amforth-devel — Development discussion and user support

Re: [Amforth] Assembler For Building Amforth

[Leon M. <leo...@gm...>]

That certainly sounds like an interesting idea. At present, my Python
skills are better than my Forth skills, so -- if you do go ahead -- let me
know if I can help. (I've been meaning to learn about AVR assembly for some
time.)

-Leon

P.S. FWIW, I contacted Atmel customer support, and -- after consulting
their legal team -- they said it was fine to redistribute avrasm2.exe.


On Thu, Oct 17, 2013 at 10:47 PM, Tom Harris <cel...@gm...> wrote:

> It seems to be a common perceived problem that building Amforth requires
> the Atmel assembler that is proprietary, only runs on Windows, is not open
> source, may contain traces of nuts, etc. Would there be any value is
> writing a simple portable assembler in (e.g.) Python that could build
> Amforth from sources? The assembler syntax is fairly simple, and it looks
> like a simple two pass assembler could do the job.
>
> Tom Harris <cel...@gm...>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60135031&iu=/4140/ostg.clktrk
> _______________________________________________
> Amforth-devel mailing list for http://amforth.sf.net/
> Amf...@li...
> https://lists.sourceforge.net/lists/listinfo/amforth-devel
>

[Amforth] Assembler For Building Amforth

[Tom H. <cel...@gm...>]

It seems to be a common perceived problem that building Amforth requires
the Atmel assembler that is proprietary, only runs on Windows, is not open
source, may contain traces of nuts, etc. Would there be any value is
writing a simple portable assembler in (e.g.) Python that could build
Amforth from sources? The assembler syntax is fairly simple, and it looks
like a simple two pass assembler could do the job.

Tom Harris <cel...@gm...>

Re: [Amforth] Build Issues

[Michael P. <mp...@rc...>]

Hi Craig,

Just catching up after a few days away from the computer ... I trust
you've been able to get 5.1 working on your Uno?

The link you posted (by Neal Crook), is well-done ... it got me started
and finally working, but it's out of date and some of the
items have changed.  At home, I have made notes on a printed copy of
Neal's piece, which point out the changes required to get
that process working with the current versions of Atmel Studio and amForth.

Once you are able to work through the process of flashing a chip that
will run, the next is to go through the multitude of files
that will give your Uno further functionality and learning exactly where
to stick those source file declarations so everything works!

Good luck!  The list has been very helpful to me.

Regards,

Michael
K6MLE



On 10/15/2013 12:32 PM, craig and heather wrote:
> Never mind, I found the problem. The generate eep file checkbox was
> unchecked. Sorry for the noise.
>
> Craig Lindley
>
>
> On Tue, Oct 15, 2013 at 12:58 PM, craig and heather <ca...@gm...>wrote:
>
>> I searched the entire C drive and could not find an eep file. Attached is
>> a screen shot of the output directory.
>>
>> To answer your question Matthias, if I'm going to use some code I really
>> like to know how it works and how to build it. In Amforth's case, I might
>> want to try and change some of the included words so I have to be able to
>> rebuild from source.
>>
>> Thanks for you help with this and I will first try the pre-canned Arduino
>> Uno image(s).
>>
>> Craig Lindley
>>
>>
>>
>>
>> On Tue, Oct 15, 2013 at 11:33 AM, craig and heather <ca...@gm...>wrote:
>>
>>> Looking closer at the build results I assume the eseg would be destined
>>> for the eeprom. Could it be that the hex file has both parts in it? And if
>>> so, what are the chances of success by just programming the single file
>>> onto the Arduino? I would try this but I'm fresh out of Arduino Uno's just
>>> now.
>>>
>>> Segment   Begin    End      Code   Data   Used    Size   Use%
>>>         ---------------------------------------------------------------
>>>         [.cseg] 0x000000 0x007edc   1822  10500  12322   32768  37.6%
>>>         [.dseg] 0x000100 0x0001e1      0    225    225    2048  11.0%
>>>         [.eseg] 0x000000 0x000054      0     84     84    1024   8.2%
>>>         Assembly complete, 0 errors. 1 warnings
>>>
>>> --
>>> Craig Lindley
>>>
>>>
>>
>> --
>> Craig Lindley / Heather Hubbard
>>
>>
>> New Recordings: craigandheather.net/cnmpage.html
>> Latest CD: craigandheather.net/songsilike2013cd.html
>>
>> Personal Website: craigandheather.net
>> Business Website: clockwork.craigandheather.net
>> Phone: (719) 495-1873
>> Cell: (719) 502-7925
>>
>> If you’re one in a million, there are now seven thousand people exactly
>> like you.
>>
>
>

Re: [Amforth] Build Issues

[craig a. h. <ca...@gm...>]

Never mind, I found the problem. The generate eep file checkbox was
unchecked. Sorry for the noise.

Craig Lindley


On Tue, Oct 15, 2013 at 12:58 PM, craig and heather <ca...@gm...>wrote:

> I searched the entire C drive and could not find an eep file. Attached is
> a screen shot of the output directory.
>
> To answer your question Matthias, if I'm going to use some code I really
> like to know how it works and how to build it. In Amforth's case, I might
> want to try and change some of the included words so I have to be able to
> rebuild from source.
>
> Thanks for you help with this and I will first try the pre-canned Arduino
> Uno image(s).
>
> Craig Lindley
>
>
>
>
> On Tue, Oct 15, 2013 at 11:33 AM, craig and heather <ca...@gm...>wrote:
>
>> Looking closer at the build results I assume the eseg would be destined
>> for the eeprom. Could it be that the hex file has both parts in it? And if
>> so, what are the chances of success by just programming the single file
>> onto the Arduino? I would try this but I'm fresh out of Arduino Uno's just
>> now.
>>
>> Segment   Begin    End      Code   Data   Used    Size   Use%
>>         ---------------------------------------------------------------
>>         [.cseg] 0x000000 0x007edc   1822  10500  12322   32768  37.6%
>>         [.dseg] 0x000100 0x0001e1      0    225    225    2048  11.0%
>>         [.eseg] 0x000000 0x000054      0     84     84    1024   8.2%
>>         Assembly complete, 0 errors. 1 warnings
>>
>> --
>> Craig Lindley
>>
>>
>
>
> --
> Craig Lindley / Heather Hubbard
>
>
> New Recordings: craigandheather.net/cnmpage.html
> Latest CD: craigandheather.net/songsilike2013cd.html
>
> Personal Website: craigandheather.net
> Business Website: clockwork.craigandheather.net
> Phone: (719) 495-1873
> Cell: (719) 502-7925
>
> If you’re one in a million, there are now seven thousand people exactly
> like you.
>



-- 
Craig Lindley / Heather Hubbard

New Recordings: craigandheather.net/cnmpage.html
Latest CD: craigandheather.net/songsilike2013cd.html

Personal Website: craigandheather.net
Business Website: clockwork.craigandheather.net
Phone: (719) 495-1873
Cell: (719) 502-7925

If you’re one in a million, there are now seven thousand people exactly
like you.

Re: [Amforth] Build Issues

[craig a. h. <ca...@gm...>]

I searched the entire C drive and could not find an eep file. Attached is a
screen shot of the output directory.

To answer your question Matthias, if I'm going to use some code I really
like to know how it works and how to build it. In Amforth's case, I might
want to try and change some of the included words so I have to be able to
rebuild from source.

Thanks for you help with this and I will first try the pre-canned Arduino
Uno image(s).

Craig Lindley




On Tue, Oct 15, 2013 at 11:33 AM, craig and heather <ca...@gm...>wrote:

> Looking closer at the build results I assume the eseg would be destined
> for the eeprom. Could it be that the hex file has both parts in it? And if
> so, what are the chances of success by just programming the single file
> onto the Arduino? I would try this but I'm fresh out of Arduino Uno's just
> now.
>
> Segment   Begin    End      Code   Data   Used    Size   Use%
>         ---------------------------------------------------------------
>         [.cseg] 0x000000 0x007edc   1822  10500  12322   32768  37.6%
>         [.dseg] 0x000100 0x0001e1      0    225    225    2048  11.0%
>         [.eseg] 0x000000 0x000054      0     84     84    1024   8.2%
>         Assembly complete, 0 errors. 1 warnings
>
> --
> Craig Lindley
>
>


-- 
Craig Lindley / Heather Hubbard

New Recordings: craigandheather.net/cnmpage.html
Latest CD: craigandheather.net/songsilike2013cd.html

Personal Website: craigandheather.net
Business Website: clockwork.craigandheather.net
Phone: (719) 495-1873
Cell: (719) 502-7925

If you’re one in a million, there are now seven thousand people exactly
like you.

Re: [Amforth] Build Issues

[Matthias T. <mt...@we...>]

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

Hi Craig,

> Looking closer at the build results I assume the eseg would be
> destined for the eeprom. Could it be that the hex file has both
> parts in it?

Unfortunately not. hex files are too stupid to contain such information

> And if so, what are the chances of success by just programming the
> single file onto the Arduino?

Wont work.

> I would try this but I'm fresh out of Arduino Uno's just now.

Why do you want to compile your own files in this case? There are both
hex files in the distribtion appl/arduino/ directory! I'd start with
them...

Matthias

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlJdgAsACgkQ9bEHdGEMFjM5zgCgnI3Q9/JzOabgBO/id+ZTpSzM
LxEAnjR3+4ZBkugfBPHysrzhVdptp/vn
=2z04
-----END PGP SIGNATURE-----

Re: [Amforth] Build Issues

[Matthias T. <mt...@we...>]

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

Hi Craig,

> 
> And the build when fairly well. I only had one warning:
> 
> Warning    1    Use of undefined or forward referenced symbol
> 'XT_NOOP' in .equ/.set
> 
> I don't know whether or not this warning is a problem

No, it is not. We discussed this only recently.

> but my biggest problem is that the assembler output a 24K hex file
> but no eep file.

That's strange indeed. Can you double check the directory? There should
be 2 (two) hex files, one of them with a suffix eep: eep.hex.

Matthias

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlJdfnQACgkQ9bEHdGEMFjPkPwCeMnFRlnuWhrcUPBMoKcFp62Dp
X/8AnRfKcv+QV1Cal41XZUG7thBFXslC
=S0lq
-----END PGP SIGNATURE-----

Re: [Amforth] Build Issues

[craig a. h. <ca...@gm...>]

Looking closer at the build results I assume the eseg would be destined for
the eeprom. Could it be that the hex file has both parts in it? And if so,
what are the chances of success by just programming the single file onto
the Arduino? I would try this but I'm fresh out of Arduino Uno's just now.

Segment   Begin    End      Code   Data   Used    Size   Use%
        ---------------------------------------------------------------
        [.cseg] 0x000000 0x007edc   1822  10500  12322   32768  37.6%
        [.dseg] 0x000100 0x0001e1      0    225    225    2048  11.0%
        [.eseg] 0x000000 0x000054      0     84     84    1024   8.2%
        Assembly complete, 0 errors. 1 warnings

-- 
Craig Lindley

[Amforth] Build Issues

[craig a. h. <ca...@gm...>]

Hello Amforth Developers

I just discovered Amforth from a Hack A Day post and I want to try it out
on the Arduino Uno. I downloaded the 5.1 version of Amforth and downloaded
the Atmel Studio 6.1.

I followed the directions given in this post:

http://www.mail-archive.com/amf...@li.../msg00556.html

And the build when fairly well. I only had one warning:

Warning    1    Use of undefined or forward referenced symbol 'XT_NOOP' in
.equ/.set

I don't know whether or not this warning is a problem but my biggest
problem is that the assembler output a 24K hex file but no eep file.

Anyone had this issue and know how to fix it?

Thanks,

Craig Lindley

New Recordings: craigandheather.net/cnmpage.html
Latest CD: craigandheather.net/songsilike2013cd.html

Personal Website: craigandheather.net
Business Website: clockwork.craigandheather.net
Phone: (719) 495-1873
Cell: (719) 502-7925

If you’re one in a million, there are now seven thousand people exactly
like you.

Re: [Amforth] AmForth Mentioned in Hackaday

[Leon M. <leo...@gm...>]

As with any time Forth comes up, that post generated some heat in the
comments, but it seems reasonably civilized.

FWIW, the managing editor of Hack a Day lives in the same city as me, and
he asked me to write that article.

Anyhow, I've been lurking here for a while without using AmForth, but I'm
probably going to be getting back into it -- actually using it in projects
as opposed to really playing with floating point math. (Although I want to
clean up that library some.)

On Sun, Oct 13, 2013 at 5:17 PM, pito <pi...@vo...> wrote:

> BTW, I created full float library in asm for amforth, but not toyed with
> amforth for years, so no idea if it would work with the newest releases..
> My 5primitives library will definitely help you, as replacing the 5 forth
> primitives with their asm friends speeds up the float stuff significantly,
> so it becomes even useful.. Of course, the puristic amforth community would
> not appreciate such asm approach, though..
>

I'd be interested to test them. I think the only changes I had to make for
5.1 compatibility were in things like the recognizer.


> There are smaller (and cheaper probably) FTDI chips available, ie. FT230X
> etc.
>

Are they hand solderable?


> Pito (not Czech,  but my email is..)
> ;)


Ok, I've changed it from Czech to "International Man of Mystery" ;)
-Leon

Re: [Amforth] AmForth Mentioned in Hackaday

[pito <pi...@vo...>]

Hi Leon, nice to read you are 4th alive :)..

BTW, I created full float library in asm for amforth, but not toyed with amforth for years, so no idea if it would work with the newest releases.. My 5primitives library will definitely help you, as replacing the 5 forth primitives with their asm friends speeds up the float stuff significantly, so it becomes even useful.. Of course, the puristic amforth community would not appreciate such asm approach, though..

There are smaller (and cheaper probably) FTDI chips available, ie. FT230X etc.

Pito (not Czech,  but my email is..)
;)



----- PŮVODNÍ ZPRÁVA -----
Od: "Leon Maurer" <leo...@gm...>
Komu: "Amforth Dev List" <amf...@li...>
Předmět: Re: [Amforth] AmForth Mentioned in Hackaday
Datum: 13.10.2013 - 22:23:18

> Glad you liked it. I thought they weren't going to publish that until
> tomorrow. I just got out of the woods...
> -Leon
> On Oct 13, 2013 3:15 AM, "Matthias Trute" <mt...@we...> wrote:
> 
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > Hi Paulo,
> >
> > > Here is the link:
> > > http://hackaday.com/2013/10/12/a-simple-forth-development-board
> >
> > Cool.
> >
> > Many thanks to Leon. I took the freedom to place a link to his video
> > onto amforth.sf.net (will no doubt offence all youtube-haters)
> >
> >
> > Matthias
> >
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.4.11 (GNU/Linux)
> > Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
> >
> > iEYEARECAAYFAlJaVncACgkQ9bEHdGEMFjMFUwCdEGr3viPbk/J+NYSYHQqOE1nd
> > 8skAn1lfAtoZ/jbKyyAbvK8igiX5FbP4
> > =jV5W
> > -----END PGP SIGNATURE-----
> >
> >
> > ------------------------------------------------------------------------------
> > October Webinars: Code for Performance
> > Free Intel webinars can help you accelerate application performance.
> > Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> > from
> > the latest Intel processors and coprocessors. See abstracts and register >
> > http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> > _______________________________________________
> > Amforth-devel mailing list for http://amforth.sf.net/
> > Amf...@li...
> > https://lists.sourceforge.net/lists/listinfo/amforth-devel
> >
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from 
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> Amforth-devel mailing list for http://amforth.sf.net/
> Amf...@li...
> https://lists.sourceforge.net/lists/listinfo/amforth-devel

Re: [Amforth] AmForth Mentioned in Hackaday

[Leon M. <leo...@gm...>]

Glad you liked it. I thought they weren't going to publish that until
tomorrow. I just got out of the woods...
-Leon
On Oct 13, 2013 3:15 AM, "Matthias Trute" <mt...@we...> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi Paulo,
>
> > Here is the link:
> > http://hackaday.com/2013/10/12/a-simple-forth-development-board
>
> Cool.
>
> Many thanks to Leon. I took the freedom to place a link to his video
> onto amforth.sf.net (will no doubt offence all youtube-haters)
>
>
> Matthias
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.11 (GNU/Linux)
> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>
> iEYEARECAAYFAlJaVncACgkQ9bEHdGEMFjMFUwCdEGr3viPbk/J+NYSYHQqOE1nd
> 8skAn1lfAtoZ/jbKyyAbvK8igiX5FbP4
> =jV5W
> -----END PGP SIGNATURE-----
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> Amforth-devel mailing list for http://amforth.sf.net/
> Amf...@li...
> https://lists.sourceforge.net/lists/listinfo/amforth-devel
>

Re: [Amforth] AmForth Mentioned in Hackaday

[Matthias T. <mt...@we...>]

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

Hi Paulo,

> Here is the link: 
> http://hackaday.com/2013/10/12/a-simple-forth-development-board

Cool.

Many thanks to Leon. I took the freedom to place a link to his video
onto amforth.sf.net (will no doubt offence all youtube-haters)


Matthias

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlJaVncACgkQ9bEHdGEMFjMFUwCdEGr3viPbk/J+NYSYHQqOE1nd
8skAn1lfAtoZ/jbKyyAbvK8igiX5FbP4
=jV5W
-----END PGP SIGNATURE-----

[Amforth] AmForth Mentioned in Hackaday

[Paulo F. <pa...@ke...>]

Here is the link:  
http://hackaday.com/2013/10/12/a-simple-forth-development-board

Best regards

Paulo Ferreira 

Re: [Amforth] Watchdog prescalers

[Matthias T. <mt...@we...>]

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

Hi David,


> If I get watchdog sleep running OK I'll send some text and the code
> to add to your documentation if that's useful.

Everything that a bloody newby may understand immediatly is fine ;)
(I doubt that the watchdog is something in that category, but I may be
wrong here ;) )

I just uploaded a very early version:
http://amforth.sourceforge.net/TG/recipes/Assembler.html, I'll improve
it ASAP (and fix the typos as
well, sorry); I hope it is not completely useless in its current
state...

Matthias

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlJZnVMACgkQ9bEHdGEMFjNJbQCgzMhRLYTH+NU1FzHO2WywC6XF
mtAAn19rnOIHkLmyIEcpQvhGGKmgKGJS
=j4pd
-----END PGP SIGNATURE-----

Re: [Amforth] Watchdog prescalers

[<vu...@em...>]

  It works!

> TimeOuts @ .
0  ok
> decimal
  ok
> 1s wdt=prescaler!
  ok
> ?wdt.
wdt control&status = 6
  ok
> &45 TimeOuts ! TimeOuts @ .
45  ok
> wdt!interrupt
  ok
> TimeOuts @ .
29  ok
> TimeOuts @ . ?wdt.
15 wdt control&status = 46
  ok
> TimeOuts @ .
0  ok
> ?wdt.
wdt control&status = 46
  ok
> ?wdt. wdt!stop ?wdt.
wdt control&status = 46
wdt control&status = 6
  ok
>


Vladimir.

Re: [Amforth] Watchdog prescalers

[<vu...@em...>]

> ...
> A quick start: It's not rocket science. You need the dict_wl.inc
> included, load vocabulary.frt and assembler.frt from the lib directory.
> ...
> Matthias

  I have loaded:

  marker.frt
  postpone.frt
  vocabulary.frt
  assembler.frt
  multitask.frt
  bitnames.frt
  atmega328p.frt
  misc.frt
  ...
  and more...

  moving forth,
  Vladimir.

Re: [Amforth] Watchdog prescalers

[David W. <ins...@gm...>]

Thanks to both Vladimir and Matthius for your input. I'll try and get it
all up and running on Monday.

If I get watchdog sleep running OK I'll send some text and the code to add
to your documentation if that's useful.

Cheers,

David

Re: [Amforth] Watchdog prescalers

[<vu...@em...>]

> ...can you give me a pointer on what to do with the
> assembler code that you posted?...
> ...
> David

  You need to use "...\amforth-5.1\lib\assembler.frt"
  Good luck!
  moving forth,
  Vladimir.

Re: [Amforth] Watchdog prescalers

[Matthias T. <mt...@we...>]

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

Hi David,

> Sorry for being dumb, but can you give me a pointer on what to do 
> with the assembler code that you posted? I've never had to use 
> assembler with forth before and can't find a page that describes 
> what to do on the site.

You're right, the assembler is very poorly documented. I never used it
myself, so I gladly accepted the work Lubos has been made developing it.
I wrote an article about it for the German forth society a few years
ago, but never spend the time to translate it into English. I just
started to write a recipe, but it may take some time to finish.

A quick start: It's not rocket science. You need the dict_wl.inc
included, load vocabulary.frt and assembler.frt from the lib directory.
With that, Vladimir's examples should work. More examples are available
behind the link he sent yesterday. I hope your government does not do
any harm on you for reading a Russian website ;)

Matthias

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlJZjmMACgkQ9bEHdGEMFjNllwCgkt1mC1G+SiSyJaYeEOWsuQdj
wQYAoOwoyHekjvVTFAH+av23tS8YF+55
=Kcl8
-----END PGP SIGNATURE-----

Re: [Amforth] Watchdog prescalers

[David W. <ins...@gm...>]

Sorry for being dumb, but can you give me a pointer on what to do with the
assembler code that you posted? I've never had to use assembler with forth
before and can't find a page that describes what to do on the site.

Many thanks for your help.

David

Re: [Amforth] Watchdog prescalers

[David W. <ins...@gm...>]

Fantastic. Looks as if this is just what I need. I'll give it a try.

Thanks,

David

Re: [Amforth] Watchdog prescalers

[<vu...@em...>]

> Message: 1
> Date: Sat, 12 Oct 2013 12:05:53 +0100
>From: David Wallis <ins...@gm...>
> Subject: Re: [Amforth] Watchdog prescalers
> To: amf...@li...
> 
> ...
> ... I was keen to use the 
>watchdog timer because it
> works with the lowest power mode (power down) and 
>doesn't need external
> components.
> ...
> 
> David
> 


Hello, David.
I have Freaduino Pro Mini ( atmega328p @16MHz)
http://www.hwkitchen.com/products/freaduino-pro-328-5v-16mhz/
I used WatchDog Timer as 1s Interval timer with this code

\
\ WDTON Fuse  unprogrammed

   marker =wdt-timeouts=

\ Watchdog Timer Control Register bits:
\ WDIF WDIE WDP3 WDCE WDE WDP2 WDP1 WDP0

\ am4th51 constants:
\ WDTCSR_WDE WDTCSR_WDCE WDTCSR_WDP WDTCSR_WDIE 
WDTCSR_WDIF WDTCSR

   WDTCSR WDTCSR_WDIF  bitmask: wdt.if
   WDTCSR WDTCSR_WDIE  bitmask: wdt.ie
   WDTCSR WDTCSR_WDCE  bitmask: wdt.ce
   WDTCSR WDTCSR_WDE   bitmask: wdt.e
   WDTCSR WDTCSR_WDP   bitmask: wdt.psc

\
\
     also assembler

   code wdtcsr! ( # --- )

     R16 SREG     lds, cli,
     R25 WDTCSR        lds,
     R25 WDTCSR_WDCE
         WDTCSR_WDE or ori,
     WDTCSR     R25    sts,
     WDTCSR     R24    sts,

        R24     Y+      ld,   \ poptos
        R25     Y+      ld,
     SREG       R16    sts,
   end-code

     previous

\
\ Watchdog Timer Prescale Select
\
\ WDTCSR_WDP bits mask

   $0 constant 1/64s
   $1 constant 1/32s
   $2 constant 1/16s
   $3 constant 1/8s
   $4 constant 1/4s
   $5 constant 1/2s
   $6 constant 1s
   $7 constant 2s
   $8 constant 4s
   $9 constant 8s

   : wdt=prescaler!  ( n -- )

      8s over < if  .x abort" illegal"  then
      dup 2 lshift $20 ( wdp3 ) and or
      wdt.psc drop and
      wdtcsr!
      ;
\
\  mode              action
\  wdt-stop          none
\  wdt-interrupt     interrupt
\  wdt-reset         system reset

   : wdt!interrupt wdt.ie high ;
   : wdt!reset     wdt.e c@ or wdtcsr! ;
   : wdt!stop
        wdt.e wdt.ie
        drop rot or invert swap c@ and wdtcsr! ;

\
\

   variable TimeOuts
   variable StopWatch

   : 0TimeOuts! 0 TimeOuts ! ;

   : 0StopWatch! 0 StopWatch ! ;

\
\ waiting for downcounter
\
   : time?out   begin pause TimeOuts @ 0= until ;

\
\ variable decrement
\
   : ?--!      ( a --- )
        dup @
       ?dup if 1- swap !
          else drop
          then
        ;

\
\   wdt interrupt service routine
\
   : tick-tack
        wdt.if   high   \ очень жаль тактов...
        TimeOuts ?--!
        ;
\
\
   : interval  ( n --- )
       TimeOuts !
       wdt!interrupt
       time?out
       wdt!stop
      ;

\
\ -----------------------------------------------
\ wdt info

   : ?wdt. WDTCSR c@ ." wdt control&status = " .x cr ;

\

Re: [Amforth] Watchdog prescalers

[Erich W. <ew....@na...>]

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

Hi David,

On 10/12/2013 01:05 PM, David Wallis wrote:
>> Are you sure, you want to do this with the watchdog timer? Isn't the
>> watchdog timer only able to generate a reset?
> 
> 
> Thanks for your code Erich. I was keen to use the watchdog timer because it
> works with the lowest power mode (power down) and doesn't need external
> components.
> 
> It seems to be a fairly common method. The watchdog can either generate a
> reset or trigger an interrupt, or both. This interrupt can also be used as
> a wake up from power down. (At least on the 324P that I'm using).

"or an interrupt". Aha. Looking into the datasheets again:
atmega32 (the on I use most often): WDT generates a reset only.
atmega644p (of which I have some):  WDT generates an interrupt or a reset.

So I learned something. Thanks for enlightening me. :-)

> The other way I was considering was with timer 2 in synchronous mode (so I
> don't need external components) and use power save mode. Did you ever look
> at synchronous mode?

No, not yet. But maybe I should ...


Cheers,
Erich

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.15 (GNU/Linux)
Comment: Using GnuPG with Icedove - http://www.enigmail.net/
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=tFuz
-----END PGP SIGNATURE-----

Re: [Amforth] Watchdog prescalers

[David W. <ins...@gm...>]

>Are you sure, you want to do this with the watchdog timer? Isn't the
>watchdog timer only able to generate a reset?


Thanks for your code Erich. I was keen to use the watchdog timer because it
works with the lowest power mode (power down) and doesn't need external
components.

It seems to be a fairly common method. The watchdog can either generate a
reset or trigger an interrupt, or both. This interrupt can also be used as
a wake up from power down. (At least on the 324P that I'm using).

The other way I was considering was with timer 2 in synchronous mode (so I
don't need external components) and use power save mode. Did you ever look
at synchronous mode?

Many thanks,

David