usbpicprog-technical

[Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Jasper W. <ja...@po...>]

Hi,

I've had some more time to look into this and i've got the Microchip C18 
compiler installed now.

Firstly it looks as if the pic16F88 data programming is missing the 
end programming command, see attached patch.

Unfortunatly i cant test it beacause the C18 compiler failes to link 
usbdsc.o with this error:

Error - section '.romdata_usbdsc.o' can not fit the section. Section 
'.romdata_usbdsc.o' length=0x000000ae 
Errors : 1 

I get these warnings when i compile it:

usbdsc.c:269: warning: [2054] suspicious pointer conversion 
usbdsc.c:269: warning: [2054] suspicious pointer conversion 
usbdsc.c:273: warning: [2054] suspicious pointer conversion 
usbdsc.c:273: warning: [2054] suspicious pointer conversion 
usbdsc.c:273: warning: [2054] suspicious pointer conversion 

but i don't know if thats significant.

Also - there are lots of references to sdcc in the firmware source, does 
compiling with sdcc work now? I tried breifly but got loads of errors...

-- 
[http://pointless.net/]                                   [0x2ECA0975]

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Frans S. <fra...@gm...>]

Dear Jasper Wallace,

You are right, thank you for the patch! I have applied the changes in
the subversion repos and in the 0.4.0 firmware release.

About the warnings, this is part of the Microchip usb stack, and I have
seen them but let's just ignore those warnings.

Frans

On 05/25/2010 10:30 AM, Jasper Wallace wrote:
> Hi,
>
> I've had some more time to look into this and i've got the Microchip C18 
> compiler installed now.
>
> Firstly it looks as if the pic16F88 data programming is missing the 
> end programming command, see attached patch.
>
> Unfortunatly i cant test it beacause the C18 compiler failes to link 
> usbdsc.o with this error:
>
> Error - section '.romdata_usbdsc.o' can not fit the section. Section 
> '.romdata_usbdsc.o' length=0x000000ae 
> Errors : 1 
>
> I get these warnings when i compile it:
>
> usbdsc.c:269: warning: [2054] suspicious pointer conversion 
> usbdsc.c:269: warning: [2054] suspicious pointer conversion 
> usbdsc.c:273: warning: [2054] suspicious pointer conversion 
> usbdsc.c:273: warning: [2054] suspicious pointer conversion 
> usbdsc.c:273: warning: [2054] suspicious pointer conversion 
>
> but i don't know if thats significant.
>
> Also - there are lots of references to sdcc in the firmware source, does 
> compiling with sdcc work now? I tried breifly but got loads of errors...
>
>   
>
>
> ------------------------------------------------------------------------------
>
>   
>
>
> _______________________________________________
> Usbpicprog-technical mailing list
> Usb...@li...
> https://lists.sourceforge.net/lists/listinfo/usbpicprog-technical
>   

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Jasper W. <ja...@po...>]

On Tue, 25 May 2010, Frans Schreuder wrote:

> Dear Jasper Wallace,
> 
> You are right, thank you for the patch! I have applied the changes in
> the subversion repos and in the 0.4.0 firmware release.

Thanks, if i use uc_code.hex from svn I can write the data ok!

However now the 2nd config word has stoped working!

I think the attached patch will work (I'm looking at the 39607c 
datasheet). I wonder how it ever worked in the 1st place!

> About the warnings, this is part of the Microchip usb stack, and I have
> seen them but let's just ignore those warnings.

yup, but the linker error means i can't generate firmware to test with :(

I've using:

mcc18:

MPLAB C18 v3.35 academic Copyright 2000-2010 Microchip Technology Inc.

mplink:

MPLINK 4.35, Linker

the linker command line is:

wine mplink /k/ rm18f4550.lkr /l/home/jasper/.wine/drive_c/MCC18/lib/ 
/ouc_code.cof /muc_code.map main.o usbdsc.o interrupt.o usbmmap.o usbgen.o 
usb9.o usbctrltrf.o usbdrv.o upp.o prog.o prog_lolvl.o 

> > Unfortunatly i cant test it beacause the C18 compiler failes to link 
> > usbdsc.o with this error:
> >
> > Error - section '.romdata_usbdsc.o' can not fit the section. Section 
> > '.romdata_usbdsc.o' length=0x000000ae 
> > Errors : 1 

-- 
[http://pointless.net/]                                   [0x2ECA0975]

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Frans S. <fra...@gm...>]

On 05/26/2010 05:35 AM, Jasper Wallace wrote:
> On Tue, 25 May 2010, Frans Schreuder wrote:
>
>   
>> Dear Jasper Wallace,
>>
>> You are right, thank you for the patch! I have applied the changes in
>> the subversion repos and in the 0.4.0 firmware release.
>>     
> Thanks, if i use uc_code.hex from svn I can write the data ok!
>
> However now the 2nd config word has stoped working!
>
> I think the attached patch will work (I'm looking at the 39607c 
> datasheet). I wonder how it ever worked in the 1st place!
>   
Hmm, this is weird... but your patch seems to be right. I have built it
and put in the svn. Could you try again?
>   
>> About the warnings, this is part of the Microchip usb stack, and I have
>> seen them but let's just ignore those warnings.
>>     
> yup, but the linker error means i can't generate firmware to test with :(
>   
Also this is weird. Maybe the difference is in the student version. I
did get a full version of the c18 compiler from a formal job (they
didn't use it anymore). And now I am still able to upgrade with the
latest upgrades.
> I've using:
>
> mcc18:
>
> MPLAB C18 v3.35 academic Copyright 2000-2010 Microchip Technology Inc.
>
> mplink:
>
> MPLINK 4.35, Linker
>
> the linker command line is:
>
> wine mplink /k/ rm18f4550.lkr /l/home/jasper/.wine/drive_c/MCC18/lib/ 
> /ouc_code.cof /muc_code.map main.o usbdsc.o interrupt.o usbmmap.o usbgen.o 
> usb9.o usbctrltrf.o usbdrv.o upp.o prog.o prog_lolvl.o 
>
>   
>>> Unfortunatly i cant test it beacause the C18 compiler failes to link 
>>> usbdsc.o with this error:
>>>
>>> Error - section '.romdata_usbdsc.o' can not fit the section. Section 
>>> '.romdata_usbdsc.o' length=0x000000ae 
>>> Errors : 1 
>>>       
>   
>
>
> ------------------------------------------------------------------------------
>
>   
>
>
> _______________________________________________
> Usbpicprog-technical mailing list
> Usb...@li...
> https://lists.sourceforge.net/lists/listinfo/usbpicprog-technical
>   

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Frans S. <fra...@gm...>]

Dear Jasper Wallace,

Have you tested the latest build?

Frans

On 05/25/2010 10:30 AM, Jasper Wallace wrote:
> more time to look into this and i've got the Microchip C18 
> compiler installed now.
>
> Firstly it looks as if the pic16F88 data programming is missing the 
> end programming command, see attached patch.
>
> Unfortunatly i cant test it beacause the C18 compiler failes to link 
> usbdsc.o with this error:
>
> Error - section '.romdata_usbdsc.o' can not fit the section. Section 
> '.romdata_usbdsc.o' length=0x000000ae 
> Errors : 1 
>
>   

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Jasper W. <ja...@po...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 28 May 2010, Frans Schreuder wrote:

> Dear Jasper Wallace,
> 
> Have you tested the latest build?

Hi,

Yes, and both config words don't work now!

Well they do sometimes but not always, I'm not sure whats going on.

Luckily this ties into another project of mine which is interfacing with a 
HP1650A logic analyser and converting the data into a format the open 
source sump logic analyser client can read.

Unfortunatly thats still a work in progress, i can't get the analyser to 
record exactly what i want (i need to read the manual more).

the client is here:

http://www.sump.org/projects/analyzer/client/

attached is a .slp file that labels the traces, and some .sla files which 
are the actual traces.

btw - one of the problems with my converter is that i think i've got the 
time units wrong.

One thing that might be a problem is that the 1st and last clock pulse for 
the 14/16 data bits are shorter than the others, it dosn't look like the 
diagrams in the datasheet.


- -- 
[http://pointless.net/]                                   [0x2ECA0975]

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iQEVAwUBTABqywCB+Qwuygl1AQJWYwf+P91d7z15XEvV9vtoUvmNOu7tN0cI1WIA
+52FRJyghHUfEB6BgENxo3Tj4WaWWTmMNE/95j67/Kd3hNFUV8bhsZvcWnJANYoL
uczXsAukTiQjhObOXixevgGJCMcSdAhVMahiWKvHncXdcXzVTZuHUV15Fx3rDN9u
bkVAPNuK1NUz7Z5DUf0Su11Muv7igX2o7iFSMuGBvGWtXCCFSlG3+iiuulSM9eUo
nDIqs31N1bMLD8s9h+A8ZE/CcaHSwcwl/cvTL6PqYfk7cXuJZ6z2Wgr5fx0PaG+m
k6o1gEoDx0Mzyb+e/RjFdp8WMqQ69L/8VG4y5MatQFXVUXmzxLy0LQ==
=SXrC
-----END PGP SIGNATURE-----

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Jasper W. <ja...@po...>]

On Fri, 28 May 2010, Frans Schreuder wrote:

> Dear Jasper Wallace,
> 
> Have you tested the latest build?

Still not sure whats going on, but:

If i have all the boxes under Preferences ticked, writing and verifying 
the config registers fails (writing the data and code works, and can be 
verified if 'Verify config memory' is unticked).

If i just have 'Write config memory' and 'Verify after programming' ticked 
it writes the config words ok.

So i guess there is something subtle wrong with the way changing from 
writing code to writing config is handled or something.

-- 
[http://pointless.net/]                                   [0x2ECA0975]

Re: [Usbpicprog-technical] Pic16F88 and eeprom/data programming

[Frans S. <fra...@gm...>]

Dear Jasper Wallace,

After writing code or data, the target processor should be reset
completely (VPP goes low)
Unfortunately I don't have a 16F87/88 PIC right now. I will order one
when I have to order other components in order to test it myself.

Frans

On 29-5-2010 18:46, Jasper Wallace wrote:
> On Fri, 28 May 2010, Frans Schreuder wrote:
>
>   
>> > Dear Jasper Wallace,
>> > 
>> > Have you tested the latest build?
>>     
> Still not sure whats going on, but:
>
> If i have all the boxes under Preferences ticked, writing and verifying 
> the config registers fails (writing the data and code works, and can be 
> verified if 'Verify config memory' is unticked).
>
> If i just have 'Write config memory' and 'Verify after programming' ticked 
> it writes the config words ok.
>
> So i guess there is something subtle wrong with the way changing from 
> writing code to writing config is handled or something.
>