dsPIC33EP128MC502 PWM模块被禁用

六羟甲基三聚氰胺六甲醚。。。在我的DSPIC33 EP256MC502上，PWM3输出是在RB10和RB11（引脚21和22的28引脚封装）。因为你的芯片和我的数据表一样，我想知道，我的意思是，它是一个16位的芯片。甚至没有RB18，RB19。无论如何，我没有仔细看你的代码，也没有尝试过任何测试它，但是：你的意思是要设置PDC3在最后一个声明？问候，戴夫



      以下为原文

   
Hmmm... On my dsPIC33EP256MC502, PWM3 output is on RB10 and RB11 (Pins 21 and 22 of the 28-pin package).  Since your chip shares the same data sheet as mine, I'm wondering...
 
I mean, it's a 16-bit chip.  Wouldn't even have RB18, RB19.
 
Anyhow, I haven't looked very carefully at your code, and haven't tried any testing with it but:

Didn't you mean to set PDC3 in that last statement?
 
Regards,
 
Dave

如上所述，没有标记为RB18或RB19的PIN。大概您所说的是：PIN 21：TDO/RP42/RB10和PIN 22：TDI/RP43/RB11，您是如何禁用JTAG的？



      以下为原文

    As above, there are no pins labelled RB18 or RB19.
 
Presumably you are talking about:
Pin 21: TDO/RP42/RB10
and
Pin 22: TDI/RP43/RB11
 

What exactly did you do to disable JTAG?
 

亲爱的DaveKW7X和Qub，抱歉我的TyPo。它应该是PDC3= 150..是的，PWM3应该是PIN 18（RB10）和PIN 19（RB11）。我用下面的代码禁用JTAG。它与软件无关。现在，我可以从这两个引脚得到PWM输出。但新的问题是，我想交换PWM3H和PWM3L，也就是说，如果旧的PWM3H高，旧的PWM3L低，我希望它们的输出是相反的。代码为ICON3BITS，SWAP＝1；测试结果为，旧占空比为15%，新占空比仍为15%。也就是说，PWM3H和PWM3L没有交换，事实上，PWM2也是如此。你知道原因吗？谢谢，酷乐



      以下为原文

    Dear Davekw7x and qub,
 
Sorry for my typo.
it should be PDC3 =150.
Yes, the PWM3 should be Pin 18(RB10) and Pin 19(RB11).
I use the below code to disable JTAG.
#pragma config JTAGEN=OFF    //JTAG Enable bit(JTAG is disabled)
 
The above issue is from hardware. it is not related to software. right now, I can get the PWM output from these two pins. But the new issue is,
I would like to swap PWM3H and PWM3L. that's to say, if old PWM3H is High and old PWM3L is low, I would like the output of them is reverse. the code is,
 
IOCON3bits.SWAP=1;
 
The testing result is, the old duty cycle is 15% and the new duty cycle still is 15%. that's to say, there is no swapping of PWM3H and PWM3L. in fact, the PWM2 is also so. do you know the reason for it? Thanks,
 
-Coolhe
 

在更改IOCONx之前，您是否发出了需要的解锁序列？如果你这样做了，请发布你的代码的一部分。问候，戴夫



      以下为原文

   
Did you issue the unlock sequence that is required before making changes to IOCONx?
 
If you did, please post that part of your code.
 
 
Regards,

Dave

你好，戴夫，是的，我把PWMKEY解锁为下面的代码。PTENP.PTEN＝0；//PWM模块被禁用。PTPH＝1000；/ /解锁PWMLOK。CON2= 0x003；pWMKEY＝0x4321；IOCON2= 0xC02；/0xF000阶段＝0；PdC2＝150；PDC3＝150；DTR2= DTR3＝0；pWMCON2= pWMCON3= 0x000；//PTCON2= 0；PTCON＝0x8000；/ /启用PWM。（1）；PWMKEY＝0xABCD；PWMKEY＝0x4321；FCLKC:PWM2的占空比测试结果总是15%。



      以下为原文

    Hi Dave,
 
Yes, I unlock the  PWMKEY as the below code. 
*****************************************************************888
//PWM Setting.
PTCONbits.PTEN = 0; //PWM module is disabled.
PTPER=1000;
//Unlock the PWMLOCK.
PWMKEY = 0xabcd;
PWMKEY = 0x4321;
FCLCON2 = 0x0003;
PWMKEY = 0xabcd;
PWMKEY = 0x4321;
IOCON2 = 0xC002; //0xF000

PHASE2 = 0;
PHASE3 = 0;
PDC2= 150;
PDC3= 150;
DTR2 = DTR3 = 0;
ALTDTR2 = ALTDTR3 = 0;
PWMCON2=PWMCON3=0x0000; //
PTCON2=0;

PTCON=0x8000; //Enable PWM.
while(1);
*************************************************
The test result of duty cycle on PWM2 is always 15%.
 
-Coolhe
 
 

我在优化级别0中使用XC16的经验是，这些方法不起作用。编译器在序列中生成的额外指令将PooCh拧紧。在优化级别设置为1（我似乎记得它确实如此）时，它可能会或可能无法工作，但我不指望它。我可以想出几个更合理的方法：做到这一点：这是可以的快速和简单的测试：不需要任何解锁序列。但是在一个“真实”的应用程序中，您可能不想禁用由PWMLIK提供的保护。不小心的占空比变化可能是灾难性的。使用数据表中显示的汇编语言序列。有点笨拙，如果这是唯一的方法，那么它是有效的。（保存最好的最后一个）使用做这个动作的内置函数（在编译器用户指南中查找）这是我的初始化序列的相关部分，然后是我的测试（）：我可以看到每一秒的占空比变化。在我的DSPIC33 EP256MC502上使用XC16版本1.33，优化级别0和1），戴夫



      以下为原文

   
 
My experience with XC16 in optimization level 0 is that these don't work.  Extra instructions that the compiler generates in the sequence screw the pooch.  It may or may not work with optimization level set to 1 (I seem to recall that it does), but I wouldn't count on it.
 
I can think of several more reasonable approaches:

• Do this:
  #pragma config PWMLOCK = OFF            // PWM Lock Enable bit (PWM registers may be written without key sequence)
  
  This is OK for quick and easy testing: Doesn't need any unlock sequence.  But in a "real" application you might not want to disable the protection offered by PWMLOCK.  Inadvertent change of duty cycle could be disastrous.
  
• Use the assembly language sequences shown in the Data Sheet. Kind of awkward, and if this were the only way, well, it works.
  
• (Saved the best for last)
  Use the builtin function that does the deed (Look it up in the Compiler User's Guide)
  

Here's the relevant part of my initialization sequence
    // The C assignment sequence doesn't work with XC16 version 1.33,
    // at optimization level 0, so use the builtin
    //PWMKEY = 0xABCD;
    //PWMKEY = 0x4321;
    //IOCON2 = 0xC000;
    __builtin_write_PWMSFR(&IOCON2, 0xC000, &PWMKEY);
 
    //PWMKEY = 0xABCD;
    //PWMKEY = 0x4321;
    //FCLCON2 = 0x0003;
    __builtin_write_PWMSFR(&FCLCON2, 0x0003, &PWMKEY);
    // Other initialization stuff
   
 
Then my test in main(): I can see the duty cycle change every second.
    // Initialize clock, I/O, Uart, PWM, etc...
    init_system();
    
    while (1) {
        uint16_t temp = IOCON2;
        temp ^= 0x0002;
        __builtin_write_PWMSFR(&IOCON2, temp, &PWMKEY);
        //IOCON2 ^= 0x0002;
        LED1 ^= 1;
        __delay_ms(1000);       
    } // End of main while(1){} loop

(Tested on my dsPIC33EP256MC502 with XC16 version 1.33, optimization level 0 and 1)
 
 
Regards,

Dave
 

嗨，戴夫，你真的很专业。你的建议很好。现在，我可以用我的思想灵活地控制PWM输出。再次谢谢。-酷！



      以下为原文

    Hi Dave,
 
You are really professional. your suggestions are good. right now, I can flexibly control the PWM output with my thought.
Thank you again.
 
-Cool