View previous topic :: View next topic |
Author |
Message |
hmmpic
Joined: 09 Mar 2010 Posts: 314 Location: Denmark
|
ICD64 blocking clk & data after programming? |
Posted: Sat Nov 04, 2017 11:56 am |
|
|
I use CCS and ICD64 Rev2.
Why is the pin clk & data not released (high-Z) after programming?
I normally use the pin to other function too with 1k in series to other, and direct to the ICD programmer.
When using "CCS device programmer" software and select "RUN" then all is nice, but not from CCS IDE it is blocking the pin.
I can see if using it as debugger then it must block and use the pin, but not when just using it as programmer.
Any hints? |
|
|
temtronic
Joined: 01 Jul 2010 Posts: 9229 Location: Greensville,Ontario
|
|
Posted: Sat Nov 04, 2017 1:18 pm |
|
|
I never use an 'IDE' but would assume it would set 'bits' so it can take 'control' of the PIC.
I know MPLAB has 2 build configurations, debug or release.
If you select debug( the factory default) then your code will NOT work in the 'Real World'. For that to happen you must select 'release' and recompile.
Perhaps ther's a 'command' or 'bullet' or 'something' to deselect the IDE 'box' from being and IDE ?? |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19520
|
|
Posted: Sat Nov 04, 2017 1:39 pm |
|
|
What chip?.
What compiler version?.
The IDE normally uses CCSLoad to actually do the programming (via it's command line). and the code will usually run after the chip is programmed, unless the debugger is enabled.
You have got the ICD disabled?. The debug window needs to be closed, and ICD turned off in the code. |
|
|
hmmpic
Joined: 09 Mar 2010 Posts: 314 Location: Denmark
|
|
Posted: Sat Nov 04, 2017 1:55 pm |
|
|
Pic is, 16f1825.
CCS IDE and compiler is: 5066. My maintenance is out therefore no update.
As i remember i have used it as described before without any problem. Maybe it was an older compiler version and/or older firmware in the ICD. |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19520
|
|
Posted: Sat Nov 04, 2017 2:00 pm |
|
|
You still will have support from CCS. Point out the problem. It certainly should work. Have you updated CCSLoad to the latest version? |
|
|
hmmpic
Joined: 09 Mar 2010 Posts: 314 Location: Denmark
|
|
Posted: Sat Nov 04, 2017 2:03 pm |
|
|
CCSLoad and ICD firmware is updated after i found the problem, no change it is the same. |
|
|
newguy
Joined: 24 Jun 2004 Posts: 1908
|
|
Posted: Sat Nov 04, 2017 5:33 pm |
|
|
Try rolling back ICD FW to an earlier version. |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19520
|
|
Posted: Sun Nov 05, 2017 3:49 am |
|
|
Yes.
I'd suggest uninstalling, and re-installing the 5.060, with the CCSload that comes with it and the firmware, and trying with this (don't let it update from the web, and install this onto the ICD).
I think if this had been a common problem at the time, other people would have reported it. There have been a lot of updates since, with new chips added etc., and sounds as if one may be giving problems with the older compiler. I had the 'opposite' recently, with a new compiler introducing problems, and having to run a couple of versions 'back in time' for the compiler, till these were fixed, but needing the very latest CCSLoad to make the programming work properly. |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19520
|
|
Posted: Sun Nov 05, 2017 1:47 pm |
|
|
It's worth also perhaps saying that you can 'work round' very easily.
Just launch CCSLoad and open the output file. Work in the CCS IDE, and when you compile the file will automatically update in CCSLoad. Hit the program button in CCSLoad to write, instead of the button in the IDE. |
|
|
newguy
Joined: 24 Jun 2004 Posts: 1908
|
|
Posted: Sun Nov 05, 2017 6:40 pm |
|
|
Ttelmah wrote: | It's worth also perhaps saying that you can 'work round' very easily.
Just launch CCSLoad and open the output file. Work in the CCS IDE, and when you compile the file will automatically update in CCSLoad. Hit the program button in CCSLoad to write, instead of the button in the IDE. |
This is my standard procedure. But it's also my standard procedure to leave the programming lines alone - no other connection other than to the ICD, so I wouldn't know if the programmer wasn't fully releasing the lines.
I have been bitten by blindly accepting an ICD update before, though. Ever since, I never accept the CCS update without testing first. Nothing quite so panic inducing as trying to update a product in front of customers and finding that the product is "bricked". |
|
|
hmmpic
Joined: 09 Mar 2010 Posts: 314 Location: Denmark
|
|
Posted: Tue Nov 07, 2017 6:53 am |
|
|
Thanks for all the nice reply.
Problem is at CCS support and they are working on a solution. |
|
|
temtronic
Joined: 01 Jul 2010 Posts: 9229 Location: Greensville,Ontario
|
|
Posted: Wed Nov 08, 2017 6:21 am |
|
|
I'm with newguy on the 'pin problem'. Since PICs are so cheap, I also leave the pgm pins alone and if need be, use a 'bigger' PIC.
While it costs a few pennies more for the bigger PIC, it eliminates a nasty problem that could cost you a LOT of time (R&D $$$), rework PCBs ($$$) or a missed dedline (lose client).
I don't know if the 'core' of the ICSP is exactly the same from PIC to PIC, across ALL families. If it's not( say change in pullups) THAT could lead you to losing hair ! |
|
|
|