Searching \ for '[EE]: Low Battery LED' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/power/batterys.htm?key=battery
Search entire site for: 'Low Battery LED'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Low Battery LED'
2002\07\23@184642 by Kieren Johnstone

picon face
Howdy..

I want to create a "low battery" warning LED, to warn me when the voltage drops below say 4V (from 4.5 supplied from 3xAA)... I told my brain to work something out, the best it came up with was putting some very high resistance resistor feeding the +ve supply to the base of a transistor, then the collector-emitter current would die (supposedly) so I could do that inverter thingy :)  Then I told my brain to construct an example, it built what I just described (minus an inversion), all that arose was an LED slightly dimmer than the "power on" indicator LED.  Anyone wanna help me solve this? :)

Oh... and.. please *sob* help me *sniff* with my [PIC]: Flashing LED In Practise :)

Thanks again,
Kieren

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@192055 by Brendan Moran

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

> Howdy..
>
>I want to create a "low battery" warning LED, to warn me when the
>voltage drops  below say 4V (from 4.5 supplied from 3xAA)... I told
>my brain to work something  out, the best it came up with was
>putting some very high resistance resistor feeding the  +ve supply
>to the base of a transistor, then the collector-emitter current
>would die  (supposedly) so I could do that inverter thingy :)  Then
>I told my brain to construct an >example, it built what I just
>described (minus an inversion), all that arose was an LED >slightly
>dimmer than the "power on" indicator LED.  Anyone wanna help me
>solve >this? :)

Have you checked the http://www.piclist.com PICList archives for someone
else's solution yet?

What your brain neglected to remind you of is that a transistor is
not always a switch.  It can also be an amplifier.  This is what
happens when a transistor comes into the "linear region" of its
operation, and this is what happened to you.

I think that you need to use an opamp for this.  You will also need a
reference voltage.

A reference voltage can be acquired in many ways.  The simplest of
which requires a zener and a resistor.  All it is is a constant
voltage source, lower than the supply voltage can ever get.

Here comes the ASCII art (copy it into some editor with a monospaced
font, such as notepad to view it if it doesn't appear correctly in
your email program.

                                                      Vcc
            R1         Vin     |\                      o
 Vcc o----/\/\/\/---+----------|+\                     |
                    |          |  \    LED             |
                    \          |   >---|<|---/\/\/\/---+
                    /          |  /
                R2  \  Vref o--|-/
                    /          |/
                    \
                    |
                   ---
                  ///

Now what you need to know is this:  The opamp (the triangle thingy)
is set up as a comparator in this circuit (opamps can be amplifiers
too).  It compares the reference voltage (Vref) and a fraction of Vcc
(Vin).

Vin is connected to the non-inverting input of the opamp, while Vref
is connected to the inverting input.

Vin is determined by the following equation:

Vin = Vcc*R2/(R1+R2)

R2 and R1 need to be configured such that 4.0V(if that is the voltge
you choose to count as a "low battery") triggers the opamp, or

4.0*R2/(R1+R2) = Vref

When Vin becomes lower than Vref, the opamp will turn the LED by
setting its output low, which forward biases the LED.

The led is on if:

Vref - Vcc*R2/(R1+R2) > 0

This should help you, I think.  You should be able to use an LM741 to
do this.  Almost any opamp will do the job.

>Oh... and.. please *sob* help me *sniff* with my [PIC]: Flashing LED
>In Practise :)

Why don't you search for "FPP" and "tait" on google, and see what you
find.  There should be some information (And maybe even a message
board ;)  If that doesn't help, try adding something like tutorial,
or FAQ or troubleshooting.

Good luck,

- --Brendan

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>

iQA/AwUBPT3ktQVk8xtQuK+BEQJ/TwCg+azX696efBg4B82DXAuNrXUuN3UAoNT9
vebvwm9J4hY2aOwBbkJot3FQ
=javJ
-----END PGP SIGNATURE-----

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@194315 by Alexandre Souza

flavicon
face
>
>
>I want to create a "low battery" warning LED, to warn me when the voltage drops below say 4V (from 4.5 supplied from 3xAA)... I told my brain to work something out, the best it came up with was putting some very high resistance resistor feeding the +ve supply to the base of a transistor, then the collector-emitter current would die (supposedly) so I could do that inverter thingy :)  Then I told my brain to construct an example, it built what I just described (minus an inversion), all that arose was an LED slightly dimmer than the "power on" indicator LED.  Anyone wanna help me solve this? :)
>
>
   The "easier" way of doing that is using a LM339 comparator, or
something like that. :o)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@201343 by Andrew Warren

flavicon
face
Kieren Johnstone <spam_OUTPICLISTTakeThisOuTspammitvma.mit.edu> wrote:

> I want to create a "low battery" warning LED, to warn me when the
> voltage drops below say 4V (from 4.5 supplied from 3xAA)... [When I
> tried,] all that arose was an LED slightly dimmer than the "power
> on" indicator LED.

Kieren:

Doing this with two LEDs (one always on and another that turns on
when the battery's low) draws MORE current from the battery when you
can least afford to do so:  During the period that your user is being
warned to replace the battery soon.

If it's important to give your users as much time as possible to
replace the batteries, you might want instead to use just one LED
which turns OFF when the battery's low.

Your "power-on LED" circuit probably looks like this now (please
forgive the lack of pretty ASCII graphics; I'm in a bit of a hurry):


   Vbatt ---- switch ----a LED c---- Resistor ---- GND

If you add a zener diode, the LED will indicate a low-battery
condition by refusing to light when Vbatt is below the zener voltage:


   Vbatt ---- Switch ----a LED c---- Resistor ----c Zener a---- GND

-Andy

=== Andrew Warren -- .....aiwKILLspamspam@spam@cypress.com
=== Principal Design Engineer
=== Cypress Semiconductor Corporation
===
=== Opinions expressed above do not
=== necessarily represent those of
=== Cypress Semiconductor Corporation

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@210559 by Jinx

face picon face
part 1 1824 bytes content-type:text/plain; (decoded 7bit)

> Howdy..

Ayoop

> I want to create a "low battery" warning LED, to warn me when the
> voltage drops below say 4V (from 4.5 supplied from 3xAA)...

I use this circuit on 6V gel cells. There's a little more capacity available
than AAs but the TLC or any low A op amp should run at a few uA. Also
in the circuit is an LDO regulator (Vreg), which may not be suitable in
your case, but a zener might do. Depends on what lifetime you expect
from the batteries. B1 in the diagram is power supplied by a PIC pin,
so normally the only current consumed is across the resistors. Depending
on the input current of op amp used you could probably increase those
resistors considerably to reduce quiescent current. The unmarked
resistor sets the trip voltage

There are some special function ICs that have a low battery warning
as a secondary function, but that probably won't do you

You could try a low current comparator, like some of Microchip's

www.microchip.com/1000/pline/analog/anicateg/linear/linercat/comp/ind
ex.htm

Some are sub-uA

===========================

Guess 1

Rudimentary low voltage detection by measuring the time it takes to
charge a capacitor. IF you use an ST input and use a source voltage
that just exceeds the ST threshold at the battery voltage you wish to
detect. As V+ drops, so will the charging source and the the ST threshold
will be later in the charge curve, thus taking longer to switch. I have a
feeling though that the ST threshold is a function of V+, in which case it
might not / probably won't work

Guess 2

A resistive divider used slightly above the ST threshold. Based on same
assumption as above

Guess 3

Measure period of a VCO (VCO powered by PIC pin). Can use a
reasonably high current as it's needed only occassionally


part 2 1584 bytes content-type:image/gif; (decode)


part 3 131 bytes
--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@212607 by Tom Messenger

flavicon
face
At 11:45 PM 7/23/02 +0100, Kieren wrote:
>Howdy..
>
>I want to create a "low battery" warning LED, to warn me when the voltage
drops below say 4V (from 4.5 supplied from 3xAA)...


Best bet: define it as a software problem and make the software person deal
with it. What's that? You say *you* are the software person? Too bad:)

Software solution: requires at least one PIC pin and some resistors, LED,
and a capacitor. Hook a resistor from +V to a cap to ground.  Run a
resistor from the RC junction to a PIC pin. Now the software part.

Set the pin to be an output. Set the output LOW. A little later set it back
to an input. Now check the pin to see if it is still low. The cap that you
dumped by setting it low is now charging up. If the pin is not high, count
bananas until it is. One banana, two bananas, ...., 119 bananas, hey, now
the pin's high. Do this with your full power supply voltage to see how many
bananas get counted. Then do it with 4.5 volts and then with 4.0 volts.
Remember this. Put it into your code.

Periodically in your code, check the RC network to measure the power
supply.  If the supply is high, maybe you light up an LED 100%. If the
supply is half way down to the cutoff point, maybe you blink it 50-50. If
the power is at the point of no return, ie, where you want to warn, then
slow down the rate further. Or turn on a LED of a different color. Or...
whatever else turns you on while you eat all those bananas.

Scott Dattalo has good information on this technique at his website:

       http://www.dattalo.com/technical/software/software.html

Check out his excellent notes called a2d.asm and the next time you need a
simple analog to digital converter, try this technique.

Tom M.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@223642 by Jinx

face picon face
> Scott Dattalo has good information on this technique at his website:
>
>         http://www.dattalo.com/technical/software/software.html

I have something similar here

http://home.clear.net.nz/pages/joecolquitt/0pots.html

BTW Kieren, I put you wrong, kind of, with that attachment. The
"unmarked" resistor is in fact the 680k (it's unmarked in the pic
I thought I was attaching). It sets the low battery voltage to 5V6 by
comparing the 680k + 560k divisor to the 1/2 V+ set by the two 470k

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\23@224257 by Dave Dilatush

picon face
part 1 1587 bytes content-type:text/plain; charset=us-ascii (decoded 7bit)

Kieren Johnstone wrote...

>I want to create a "low battery" warning LED, to warn me when the
>voltage drops below say 4V (from 4.5 supplied from 3xAA)... I
>told my brain to work something out, the best it came up with
>was putting some very high resistance resistor feeding the +ve
>supply to the base of a transistor, then the collector-emitter
>current would die (supposedly) so I could do that inverter thingy :)
>Then I told my brain to construct an example, it built what I just
>described (minus an inversion), all that arose was an LED slightly
>dimmer than the "power on" indicator LED.  Anyone wanna help me
>solve this? :)

For a more or less practical (depending on what precision you're
looking for) low-battery indicator circuit, try the attached.

R1 will probably have to be adjusted to get the right trip
voltage, since transistors differ in the base-to-emitter voltage
needed to get them to conduct (typically this is about 0.6 volts,
but can be as low as 0.45 volts or as high as 0.75 volts
depending on the transistor).  Make R1 larger to increase the
trip voltage, smaller to reduce it.

Resistor R5 can be omitted if desired; it just gives the circuit
a slight amount of hysteresis for a "snap-action" response.

Be aware that with a cheap-and-dirty circuit like this, the trip
voltage is going to change a bit with temperature, to the tune of
roughly -0.3% per degree celsius.  This may or may not be
acceptable depending on your needs.

Hope this helps a bit...

DD


part 2 1861 bytes content-type:image/gif; name=lowbatt.gif (decode)


part 3 131 bytes
--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\07\24@035827 by Alan B. Pearce

face picon face
>Doing this with two LEDs (one always on and another that turns on
>when the battery's low) draws MORE current from the battery when you
>can least afford to do so:  During the period that your user is being
>warned to replace the battery soon.

However doing it this way may stop another possibly confusing situation.

Consider what happens when you have one LED on while the battery is good.
When the battery voltage drops sufficiently the LED is turned off to
indicate "low battery". Now because of the reduced current draw the battery
voltage rises, and so the LED turns on again. Now because the battery draw
has risen the voltage falls and so the LED goes off..... etc.

If the check is being done by a micro then it is easy to get this
oscillation to be a sufficiently long period to make the LED flash in a
manner that means something to the user. If the frequency is not controlled
then it may be too fast to mean anything, to a point where it may just
appear that the LED has gone dim.

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamKILLspammitvma.mit.edu with SET PICList DIGEST in the body


2002\07\24@041254 by Alan B. Pearce

face picon face
>Set the pin to be an output. Set the output LOW. A little later set it back
>to an input. Now check the pin to see if it is still low. The cap that you
>dumped by setting it low is now charging up. If the pin is not high, count
>bananas until it is. One banana, two bananas, ...., 119 bananas, hey, now
>the pin's high. Do this with your full power supply voltage to see how many
>bananas get counted. Then do it with 4.5 volts and then with 4.0 volts.
>Remember this. Put it into your code.

There is another trick here that I have thought of using, and it relates
quite easily to using battery powered equipment.

If needing to make a number of analogue measurements, set the system up to
use the Vref+ as the reference for all the analogue measurements. This
should be at a voltage that will be stable even at battery end of life. Now
all external measurements use this, and will be a valid value right through
to battery end of life, because the reference is good till then.

When measuring the battery voltage switch to using the internal 5V
reference. Now measure the value of the Vref+ pin. Remember that the value
obtained will be inverse to the battery state, i.e. the larger value given
by the A/D means a lower battery voltage. Unless you are really looking for
punishment this does not matter, as you are only looking for the battery end
point to report low battery. It may mean setting two end points in your
code - one for flat battery, and one for low battery, and these would need
to be determined experimentally, but that would not be a big deal.

Result is no extra pin needed to measure the PIC supply voltage, while
having stable measurement of everything else right through to battery end of
life.

Off course this does not work too well if there is a regulator supplying the
PIC off something like a car battery, but then you will be able to use the
internal 5V reference as it will be stable anyway, so you can use the Vref+
pin as an analogue input for the battery voltage through a suitable divider
:)

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam.....mitvma.mit.edu with SET PICList DIGEST in the body


2002\07\24@090517 by Roman Black

flavicon
face
Dave Dilatush wrote:
>
> Kieren Johnstone wrote...
>
> >I want to create a "low battery" warning LED, to warn me when the
> >voltage drops below say 4V (from 4.5 supplied from 3xAA)... I
> >told my brain to work something out,

> For a more or less practical (depending on what precision you're
> looking for) low-battery indicator circuit, try the attached.

>                               Name: lowbatt.gif


We discussed this at length on the list a few months
back, all you need for the PIC to detect battery limit
(ie threshold) is one resistor and one capacitor. The
PIC discharges the C, and counts the number of cycles
until the cap charges and activates the PIC input. It's
reliable to a few percent and costs almost nothing. :o)
-Roman

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email EraseMElistservspam_OUTspamTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body


2002\07\24@173152 by Dave Dilatush

picon face
Roman wrote...
>Dave Dilatush wrote:
>> Kieren Johnstone wrote...
>>
>> >I want to create a "low battery" warning LED, to warn me when the
>> >voltage drops below say 4V (from 4.5 supplied from 3xAA)... I
>> >told my brain to work something out,
>
>> For a more or less practical (depending on what precision you're
>> looking for) low-battery indicator circuit, try the attached.
>
>We discussed this at length on the list a few months
>back, all you need for the PIC to detect battery limit
>(ie threshold) is one resistor and one capacitor. The
>PIC discharges the C, and counts the number of cycles
>until the cap charges and activates the PIC input. It's
>reliable to a few percent and costs almost nothing. :o)

This would work OK if the PIC were operated from a regulated VCC,
in which case the PIC's logic thresholds are constant and can be
used as a reference.

But in Kieren's circuit, he's operating the PIC directly off 4.5V
from three AA cells in series and it's this voltage he wishes to
monitor.  Since his "reference" (i.e., the PIC input's VIH level)
is varying in approximate proportion to the voltage he's trying
to measure, he'd get a more or less invariant result out of this
RC scheme.

In any case, both the circuit I showed and this RC timing method
are pretty crude; anyone who wants an accurate low-battery
indicator would be best off using a proper voltage reference and
a comparator.

DD

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2002\07\24@184254 by Dwayne Reid

flavicon
face
At 09:24 PM 7/24/02 +0000, Dave Dilatush wrote:

>This would work OK if the PIC were operated from a regulated VCC,
>in which case the PIC's logic thresholds are constant and can be
>used as a reference.
>
>But in Kieren's circuit, he's operating the PIC directly off 4.5V
>from three AA cells in series and it's this voltage he wishes to
>monitor.  Since his "reference" (i.e., the PIC input's VIH level)
>is varying in approximate proportion to the voltage he's trying
>to measure, he'd get a more or less invariant result out of this
>RC scheme.

One easy way around this might be to use a LED as a reference.  Use the RC
measurement thingy to measure the apparent voltage across the power
LED.  As the supply voltage drops, the measured voltage will appear to RISE
since the reference is dropping.  This should be reasonably repeatable so
long as one uses LEDs of the same part number from the same manufacturer -
this is a reasonable limitation in most cases.

You could also use a PIC pin to flash the power LED by putting a short
across the LED - this increases the supply current and causes the battery
to die faster but does ensure that the device does not spend much time in
that nebulous region where the battery is almost dead.

dwayne

--
Dwayne Reid   <@spam@dwaynerKILLspamspamplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 18 years of Engineering Innovation (1984 - 2002)
 .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
    `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email KILLspamlistservKILLspamspammitvma.mit.edu with SET PICList DIGEST in the body


2002\07\24@195332 by Dave Dilatush

picon face
Dwayne Reid wrote...

>One easy way around this might be to use a LED as a reference.  Use the RC
>measurement thingy to measure the apparent voltage across the power
>LED.  As the supply voltage drops, the measured voltage will appear to RISE
>since the reference is dropping.  This should be reasonably repeatable so
>long as one uses LEDs of the same part number from the same manufacturer -
>this is a reasonable limitation in most cases.

There's a catch, though: this RC voltage measuring technique can
only measure voltages that are higher than the VIH threshold of
the PIC's input pin.  And I think the LED voltage is going to be
substantially lower than that.

But you could turn it upside-down, and use the RC technique to
measure the drop across the LED's feed resistor.  Not as
straightforward, but it might work well enough to be useable.

>You could also use a PIC pin to flash the power LED by putting a short
>across the LED - this increases the supply current and causes the battery
>to die faster but does ensure that the device does not spend much time in
>that nebulous region where the battery is almost dead.

The circuit I posted yesterday can be operated from a PIC output
pin and made to flash.  In fact, if the collector of Q2 is
connected to a second PIC pin configured as an input, the PIC can
drastically reduce the power consumption of the low battery
indicator by strobing it every second or two with a short pulse
(just a couple of microseconds long); and leaving it on long
enough to make the LED visible only when Q2's collector is seen
to go low.

Takes two PIC pins, but that would give him a nearly zero-power
battery level monitor.

DD

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservTakeThisOuTspammitvma.mit.edu with SET PICList DIGEST in the body


2002\07\25@025904 by Roman Black

flavicon
face
Dave Dilatush wrote:
{Quote hidden}

Not so, that was also discussed in the old thread. :o)
The PIC input->hi threshold for the TTL inputs is
0.25Vdd + 0.8v, where the 0.8v remains constant as it
is a semiconductor junction within the device.

So as the Vdd itself changes, the input->hi trigger
point from charging the cap WILL give a simple battery
low detection. Yes it is crude, and will have a few
of percent tolerance, but in some cases that is all
you need. :o)
-Roman

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestspamBeGonespammitvma.mit.edu


2002\07\25@061059 by Dave Dilatush

picon face
Roman wrote...

>Dave Dilatush wrote:
>> This would work OK if the PIC were operated from a regulated VCC,
>> in which case the PIC's logic thresholds are constant and can be
>> used as a reference.
>
>Not so, that was also discussed in the old thread. :o)
>The PIC input->hi threshold for the TTL inputs is
>0.25Vdd + 0.8v, where the 0.8v remains constant as it
>is a semiconductor junction within the device.

0.25Vdd + 0.8v is not the input high threshold for the TTL
inputs; it is VIH(min), the minimum voltage guaranteed to be a
logic high.  The actual threshold can be anywhere between
VIH(min) and VIL(max), which is 0.15Vdd.

>So as the Vdd itself changes, the input->hi trigger
>point from charging the cap WILL give a simple battery
>low detection. Yes it is crude, and will have a few
>of percent tolerance, but in some cases that is all
>you need. :o)

From the above, the input threshold at Vcc=4.0V can be anywhere
from 0.6V to 1.8 volts, a 3:1 range.  That's a lot more than a
"few percent tolerance".

As a general rule, it's not a good idea to use logic threshold
levels to measure voltage unless you can tolerate a LOT of
uncertainty in the results.

DD

--
http://www.piclist.com hint: To leave the PICList
TakeThisOuTpiclist-unsubscribe-requestEraseMEspamspam_OUTmitvma.mit.edu


2002\07\25@141325 by Dwayne Reid

flavicon
face
At 11:45 PM 7/24/02 +0000, Dave Dilatush wrote:
>Dwayne Reid wrote...
>
> >One easy way around this might be to use a LED as a reference.  Use the RC
> >measurement thingy to measure the apparent voltage across the power
> >LED.  As the supply voltage drops, the measured voltage will appear to RISE
> >since the reference is dropping.  This should be reasonably repeatable so
> >long as one uses LEDs of the same part number from the same manufacturer -
> >this is a reasonable limitation in most cases.
>
>There's a catch, though: this RC voltage measuring technique can
>only measure voltages that are higher than the VIH threshold of
>the PIC's input pin.  And I think the LED voltage is going to be
>substantially lower than that.

VIH is about 1.4V for a non-ST input when Vdd=5V.  Vf for a red LED is
about 1.7V, Vf for green and yellow LEDs is about 2.1V.  Note: VIH drops as
Vdd drops - that is why this measurement is possible.

I'll be the first to admit that I haven't tried my suggestion yet but it
*should* work.

dwayne

--
Dwayne Reid   <RemoveMEdwaynerspamTakeThisOuTplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 18 years of Engineering Innovation (1984 - 2002)
 .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
    `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestEraseMEspam.....mitvma.mit.edu


2002\07\25@175820 by Dave Dilatush

picon face
Dwayne Reid wrote...

>VIH is about 1.4V for a non-ST input when Vdd=5V.  Vf for a red LED is
>about 1.7V, Vf for green and yellow LEDs is about 2.1V.  Note: VIH drops as
>Vdd drops - that is why this measurement is possible.
>
>I'll be the first to admit that I haven't tried my suggestion yet but it
>*should* work.

I've no doubt it might work for a one-off, or for a short run of
units.  But for any more than that, I wouldn't want to rely on
it: I've had too much experience over the years with designs that
developed production troubles because the original designer
counted on parts performing better than specified.

What it comes down to is, do you trust that this 1.4V threshold
(which you measured on a large number of units, I assume) is
going to hold from unit to unit, and from production lot to
production lot, over a long period of time?  Or do you use the
PIC specifications as your guide?

That's what I prefer to do; and the specs say that at Vdd = 4V,
which is the voltage level Kieren wants to detect, the PIC's TTL
input threshold voltage could be anywhere within a 3:1 range,
from 0.6V to 1.8V.  That's a pretty wide spread.

DD

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspammitvma.mit.edu


2002\07\25@191317 by Dwayne Reid

flavicon
face
At 09:50 PM 7/25/02 +0000, Dave Dilatush wrote:

>What it comes down to is, do you trust that this 1.4V threshold
>(which you measured on a large number of units, I assume) is
>going to hold from unit to unit, and from production lot to
>production lot, over a long period of time?  Or do you use the
>PIC specifications as your guide?

The threshold voltage seems to be reasonably repeatable.  In fact, the
RCtime type measurements that are used to such great effect in things like
the Basic Stamps and in many other similar applications rely on that
threshold voltage being repeatable.

I'm not sure how many units the original poster was planning to build.  If
it was me and the production was going to be thousands or tens of
thousands, I'd darn well characterize it to *my* satisfaction before
committing to it.

I use RCtime type measurements in several of my products.  One product in
particular uses it to set one of the calibration parameters and is thus
tested on each and every unit shipped.  The threshold has been absolutely
repeatable - this is with the 12c508A and 16C73B PICs.  I won't hazard a
guess about other PICs.

The original poster was asking about a low cost method of determining the
remaining battery life.  I would classify this as a low accuracy
requirement and I think that the RCtime type of measurement is probably
adequate for the purpose.  But only the person doing that project can
determine if it in fact meets his requirements.

Same goes for using a LED as a reference.  I've had great results in low
accuracy applications (mostly for setting the bias voltage in a 1
transistor constant current supply).  In that application, it has BETTER
temperature stability than the classic 2 transistor circuit.  But only the
person doing the project can determine if it is accurate enough for HIS
application.

dwayne



--
Dwayne Reid   <RemoveMEdwaynerEraseMEspamEraseMEplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 18 years of Engineering Innovation (1984 - 2002)
 .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
    `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspam_OUTspamKILLspammitvma.mit.edu


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...