Hi,

I am designing a module where low voltage threshold for interrupts
are desirable. A high of 0.5V and low of 0.2V are ideal but I am not sure
if current interrupt hardware can distinguish this kind of subtle
difference.I would appreciate it if anybody could tell me whether this is
possible, or how difficult it may be, or some reference systems which
has implemented such low voltage-threshold interrupts.

Thanks in advance,

lin

Lin Gu wrote:

>Hi,
>
>I am designing a module where low voltage threshold for interrupts
>are desirable. A high of 0.5V and low of 0.2V are ideal but I am not sure
>if current interrupt hardware can distinguish this kind of subtle
>difference.I would appreciate it if anybody could tell me whether this is
>possible, or how difficult it may be, or some reference systems which
>has implemented such low voltage-threshold interrupts.
>
>Thanks in advance,
>
>lin

The exact levels at which interrupt inputs trigger depends on the
technology of the parts used.  If, as you describe, you have specific
needs that differ from your part's characteristic, you could provide
additional circuitry, perhaps using op-amps, to process the input.

The major problem with your specification is that there is little
"dead band" between the high and low states, and your system will
suffer due to insufficient noise immunity.

-- 
========================================================================
          Michael Kesti            |  "And like, one and one don't make
                                   |   two, one and one make one."
          m...@gv.net            |          - The Who, Bargain

Lin Gu wrote:
> 
> I am designing a module where low voltage threshold for interrupts
> are desirable. A high of 0.5V and low of 0.2V are ideal but I am
> not sure if current interrupt hardware can distinguish this kind
> of subtle difference. I would appreciate it if anybody could tell
> me whether this is possible, or how difficult it may be, or some
> reference systems which has implemented such low
> voltage-threshold interrupts.

Interrupts have nothing to do with it.  You are interested in
logic levels.  Receivers for such low level swings will normally
be based on long-tailed pairs, op-amps, etc.  Your best bet for
cheap available ICs for the range might be RS485 receivers, which
will convert to the usual logic levels.

-- 
Chuck F (c...@yahoo.com) (c...@worldnet.att.net)
   Available for consulting/temporary embedded and systems.
   <http://cbfalconer.home.att.net>;  USE worldnet address!

CBFalconer wrote:
> Lin Gu wrote:
> 
>>I am designing a module where low voltage threshold for interrupts
>>are desirable. A high of 0.5V and low of 0.2V are ideal but I am
>>not sure if current interrupt hardware can distinguish this kind
>>of subtle difference. I would appreciate it if anybody could tell
>>me whether this is possible, or how difficult it may be, or some
>>reference systems which has implemented such low
>>voltage-threshold interrupts.
> 
> 
> Interrupts have nothing to do with it.  You are interested in
> logic levels.  Receivers for such low level swings will normally
> be based on long-tailed pairs, op-amps, etc.  Your best bet for
> cheap available ICs for the range might be RS485 receivers, which
> will convert to the usual logic levels.

Maybe what he's looking for is a comparator between the
interrupt pin and his source.  I can imagine some sort
of opto or inductive sensor that he wants to derive an
interrupt from.

"Michael R. Kesti" <m...@gv.net> wrote in message
news:4...@gv.net...
> Lin Gu wrote:
>
> >Hi,
> >
> >I am designing a module where low voltage threshold for interrupts
> >are desirable. A high of 0.5V and low of 0.2V are ideal but I am not sure
> >if current interrupt hardware can distinguish this kind of subtle
> >difference.I would appreciate it if anybody could tell me whether this is
> >possible, or how difficult it may be, or some reference systems which
> >has implemented such low voltage-threshold interrupts.
> >
> >Thanks in advance,
> >
> >lin
>
> The exact levels at which interrupt inputs trigger depends on the
> technology of the parts used.  If, as you describe, you have specific
> needs that differ from your part's characteristic, you could provide
> additional circuitry, perhaps using op-amps, to process the input.

I cannot use amplifiers. The circuit I am developing should use a low
power as possible and amplifiers or other preprocessing units usually
consume power.

> The major problem with your specification is that there is little
> "dead band" between the high and low states, and your system will
> suffer due to insufficient noise immunity.

I certainly agree. With help from google, I find some low-voltage
microcontrollers
have a "dead band" of about 0.9V but it is still to large for my device. Are
there
any microprocessors or interrupt handler circuit that can accept a high/low
of 0.5V/0.1V?

Thanks!

lin

Lin Gu wrote:
> 
> "Michael R. Kesti" <m...@gv.net> wrote in message
> news:4...@gv.net...
> > Lin Gu wrote:
> >
> > >Hi,
> > >
> > >I am designing a module where low voltage threshold for interrupts
> > >are desirable. A high of 0.5V and low of 0.2V are ideal but I am not sure
> > >if current interrupt hardware can distinguish this kind of subtle
> > >difference.I would appreciate it if anybody could tell me whether this is
> > >possible, or how difficult it may be, or some reference systems which
> > >has implemented such low voltage-threshold interrupts.
> > >
> > >Thanks in advance,
> > >
> > >lin
> >
> > The exact levels at which interrupt inputs trigger depends on the
> > technology of the parts used.  If, as you describe, you have specific
> > needs that differ from your part's characteristic, you could provide
> > additional circuitry, perhaps using op-amps, to process the input.
> 
> I cannot use amplifiers. The circuit I am developing should use a low
> power as possible and amplifiers or other preprocessing units usually
> consume power.
> 
> > The major problem with your specification is that there is little
> > "dead band" between the high and low states, and your system will
> > suffer due to insufficient noise immunity.
> 
> I certainly agree. With help from google, I find some low-voltage
> microcontrollers
> have a "dead band" of about 0.9V but it is still to large for my device. Are
> there
> any microprocessors or interrupt handler circuit that can accept a high/low
> of 0.5V/0.1V?

I don't know of *any* logic that works with IO voltages below 1 volt. 
You can get micropower analog components.  What you are trying to do is
not a normal digital function and should not be considered as such. 
Find a comparator with a low enough power consumption and work with
that.  Some even have a reference built in.

On Sun, 11 Jan 2004 19:34:51 -0500, "Lin Gu" <l g u @ i n a m e . c o
m> wrote:

>Hi,
>
>I am designing a module where low voltage threshold for interrupts
>are desirable. A high of 0.5V and low of 0.2V are ideal but I am not sure
>if current interrupt hardware can distinguish this kind of subtle
>difference.I would appreciate it if anybody could tell me whether this is
>possible, or how difficult it may be, or some reference systems which
>has implemented such low voltage-threshold interrupts.
>
>Thanks in advance,
>
>lin

If you want precise voltage thresholds like that, I think you'll need
to use an analog comparator to measure the voltage, with the
comparator's output generating a logic level signal to be used for the
processor's interrupt input.

Digital inputs to microcontrollers (or digital ICs) are not intended
for precision voltage measurements.



-- 
Peter Bennett, VE7CEI    
new newsgroup users info : http://vancouver-webpages.com/nnq
GPS and NMEA info: http://vancouver-webpages.com/peter
Vancouver Power Squadron: http://vancouver.powersquadron.ca

"Lin Gu" <l g u @ i n a m e . c o m> writes:

> I cannot use amplifiers. The circuit I am developing should use a low
> power as possible and amplifiers or other preprocessing units usually
> consume power.

This depends on what you consider low power. For example, the TLV370x family
of comparators consumes around 600 nA/channel. Is this acceptable?
(Note that the speed is rather moderate, around 40 us.) There are some
comparators with even lower quiescent current, but they tend to have an
open-drain output which then consumes some power.

Another possibility would be to use the internal comparator of a 
microcontroller. It consumes probably a lot more, but you can
switch it on only once in a while to check if something has
happened in the inputs, in case your speed requirements allow this.

Of course, your total power consumption will depend on how you make
the reference voltage for the comparator. You'll need to use megaohm
resistances in any case.

You could use an amplifier to amplify your input signals to acceptable
levels. However, linear amplifiers provide a worse power-to-speed
ratio than comparators.

- Ville

-- 
Ville Voipio, Dr.Tech., M.Sc. (EE)

Lin Gu wrote:
<snip>
> I cannot use amplifiers. The circuit I am developing should use a low
> power as possible and amplifiers or other preprocessing units usually
> consume power.
> 
  There are uC available that include very low power
Comparitors, that can drive the INTERRUPT logic.
  These allow you to define your own voltage threshold levels.
  Look at the www.Cygnal.com families, and also
the Philips P89LPC9xx families.
  IIRC, the Cygnal ones can run < 1uA of comparitor Icc adder.

-jg