|
Hi We are using the IAR workbench and Jlink to develop code for the LPC2129. I must use the PLL and debug from FLASH as my timing is criticle but the use of the PLL is preventing the Jlink connecting. The only way around this is to use the Philips Flash utility to erase the device each time before I start the debugger. On the development board this is a pain, but I would prefer our target board not to have a serial port if possible. Firstly, has anyone managed to get around this problem? (The recent Jlink DLL patch has not helped.) Secondly, has anyone used an alternative debugger with IAR that does not have this problem? Any comments would be greatly appreciated. best regards Simon |
|
|
IAR, JLINK and PLL problems
Started by ●September 4, 2004
Reply by ●September 5, 20042004-09-05
|
--- In , "itsjustimpossible" <simonjh@b...> wrote: > Hi > We are using the IAR workbench and Jlink to develop code for the > LPC2129. I must use the PLL and debug from FLASH as my timing is > criticle but the use of the PLL is preventing the Jlink connecting. > The only way around this is to use the Philips Flash utility to > erase the device each time before I start the debugger. > On the development board this is a pain, but I would prefer our > target board not to have a serial port if possible. > > Firstly, has anyone managed to get around this problem? (The recent > Jlink DLL patch has not helped.) > > Secondly, has anyone used an alternative debugger with IAR that does > not have this problem? > > Any comments would be greatly appreciated. > > best regards > Simon We had the same problem. Once you have programmed the flash with the pll enabled the processor is history as far as the debugger is concerned until you scrub the flash. I found jlink will work ok with pll running if you set the following in the jlink menu of the debugger. 1. tick philips lpc download button 2. tick reset aserts hardware reset pin 3. although this is not the case set the cclk to whatever you are running the cpu clock we have a 10mhz crystal pumped up to 60 mhz by the pll. In theory we should set the cclk setting to 10000 in the jlink menu. This doesn't work. If we set it to 60000 we have never had a problem!!!! regards Doug |
|
|
Reply by ●September 8, 20042004-09-08
|
Hi Doug Thanks for the info, and it has helped in some circumstances but I have to manually reset the board before I try to enter the debugger. I am now wondering if the board is resetting properly. The Keil evaluation board we are testing with has no reset IC, just an RC network. Are you using a reset IC on your target board? cheers Simon --- In , "douglasbolton" <doug@c...> wrote: > --- In , "itsjustimpossible" <simonjh@b...> > wrote: > > Hi > > We are using the IAR workbench and Jlink to develop code for the > > LPC2129. I must use the PLL and debug from FLASH as my timing is > > criticle but the use of the PLL is preventing the Jlink connecting. <snip> > > We had the same problem. Once you have programmed the flash with the > pll enabled the processor is history as far as the debugger is > concerned until you scrub the flash. > > I found jlink will work ok with pll running if you set the following > in the jlink menu of the debugger. > > 1. tick philips lpc download button > > 2. tick reset aserts hardware reset pin > > 3. although this is not the case set the cclk to whatever you are > running the cpu clock > > we have a 10mhz crystal pumped up to 60 mhz by the pll. In theory we > should set the cclk setting to 10000 in the jlink menu. This doesn't > work. If we set it to 60000 we have never had a problem!!!! > > regards Doug |
|
|
Reply by ●September 8, 20042004-09-08
|
--- In , "itsjustimpossible" <simonjh@b...> wrote: > Hi Doug > Thanks for the info, and it has helped in some circumstances but I > have to manually reset the board before I try to enter the debugger. > > I am now wondering if the board is resetting properly. The Keil > evaluation board we are testing with has no reset IC, just an RC > network. > > Are you using a reset IC on your target board? > > cheers > Simon > --- In , "douglasbolton" <doug@c...> wrote: > > --- In , "itsjustimpossible" <simonjh@b...> > > wrote: > > > Hi > > > We are using the IAR workbench and Jlink to develop code for the > > > LPC2129. I must use the PLL and debug from FLASH as my timing is > > > criticle but the use of the PLL is preventing the Jlink > connecting. > <snip> > > > > We had the same problem. Once you have programmed the flash with > the > > pll enabled the processor is history as far as the debugger is > > concerned until you scrub the flash. > > > > I found jlink will work ok with pll running if you set the > following > > in the jlink menu of the debugger. > > > > 1. tick philips lpc download button > > > > 2. tick reset aserts hardware reset pin > > > > 3. although this is not the case set the cclk to whatever you are > > running the cpu clock > > > > we have a 10mhz crystal pumped up to 60 mhz by the pll. In theory > we > > should set the cclk setting to 10000 in the jlink menu. This > doesn't > > work. If we set it to 60000 we have never had a problem!!!! > > > > regards Doug Yes we had to put a reset device in, we use a max809 (140mS delay) and haven't had any problems. With an RC network we had trouble with the debugger and actual boot ups of the board in standalone. |
|
|
Reply by ●September 10, 20042004-09-10
|
--- In , "douglasbolton" <doug@c...> wrote: > --- In , "itsjustimpossible" <simonjh@b...> > wrote: > > Are you using a reset IC on your target board? > > Yes we had to put a reset device in, we use a max809 (140mS delay) > and haven't had any problems. With an RC network we had trouble with > the debugger and actual boot ups of the board in standalone. Great, I will give it a try. Thanks Simon |
|
|
Reply by ●September 10, 20042004-09-10
|
Reset duration needs to be 10ms at power-up and 300ns thereafter. I have seen problems with boards where the RC reset circuit provided a reset pulse on power-up which was too short. Rich --- In , "itsjustimpossible" <simonjh@b...> wrote: > --- In , "douglasbolton" <doug@c...> wrote: > > --- In , "itsjustimpossible" <simonjh@b...> > > wrote: > > > Are you using a reset IC on your target board? > > > > > > > > > Yes we had to put a reset device in, we use a max809 (140mS delay) > > and haven't had any problems. With an RC network we had trouble > with > > the debugger and actual boot ups of the board in standalone. > > Great, I will give it a try. > > Thanks > Simon |
