Possible chip definition file error

22 views
Skip to first unread message

Mike

unread,
Jul 19, 2022, 8:24:06 PM7/19/22
to jallib
Hi Rob,

I say possible, because I've messed up on this before.  But...I'm pretty sure the chip definition file for the 18F26Q43 has definitions for PORTD and LATD that don't exist on that chip.  The other 28-pin chips in this family probably have the same error: 18F25Q43 and 18F27Q43.  Am I wrong again?

Regards,
Mike

Mike

unread,
Jul 19, 2022, 9:48:38 PM7/19/22
to jallib
Sorry, instead of giving you more work to do, here are the files I edited from the latest bee (2022-07-17).

Regards,
Mike

18f26q43.jal
18f27q43.jal
18f25q43.jal
18f25q43.jal
18f27q43.jal
18f26q43.jal

Mike

unread,
Jul 19, 2022, 9:49:41 PM7/19/22
to jallib
I attached the files twice because I couldn't tell if they were attached the first time.  (Old PC, I guess).

Rob CJ

unread,
Jul 20, 2022, 1:36:13 PM7/20/22
to jal...@googlegroups.com
Hi Mike,

This time you seemt to be right 🙂.

There is an error in the XML file from Microchip where it mentions portd as being present on a 18f26q43 while it does not have it.

It might be - however - that the port is present but since there are no pins left they are just not connected. It is sometimes cheaper to do that instead of making a separate silicon for that.  I could test if I could read and write to that port to verify this.

Or it is an error that Microchip will fix later, not sure.

I could change the Python script to leave that out but then I need to remember to remove it when it is fixed (if they do that) and the script becomes harder to maintain. Since the pinning in the datasheet does not mention the presence of a portd I do not think it will result in any problems since you cannot control anything with it.

Do you agree or did is miss something that still it may result in a problem?

Kind regards,

Rob


Van: 'Mike' via jallib <jal...@googlegroups.com>
Verzonden: woensdag 20 juli 2022 03:49
Aan: jallib <jal...@googlegroups.com>
Onderwerp: [jallib] Re: Possible chip definition file error
 
--
You received this message because you are subscribed to the Google Groups "jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jallib+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jallib/cf0f8422-848f-4ea0-b952-cc2329842169n%40googlegroups.com.

Mike

unread,
Jul 20, 2022, 2:13:04 PM7/20/22
to jallib
Hi Rob,

I will tell you that if the PORTD, TRISD, and LATD definitions didn't exist in the chip def file I would have much spent less time pulling my hair out yesterday.  :)  I couldn't figure out why my OLED wasn't working.  I was referencing PORTD pins for software I2C and the compiler didn't throw an error.

By the way, many PIC chips share the same silicon die...It's just easier/cheaper to manufacture that way.  They're turned into specific PIC chips in the final phase when they do testing and such.  The 18F25/6/7Q43 won't ever have a PORTD; it's been set in stone now.   And they share the same topology as other 28-pin PIC's such as the old 16F886...no PORTD.  Microchip isn't going to mess with that.  Any error in a library or datasheet is almost certainly an intern's mistake.

It's your call, though.  I've removed those definitions from the chip def files on my own computer.  

Regards,
Mike

Rob CJ

unread,
Jul 21, 2022, 1:23:26 PM7/21/22
to jal...@googlegroups.com
Hi Mike,

I am still a bit confused about how you work. If you define I2c pins on portd for the 18f26q43 which does not have the portd pins available how to you know that it does not work? 

For testing I assume that you first connect the OLED and then find out that there are no portd pins on the PIC to connect it to and I would then assume (any assumption is dangerous, I know) that you would select other pins for that.

But maybe I totally miss what you are doing.

Kind regards,

Rob



Van: 'Mike' via jallib <jal...@googlegroups.com>
Verzonden: woensdag 20 juli 2022 20:13
Aan: jallib <jal...@googlegroups.com>
Onderwerp: Re: [jallib] Re: Possible chip definition file error
 
Reply all
Reply to author
Forward
0 new messages