ICD-U64 & PIC16F18855 (also posted in the CCS ICD forum)

guy · Joined: 21 Oct 2005 Posts: 297

Has anyone tried to debug with ICD-U64 & PIC16F18855 ?
I've been having trouble and I'm not sure if it has to do with my equipment or a CCS issue.

Ttelmah · Joined: 11 Mar 2010 Posts: 19477

I had problems with a PIC24FJ512GB606. Another similarly 'new' chip.

First, check what version firmware you have in the ICD, and try updating to the latest if you haven't already got this. Also, what compiler version?. Apparently there are some changes in the IDE that also affect handling for the latest chips (this was a problem for me, as 5.072/3 introduced a problem with the PIC24, which was not there in 5.070 - touch wood, 5.074 seems to have this fixed).

If you still have the problem with the latest firmware and IDE, then talk to CCS. They sent me a beta firmware, which at present is working OK.

guy · Joined: 21 Oct 2005 Posts: 297

Hi Mr. T.
Did you switch to CCS ICD?!
I've been in contact with them, also tried the beta firmware, 5.074 etc. but no luck so far.

Ttelmah · Joined: 11 Mar 2010 Posts: 19477

I have a number of debuggers. Pickit3, ICD-U40, ICD-U64, ICD2, ICD3, and Mach-X (run as a debugger). Currently this latter is the one that works reliably!...

The Pickit3 works from MPLAB-X, but the package gives problems everywhere.
This debugger hangs intermittently if used from the CCS IDE.
The CCS ICD-U40, and the MicroChip ICD2, don't support this chip.
The U64, run from MPLAB crashes. Run from the CCS IDE, it works, but intermittently hangs when programming the chip. However 90% of the time it is OK.
Mach-X works from the CCS IDE, and has been fine.

Switch down to an older chip, and the crashes/hangs disappear.

guy · Joined: 21 Oct 2005 Posts: 297

This is a useful review for many folks! Although it might not be pinned on the top of the forum Razz

Still waiting to hear if someone got the 16F18855 to work with one of CCS's debuggers??

Ttelmah · Joined: 11 Mar 2010 Posts: 19477

What is your unit actually doing?.
One thing I found, was I had to shorten the supplied cable. The new chips support much faster programming algorithms than the older units, and the original cable (about 150mm), was causing noticeable signal degradation. I made one as short as I could (only about 75mm), and things improved noticeably when this was used....

guy · Joined: 21 Oct 2005 Posts: 297

CCS managed to fix the problem mostly and it will be incorporated in the next version. I think the situation now is like you describe Ttelmah - the debugger works but will occasionally lose comm.
I hate shortening the RJ11 cable since 6-wire RJ11s are hard to come by... I might cut it in one of my rage bursts if the debugger needs restarting Twisted Evil