langwadt@fonz.dk skrev:
> On 13 Sep., 20:48, Anton Erasmus <nob...@spam.prevent.net> wrote:
>> On Sun, 12 Sep 2010 11:08:02 -0700 (PDT), larwe <zwsdot...@gmail.com>
>> wrote:
>>
>>> I'm using ATmega169P in an application and just came across a need for
>>> an additional two ADC channels (0-3 are in use). Unfortunately on this
>>> part, ADC4-7 are muxed with JTAG pins. I've had no luck getting the
>>> micro to read those channels
>>> Is it necessary to disable the JTAG interface via fuse settings in
>>> order to use these upper ADCs? That's a real drag for me... means I
>>> have to dig out my AVR-ISP to program it over SPI, and I won't be able
>>> to do any further debugging :(
>>> [Obviously I am not expecting to use the JTAG interface simultaneously
>>> with having those pins work as ADCs. I just don't want to lock off the
>>> JTAG just yet...]
>>> thanks.
>> You can disable the JTAG pins in software. Obviously you cannot then
>> debug via JTAG from that point, but it means you can still use the
>> JTAG pins fro programming the device.
>>
>> Try this snippet I got off the web.
>>
>> void disable_JTAG(void)
>> {
>>         unsigned char sreg;
>>
>>         sreg = SREG;
>>         cli();
>>         MCUCR |= ( 1 <<JTD );
>>         MCUCR |= ( 1 <<JTD );
>>         SREG = sreg;
>>
>> }
>>
>> You have to access the register twice in 4 cycles.
>>
> 
> yep the 4 cycles is important, remember doing multiple variations of
> code and compiler
> setting before I gave up and wrote it in asm to make it work
> 
> -Lasse

Obviously, you can use the pin change interrupt on another pin to 
enable/disable the JTAG as needed.
If the pin is also enables a buffer, you don't have to remove the
JTAGICE Mk II.

-- 
Best Regards
Ulf Samuelsson
These are my own personal opinions, which may
or may not be shared by my employer Atmel Nordic AB

On 13 Sep., 20:48, Anton Erasmus <nob...@spam.prevent.net> wrote:
> On Sun, 12 Sep 2010 11:08:02 -0700 (PDT), larwe <zwsdot...@gmail.com>
> wrote:
>
> >I'm using ATmega169P in an application and just came across a need for
> >an additional two ADC channels (0-3 are in use). Unfortunately on this
> >part, ADC4-7 are muxed with JTAG pins. I've had no luck getting the
> >micro to read those channels
>
> >Is it necessary to disable the JTAG interface via fuse settings in
> >order to use these upper ADCs? That's a real drag for me... means I
> >have to dig out my AVR-ISP to program it over SPI, and I won't be able
> >to do any further debugging :(
>
> >[Obviously I am not expecting to use the JTAG interface simultaneously
> >with having those pins work as ADCs. I just don't want to lock off the
> >JTAG just yet...]
>
> >thanks.
>
> You can disable the JTAG pins in software. Obviously you cannot then
> debug via JTAG from that point, but it means you can still use the
> JTAG pins fro programming the device.
>
> Try this snippet I got off the web.
>
> void disable_JTAG(void)
> {
> =A0 =A0 =A0 =A0 unsigned char sreg;
>
> =A0 =A0 =A0 =A0 sreg =3D SREG;
> =A0 =A0 =A0 =A0 cli();
> =A0 =A0 =A0 =A0 MCUCR |=3D ( 1 <<JTD );
> =A0 =A0 =A0 =A0 MCUCR |=3D ( 1 <<JTD );
> =A0 =A0 =A0 =A0 SREG =3D sreg;
>
> }
>
> You have to access the register twice in 4 cycles.
>

yep the 4 cycles is important, remember doing multiple variations of
code and compiler
setting before I gave up and wrote it in asm to make it work

-Lasse

On Sep 13, 11:48=A0am, Anton Erasmus <nob...@spam.prevent.net> wrote:
> On Sun, 12 Sep 2010 11:08:02 -0700 (PDT), larwe <zwsdot...@gmail.com>
> wrote:
>
> >I'm using ATmega169P in an application and just came across a need for
> >an additional two ADC channels (0-3 are in use). Unfortunately on this
> >part, ADC4-7 are muxed with JTAG pins. I've had no luck getting the
> >micro to read those channels
>
> >Is it necessary to disable the JTAG interface via fuse settings in
> >order to use these upper ADCs? That's a real drag for me... means I
> >have to dig out my AVR-ISP to program it over SPI, and I won't be able
> >to do any further debugging :(
>
> >[Obviously I am not expecting to use the JTAG interface simultaneously
> >with having those pins work as ADCs. I just don't want to lock off the
> >JTAG just yet...]
>
> >thanks.
>
> You can disable the JTAG pins in software. Obviously you cannot then
> debug via JTAG from that point, but it means you can still use the
> JTAG pins fro programming the device.

Yes, for the first time.  If you have the software running with jtag
disabled, you still have to erase/reprogram it with SPI.

>
> Try this snippet I got off the web.
>
> void disable_JTAG(void)
> {
> =A0 =A0 =A0 =A0 unsigned char sreg;
>
> =A0 =A0 =A0 =A0 sreg =3D SREG;
> =A0 =A0 =A0 =A0 cli();
> =A0 =A0 =A0 =A0 MCUCR |=3D ( 1 <<JTD );
> =A0 =A0 =A0 =A0 MCUCR |=3D ( 1 <<JTD );
> =A0 =A0 =A0 =A0 SREG =3D sreg;
>
> }
>
> You have to access the register twice in 4 cycles.
>
> Regards
> =A0 =A0Anton Erasmus

On Sun, 12 Sep 2010 11:08:02 -0700 (PDT), larwe <zwsdotcom@gmail.com>
wrote:

>I'm using ATmega169P in an application and just came across a need for
>an additional two ADC channels (0-3 are in use). Unfortunately on this
>part, ADC4-7 are muxed with JTAG pins. I've had no luck getting the
>micro to read those channels
>
>Is it necessary to disable the JTAG interface via fuse settings in
>order to use these upper ADCs? That's a real drag for me... means I
>have to dig out my AVR-ISP to program it over SPI, and I won't be able
>to do any further debugging :(
>
>[Obviously I am not expecting to use the JTAG interface simultaneously
>with having those pins work as ADCs. I just don't want to lock off the
>JTAG just yet...]
>
>thanks.

You can disable the JTAG pins in software. Obviously you cannot then
debug via JTAG from that point, but it means you can still use the
JTAG pins fro programming the device.

Try this snippet I got off the web.

void disable_JTAG(void)
{
	unsigned char sreg;

	sreg = SREG;
	cli();
	MCUCR |= ( 1 <<JTD );
	MCUCR |= ( 1 <<JTD );
	SREG = sreg;
}


You have to access the register twice in 4 cycles.

Regards
   Anton Erasmus

On Sep 12, 11:17=A0pm, kevin93 <ke...@whitedigs.com> wrote:

> You can disable the JTAG interface under program control even if the
> fuses have it enabled. =A0So depending on your application it may be

I didn't know that! Thanks for the info.

"Hans-Bernhard Br&#4294967295;ker" <HBBroeker@t-online.de> wrote in message 
news:i6jb84$tli$00$1@news.t-online.com...
> On 12.09.2010 20:51, larwe wrote:
>
>> Bugger. That answer is concise, accurate, and a complete pain in my
>> rump.
>
> Hmm... reminds me of that joke about an answer that must have been given 
> by a mathematician because it's so wonderfully short, extremely correct, 
> .... and completely useless.

Hmm, I'm not short, when correct extremely so, and completely useless 
(according to the wife)... and of course a mathematician by training. I take 
offense that mathematical answers can be the butt of jokes. Many modern 
theorems are extremely long. The are correct or not, so 'extremely' is 
superfluous. Further, the answer can't be useless if it is an answer to the 
question unless the question itself was set up that way.

:-)
Peter

On Sep 12, 11:08=A0am, larwe <zwsdot...@gmail.com> wrote:
> I'm using ATmega169P in an application and just came across a need for
> an additional two ADC channels (0-3 are in use). Unfortunately on this
> part, ADC4-7 are muxed with JTAG pins. I've had no luck getting the
> micro to read those channels
>
> Is it necessary to disable the JTAG interface via fuse settings in
> order to use these upper ADCs? That's a real drag for me... means I
> have to dig out my AVR-ISP to program it over SPI, and I won't be able
> to do any further debugging :(
>
> [Obviously I am not expecting to use the JTAG interface simultaneously
> with having those pins work as ADCs. I just don't want to lock off the
> JTAG just yet...]
>
> thanks.

You can disable the JTAG interface under program control even if the
fuses have it enabled.  So depending on your application it may be
appropriate to have JTAG disabled when needed by the program and
enabled when you need it for programming/debugging.

kevin

Hans-Bernhard Br=F6ker wrote:

> On 12.09.2010 20:51, larwe wrote:
>=20
>> Bugger. That answer is concise, accurate, and a complete pain in my
>> rump.
>=20
>=20
> Hmm... reminds me of that joke about an answer that must have been give=
n=20
> by a mathematician because it's so wonderfully short, extremely correct=
,=20
> ..... and completely useless.

Reminds me of that joke about an answer that must have been given by a=20
manager: somebody else's design problem from now on must become my=20
personal problem.

VLV

On 12.09.2010 20:51, larwe wrote:

> Bugger. That answer is concise, accurate, and a complete pain in my
> rump.

Hmm... reminds me of that joke about an answer that must have been given 
by a mathematician because it's so wonderfully short, extremely correct, 
.... and completely useless.