BUG：ICD4代码在调试时工作，但在ICD4供电时不是Ru

是的，MCLR有一个上拉电阻。我从来没有'启用' MCLR引脚之前..这是一个新的…你必须做什么？



      以下为原文

    Yes, MCLR has a pullup resistor.   I've never 'enabled' the MCLR pin before..thats a new one.. what do you have to do?

抱歉，忽略了我的评论。大多数PIC有一个配置，但是允许MCLR引脚被禁用，并用作通用输入。然而，PIC32 MZ有这么多的引脚，它们不费心这样做，你只需要一个配置位，让MCLR可以：（1）产生一个上电复位，或者（2）。生成一个正常的系统复位。



      以下为原文

   
Sorry, disregard that part of my comment.
Most PIC's have a CONFIG but that allows the MCLR pin to be disabled, and used as a general purpose input.
However, PIC32MZ have so many pins, they don't bother doing that, you just have a config bit that lets MCLR either:
[1] Generate a power on reset, or
[2] Generate an normal system reset.
 

我怀疑这里的问题是我使用的是一个为电路提供电源的ICD4。当ICD4在编程之后继续提供电路的电源时，它不释放MCLR。



      以下为原文

    I suspect the problem here is that i was using a ICD4 that is providing power to the circuit. WHile the ICD4 continues to provide power to the circuit after it is programed, its not releasing the MCLR.
 

工具栏上有一个按钮，可以在运行模式下释放MCLR。



      以下为原文

    There's a button on the toolbar to release MCLR in run mode.

“它不发布MCLR”-调试器通过在编程部分之后声明MCLR低来重置芯片。在那之后，它应该独自离开。如果您怀疑除了调试器之外设备没有被供电，您应该检查它的VDD引脚。



      以下为原文

    "it is not releasing the MCLR" - The debugger resets the chip by asserting MCLR low after programming the part.  It should leave it alone after that.  
 
If you suspect that the device isn't being powered except from the debugger, you should check its Vdd pins.

这个设备绝对是由调试器供电的。在这种情况下没有其他电源，我把一个作用域连接到MCLR管脚上，并附上ICD。在编程之前它的位置很低。当编程开始时，（你听到点击ICD4）MCLR Byffy走高。编程继续，并完成，你听到另一个点击从ICD4，然后在一个短暂的延迟（也许5秒左右）MCLR高达约1秒，然后返回低，并保持在那里。VDD在整个过程中被提供和恒定。（在3.3.2V，MCLR保持低而不返回高，它永远不会开始。ICD4被设置为既能对目标供电，又能保持有功功率。作为检查，我在最终的预期模式中提供了设备（提供其自身的功率），并且它按照预期运行。MCLR引脚有一个12K1的上拉电阻，它与以前的所有工作一起工作，当它自己供电时，它也工作得很好。这似乎是ICD4工作方式的一个问题。这是可配置的，我找不到一种方法来重新配置。我想使用它从ICD4供电的原因是，这个电路是由一个非隔离的POE接口供电，它可以在-52V浮动相对于ICD的电源…一个坏的组合，给ICD带来死亡和破坏。（它确实对ICD3）。我确实有一个ItoNA USB隔离器，当我们绝对必须这样做，但是。



      以下为原文

    The device absolutely is being powered by the debugger as it sits.  there is no other power supply in this case. 

I attached a scope to the MCLR pin, with the ICD attached. Before programming it sits low.  When programming commences, ( you hear the click on the ICD4 ) MCLR breifly goes high.  Programming continues, and completes, you hear another click from the ICD4, and then after a short delay ( perhaps .5s or so ) MCLR goes high for approximately 1 second and then returns low, and remains there.     VDD is supplied and constant throughout the entire process.( at ~3.2V
 
With MCLR remaining low and not returning high, it never gets started.
                                                                               
The ICD4 is set to Both Power the target and Maintain active power. 
 
As a check, i powered the device up in its final intended mode ( providing its own power ) and it runs as expected.  The MCLR pin has a 12k1 pull up resistor, which has worked with everyrthing previously, and when self powered it works just fine as well.



This appears to be a problem with the way in which the ICD4 is operating?  Is this configurable, i can't find a way to reconfigure.
 
The reason i want to use it powered from the ICD4 is that this circuit is powered by a Non Isolated POE interface, and it can be floating at -52V with respect to the ICD's power supply...  A bad combination, which brings death and destruction to the ICD.. ( well it did to ICD3 )..  I do have a Intona USB isolater for when we absolyutely have to do this however.
 
 
 
 
 
 
 
 
 
 

只是为了比较，如果你把ICD4连接到PIC32 EF启动器套件，通过J7，就会发生同样的事情。



      以下为原文

    Just for a comparison, the same thing happens if you attach the ICD4 to a PIC32EF starter Kit, via J7.. 
 

NB…你不能在微芯片支持中记录一个ICD4的情况，因为支持工具的下拉列表还没有被更新以包括它。



      以下为原文

    NB.. You cna't log a case in microchip support for an ICD4 yet, becuase the drop down list of supported tools has not yet been updated to include it.
 
 

我猜你忽略了我的帖子。这个按钮叫做“关机复位”。如果按下这个按钮，程序员将按照你所描述的那样行动。



      以下为原文

    I guess you ignored my post.

 
The button is called "hold in reset". If that button is pressed, the programmer will act exactly as you described...
 
 
 

      以下为原文

   

按钮被称为“保持复位”。如果按下该按钮，程序员将按您所描述的那样操作…按钮不被按下，并且如果按下该按钮，它的行为没有区别。这种行为对于ICD4来说是独一无二的，这是由ICD3或PICTIT3引起的。它似乎不太可能在ICD4中存在硬件错误，因为它在调试模式下发布MCLR。但是无论如何谢谢您的贡献。



      以下为原文

   
 The button is called "hold in reset". If that button is pressed, the programmer will act exactly as you described...
 



The button is not pressed, and it makes no difference to its behavior if it is pressed.  This behavior is unique to the ICD4,  this did'tn occur with an ICD3,  or a Pickit3.   
 
It does not seem likely that there is a hardware bug in the ICD4, as it it releases MCLR when it is in debugging mode.


but thanks for your contribution anyway.
 

是的，正如你所说的，它必须是ICD4中的一个bug。对其他程序员很好。：）



      以下为原文

    Yeah, as you say, it must be a bug in the ICD4. Works fine on other programmers. :)
 

刚刚收到我们当地销售代表的询问，工厂反应：“我们知道这个问题，并致力于解决这个问题。”应该马上就来。“



      以下为原文

    Just received from our Local Sales Rep, who enquired about this,  
 
Factory response: “We are aware of this issue and working on fixing this.  Should be in shortly. “