Forums

Olimex ARM JTAG Signals levels

Started by johnnorgaard2003 October 20, 2004

Hi Olimex/Others

I have an Olimex ARM JTAG. It works ok on my homemade board if Vcc is
3.3 V. But if Vcc is 3.1 V is does not work.

I have look at the JTAG.
The board contain a 74AC244 (5V driver - power from PC)
The board contain a 74LCX244 (3.xV driver - power from target)

Signals from Target through 74LCX244 OK, but from 74LCX244 to 74AC244,
do not work 3.x volt to 5 volt system.

I think the 74AC244 should be replaced by a 74ACT244 ???

What is your opinion ? Best regards
John Noergaard




An Engineer's Guide to the LPC2100 Series

Hi John

If the datasheet specifies what your saying then there's not need for
opions as it's a bad design or simply a mistake brought about (let me
give you a practical example)

We have a secretary where I work and she doesn't know a resistor from a
transistor but she looks damn well good and does the accounting...

One day as she was doing the "shopping" at Arrow (I believe it was)

She was instructed to purchase the 74ACT244 (or what ever the 3.0V type
is called) but she found that she could get the 74??244 which in the
sales catalogue said "Super fast rama dama ding dong" and she thought
that it was better :o) The thing is we purchased a shit load of these
which we couldn't use and even though I'd like to punish her the dirty
way I don't think it would help...

Well back to the issue at hand :)

Maybe the same thing happened to Olimex?
I "know" that at least Tsvetan which I "know of" is a descend man and
knows what he's doing so It could well be a slip up I will recommend you
to exchange the dongle to Olimex or replace the IC your self

Regards
Lasse -----Original Message-----
From: johnnorgaard2003 [mailto:]
Sent: 20. oktober 2004 15:32
To:
Subject: [lpc2000] Olimex ARM JTAG Signals levels
Hi Olimex/Others

I have an Olimex ARM JTAG. It works ok on my homemade board if Vcc is
3.3 V. But if Vcc is 3.1 V is does not work.

I have look at the JTAG.
The board contain a 74AC244 (5V driver - power from PC)
The board contain a 74LCX244 (3.xV driver - power from target)

Signals from Target through 74LCX244 OK, but from 74LCX244 to 74AC244,
do not work 3.x volt to 5 volt system.

I think the 74AC244 should be replaced by a 74ACT244 ???

What is your opinion ? Best regards
John Noergaard

Yahoo! Groups Links




Hi Lasse

Yes I know, it was just because I mail my question directly to Olimex,
but the didn't answer. I know they read this yahoo group, that why !

best regards

John




> Hi Olimex/Others
>
> I have an Olimex ARM JTAG. It works ok on my homemade board if Vcc
is
> 3.3 V. But if Vcc is 3.1 V is does not work.
>
> I have look at the JTAG.
> The board contain a 74AC244 (5V driver - power from PC)
> The board contain a 74LCX244 (3.xV driver - power from target)
>
> Signals from Target through 74LCX244 OK, but from 74LCX244 to
74AC244,
> do not work 3.x volt to 5 volt system.

Hello John,

3.1V is near the down limit of the specification
if you have problems just increase the power to 3.3V while debugging
then if you product needs exactly 3V power use this power supply in
production and use ISP utility to load your code

> I think the 74AC244 should be replaced by a 74ACT244 ???
>
> What is your opinion ?

bad idea as LPT signals often are not 5V but down to 4.0-4.3V (at
least this is what we measured on two notebooks in office
ACT needs >4.5V to operate, so the option to rise your ARM power
supply to 3.2V while debugging seems less hassle, than to replace the
computer you use for debugging

Best regards
Tsvetan




--- In , "johnnorgaard2003" <john_2005@c...>
wrote:
>
> Hi Lasse
>
> Yes I know, it was just because I mail my question directly to
Olimex,
> but the didn't answer. I know they read this yahoo group, that why !
>
> best regards
>
> John

John,

I check and can't see any e-mail from you received in our database on
the subject above

tsvetan



On Wed, 20 Oct 2004 14:38:48 -0000, tsvetanusunov wrote:

> Hi Olimex/Others
>
> I have an Olimex ARM JTAG. It works ok on my homemade board if Vcc
is
> 3.3 V. But if Vcc is 3.1 V is does not work.
>
> I have look at the JTAG.
> The board contain a 74AC244 (5V driver - power from PC)
> The board contain a 74LCX244 (3.xV driver - power from target)
>
> Signals from Target through 74LCX244 OK, but from 74LCX244 to
74AC244,
> do not work 3.x volt to 5 volt system.

Hello John,

3.1V is near the down limit of the specification
if you have problems just increase the power to 3.3V while debugging
then if you product needs exactly 3V power use this power supply in
production and use ISP utility to load your code

> I think the 74AC244 should be replaced by a 74ACT244 ???
>
> What is your opinion ?

bad idea as LPT signals often are not 5V but down to 4.0-4.3V (at
least this is what we measured on two notebooks in office
ACT needs >4.5V to operate, so the option to rise your ARM power
supply to 3.2V while debugging seems less hassle, than to replace the
computer you use for debugging

Best regards
Tsvetan
===================================================================
I'm not sure it is a bad idea. The 74LCX244 connects to the ARM and
gets its power (~3.3v) from the target board. The 74AC244 connects to the PC
and gets its 5v from there. The problem might be that the signal(s)
FROM the 74LCX244 TO the 74AC244 are not above the 74AC244's input threshold
level. Changing the 74AC244 to a 74ACT244 would lower the input threshold
requirement for the signal(s) coming from the 74LCX244 while still maintaining
compatablility with the PC signals.

Of course making sure the target supply voltage was 3.3v would help too.

BTW: I checked my Macraigor WIGGLER and it also uses the 74AC244 and 74LCX244
parts.

Regards
-Bill Knight
R O SoftWare Yahoo! Groups Links



> I'm not sure it is a bad idea. The 74LCX244 connects to the ARM and
> gets its power (~3.3v) from the target board. The 74AC244 connects
to the PC and gets its 5v from there.

Bill,
you don't follow Johns post - he is powering the target with 3V and
the problem comes from LCX-to-AC direction as AC have threshold 3V
for "1" high level and this is on the edge when LCX is powered with
3.1V only

>The problem might be that the signal(s)
> FROM the 74LCX244 TO the 74AC244 are not above the 74AC244's input
threshold
> level. Changing the 74AC244 to a 74ACT244 would lower the input
threshold
> requirement for the signal(s) coming from the 74LCX244 while still
maintaining
> compatablility with the PC signals.

of course this is OK if *all* LPTs provide >4.5V which ACT will need
as this part is powered from LPT

Best regards
Tsvetan




On Wed, 20 Oct 2004 15:25:15 -0000, tsvetanusunov wrote:

> I'm not sure it is a bad idea. The 74LCX244 connects to the ARM and
> gets its power (~3.3v) from the target board. The 74AC244 connects
to the PC and gets its 5v from there.

Bill,
you don't follow Johns post - he is powering the target with 3V and
the problem comes from LCX-to-AC direction as AC have threshold 3V
for "1" high level and this is on the edge when LCX is powered with
3.1V only

>The problem might be that the signal(s)
> FROM the 74LCX244 TO the 74AC244 are not above the 74AC244's input
threshold
> level. Changing the 74AC244 to a 74ACT244 would lower the input
threshold
> requirement for the signal(s) coming from the 74LCX244 while still
maintaining
> compatablility with the PC signals.

of course this is OK if *all* LPTs provide >4.5V which ACT will need
as this part is powered from LPT

Best regards
Tsvetan
=========================================================================
It was my understanding that the 74AC244 got its power indirectly from
the PC's parallel port not from the target.

-Bill


>From: "Bill Knight" <>
>Reply-To:
>To: "" <>
>Subject: Re: [lpc2000] Re: Olimex ARM JTAG Signals levels
>Date: Wed, 20 Oct 2004 10:03:21 -0500 >On Wed, 20 Oct 2004 14:38:48 -0000, tsvetanusunov wrote:
>
> > Hi Olimex/Others
> >
> > I have an Olimex ARM JTAG. It works ok on my homemade board if Vcc
>is
> > 3.3 V. But if Vcc is 3.1 V is does not work.
> >
> > I have look at the JTAG.
> > The board contain a 74AC244 (5V driver - power from PC)
> > The board contain a 74LCX244 (3.xV driver - power from target)
> >
> > Signals from Target through 74LCX244 OK, but from 74LCX244 to
>74AC244,
> > do not work 3.x volt to 5 volt system.
>
>Hello John,
>
>3.1V is near the down limit of the specification
>if you have problems just increase the power to 3.3V while debugging
>then if you product needs exactly 3V power use this power supply in
>production and use ISP utility to load your code
>
> > I think the 74AC244 should be replaced by a 74ACT244 ???
> >
> > What is your opinion ?
>
>bad idea as LPT signals often are not 5V but down to 4.0-4.3V (at
>least this is what we measured on two notebooks in office
>ACT needs >4.5V to operate, so the option to rise your ARM power
>supply to 3.2V while debugging seems less hassle, than to replace the
>computer you use for debugging
>
>Best regards
>Tsvetan
>===================================================================
>I'm not sure it is a bad idea. The 74LCX244 connects to the ARM and
>gets its power (~3.3v) from the target board. The 74AC244 connects to the
>PC
>and gets its 5v from there. The problem might be that the signal(s)
>FROM the 74LCX244 TO the 74AC244 are not above the 74AC244's input
>threshold
>level. Changing the 74AC244 to a 74ACT244 would lower the input threshold
>requirement for the signal(s) coming from the 74LCX244 while still
>maintaining
>compatablility with the PC signals.
>
>Of course making sure the target supply voltage was 3.3v would help too.
>
>BTW: I checked my Macraigor WIGGLER and it also uses the 74AC244 and
>74LCX244
>parts.
>

I just use a 74HC244 on my own JTAG interface (it's spec'd down to 2.0 V
supply). I haven't checked it at anything other than a nominal 3.3V, though.

Leon