| View previous topic :: View next topic |
| Author |
Message |
Hans Wedemeyer
Joined: 15 Sep 2003
Posts: 226
|
| PWM not reliable any suggestions |
 Posted: Fri Dec 30, 2005 1:34 pm |
 |
|
PWM not reliable on PIC18F6720
Microchip tells me there are no un-published errata about PWM issues.
Some chips simply do not output anything. I'm using CCP1 and CCP5
Failure rate is approaching 5% Units that work are reliable, units that fail can be spoted in the first few minutes of testing.
Chip replacement is a cure. I've tried about every software variation.
So far I have not put a "bad chip" I as a replacement.
TQFP are no fun to change, and it does stress the PCB making the circuit board vunerable for any future (hope never) chip changes.
I'd move to a different chip but that would limit my future firmware plans.
Any suggestions or similar experiences ?
Perhaps I should rig up a socketes board and pretest the chip before use.
Anyone know where I can get a TQFP 64 pin ZIF socket ?
Happy New Year.... |
|
 |
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Fri Dec 30, 2005 4:00 pm |
|
|
It might be useful to search for PWM problems in the Erratas for
other 18F chips.
Microchip took away their Advanced Search engine, so that leaves
Google. So go to Google Advance Search page:
http://www.google.com/advanced_search?hl=en
Then set it up like this to find erratas relating to PWM:
Find all the words: Errata PWM
Without the words: microsolutions
Return results in file format: PDF
Return results from domain: microchip.com
Here is one that looks interesting:
PIC18F6390/6490/8390/8490 Rev. B3 Silicon Errata (80207b.pdf)
| Quote: | 7. CCP
The CCP1 and CCP2 configured for PWM mode
with 1:1 Timer2 Prescaler and duty cycle set to
the period minus 1, may result in the PWM outputs
remaining at a logic low level.
(it goes on with more after this, and gives
a work-around). |
This is sort of the type of problem that you're having.
i.e.: PWM shuts down for weird reason.
You could look at that errata and the rest of the 18F (or even 16F)
erratas that Google pops up, and see if they might be the cause
of your problem. My logic here is that a bug recognized in one
PIC might well exist in other PICs, but it just hasn't been verified
or officially acknowledged yet. You might be the first one to find
it in your PIC. |
|
|
ckielstra
Joined: 18 Mar 2004
Posts: 3680
Location: The Netherlands
|
|
|
Hans Wedemeyer
Joined: 15 Sep 2003
Posts: 226
|
| Thanks good tip |
| Posted: Sat Dec 31, 2005 6:51 am |
|
|
Thanks for the tip, I keep forgetting google...
This project started out with PIC18F6722 and after nightmares with the EUSART and contacting Microchip they recommended I use the older PIC18F6720 with "normal" UART. Of course I lost any ability to do auto Baud rate, but the first few batches never displayed a problem with PWM.
It is also interesting how one person at Microchip tells me about Silicon Rev. A5, and another denies it !
Microchip now recommends I move to the PIC18F6722 ! Ha Ha... that errata sheet reads like Al Capon's crime sheet, and the EUSART still has not been fixed ! In fact I think it's still the same errata sheet from 2004.
As the PWM issue is easy to detect, testing seems to be the next route. I am plannig to sell the chips preprogrammed so that gives me another reason for going down that road...
Thanks again...
Hans W |
|
|
Hans Wedemeyer
Joined: 15 Sep 2003
Posts: 226
|
| Great... just what I needed. Thanks.... |
| Posted: Sat Dec 31, 2005 6:52 am |
|
|
|
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
