之间经过多长时间，它只是跟随。）因此，您可以随时切换速度（只要主机可以处理），从属者将跟随。此外，交换机的速度将与以太网E的速度无关。xchange用于分组，但可以对分组之间的时间产生影响。苏珊



      以下为原文

    Don't forget that the SPI clock speed only relates to the exchange between the master and the slave. The individual bits are exchanged on the rising or falling bits (as determined by the CKP and CKE bit settings) and they can occur as far apart as you want them to be. (Of course you have to observe the maximum clock frequency specs.)
(This is one of the reasons why not using the SS signal for the slave can be so bad - if the slave sees noise on the SCK line then it can be interpreted as a clock pulse which means the slave gets out of sync with the master. The slave does not care how long a time elapses between clock pulses, it just follows along.)
Therefore you can switch speed any time you like (as long as the master can handle it) and the slave will tag along.
Further, the speed of the exchange will have little to do with the speed of the Ethernet exchange for a packet, but can have an impact on the time between packets.
Susan