No current draw, bat low voltage detector

Jon Jenkins

unread,

Feb 13, 1996, 3:00:00 AM2/13/96

to

Hi, we have one of those remote elctronic
door chimes in our house. Problem is the
batteries in the receiver always go flat
"silently". We never know until someone
banging at the door tells us that our
doorbell doesn't work.

What I need is a low voltage detector. I have
come up with a few ideas with zeners/dividers/
darlingtons and a 555 i.e. the zener holds a
darlington pair off which in turn holds the 555
off etc but they all draw some quiescent current.
Can anyone think of a design which draws only
miniscule current while the voltage is above
some threshold value and then intermittantly
"chirp" as the voltage goes low ??

Thanks

Jon Jenkins

--
----------------------------------------------------------------------
Name: Dr Jon Jenkins Location: Digital Equipment Corporation NaC
Voice/Fax: 61-7-55-75-0151/100 Burnett Place, Research Park,
Inet: jenk...@ozy.dec.com Bond University, Gold Coast
Close Proximity: "HEY YOU !!!" QLD, AUSTRALIA 4229
"Daddy, what's outside the Universe?" (My 5 year old.....)
-----------------------------------------------------------------------

Boris Mohar

unread,

Feb 15, 1996, 3:00:00 AM2/15/96

to

On 13 Feb 1996 20:24:47 GMT, jenk...@flotsm.ozy.dec.com (Jon Jenkins)
wrote:

>Hi, we have one of those remote elctronic
>door chimes in our house. Problem is the
>batteries in the receiver always go flat
>"silently". We never know until someone
>banging at the door tells us that our
>doorbell doesn't work.
>
>What I need is a low voltage detector.

t.
>Can anyone think of a design which draws only
>miniscule current while the voltage is above
>some threshold value and then intermittantly
>"chirp" as the voltage goes low ??
>

Smoke detector chips have low bat chirp function.

Boris

John Lundgren

unread,

Feb 17, 1996, 3:00:00 AM2/17/96

to

Jon Jenkins (jenk...@flotsm.ozy.dec.com) penned:
: Hi, we have one of those remote elctronic

: door chimes in our house. Problem is the
: batteries in the receiver always go flat
: "silently". We never know until someone
: banging at the door tells us that our
: doorbell doesn't work.

: What I need is a low voltage detector. I have

: come up with a few ideas with zeners/dividers/
: darlingtons and a 555 i.e. the zener holds a
: darlington pair off which in turn holds the 555
: off etc but they all draw some quiescent current.

: Can anyone think of a design which draws only

: miniscule current while the voltage is above
: some threshold value and then intermittantly
: "chirp" as the voltage goes low ??

I played around with some low voltage detector circuits a few years ago,
and came to the conclusion that they were easy to make but not accurate
enough. The ones I used depended on the E-B voltage of .6 V to
determine the low voltage point. This voltage is dependent on the
temperaturs, and changes substantially. So it's not a good circuit. The
ones with an accurate voltage reference cost a lot more, just for the
2.5V reference IC, so I didn't want to use one of those. And they're
hard to find: not something that you can go to Rat Shack for (they
did have them at one time many years ago).

I thought about it, and the simple circuit that is used for detecting the
off-hook of a telephone line would work great. It doesn't draw any
current when the voltage is above a certtain point. And when it goes
below that point, it lights a LED or whatever. This is the circuit from
an original post from sci.electronics, with my comments at the end.

Just eliminate the lower 2.2M resistor and ground that lead. Change the
upper 2.2M resistor to that needed to get the circuit to go on at 7 or
8V, something like 470K or so. The problem is that the 3 VDC has to be
used for the reference voltage, so you will have to come up with some
method of deriving that from the battery, without drawing substantial
current. Good luck.

I'll look around for another similar circuit.

-------
Newsgroups: sci.electronics
Subject: phone use detect
From: aurel.b...@bville.gts.org (Aurel Boisvert)
Path:
ohlone.kn.PacBell.COM!pacbell.com!well!barrnet.net!agate!darkstar.UCSC.EDU!news.hal.COM!olivea!charnel.ecst.csuchico.edu!yeshua.marcam.com!MathWorks.Com!europa.eng.gtefsd.com!howland.reston.ans.net!cs.utexas.edu!utnut!torn!uunet.ca!uunet.ca!io.org!bville!aurel.boisvert
Distribution: world
Message-ID: <1b.11126.38...@bville.gts.org>
References: <lippman.775463977@scooter>
Date: Tue, 2 Aug 94 07:15:00 -0500
Organization: Baudeville BBS: Toronto CANADA 416.283.0114
Lines: 28

Hi Louis.
Here is a simple circuit I use for that purpose.
My computer is in the basement and this device tells me if the phone
line is in use. I have inserted a N/O switch in the battery connection
so that the batteries will last longer as sometime my sons spend a lot
of time on the phone. Prior to using my modem I press the switch to find
out if the line is busy.

+ = connection
2N3906 33K 2N3904

2.2 meg /----------/\/\/\--+ /-----+
|/ | |/ \
Tip ---/\/\/\----+-------| PNP +----| NPN / 220 **
\ |\E 2N3904 |\E \
/ 330K \ \ /
\ | +-------------+ \
| | | | |
Ring--/\/\/\-----+ | v - | ^ LED
2.2 meg | | 3V | |
| | ^ + | |
| | | | |
+-------------------------------------+ |
| | |
+------------+-------------------+

** This resistor may have to be lowered to match the led used.
Use alkaline battery, they last longer.

--
End of Original Msg. Comments by jlun...@kn.pacbell.com

The above circuit really does work, and it's about the simplest
and most effective one around. It runs for a long time off
two C or D cells, and doesn't connect to any power, so it won't
burn down the house or get Ma Bell mad at you. I've seen other
circuits that are more complex and expensive, and use chips and other
stuff like optoisolators. They draw more power from the power supply
and the phone line. And they don't do the job any better. KISS and
use this circuit.

Corrections, additions, and nitpicks:

Correction:
The Tip and Ring are reversed. The tip is negative, so it must be
connected to the resistor that is connected to the minus 3V. The
ring should go to the base of the 2N3906. But whichever way your
phone line is, connecting it one way will work, and the other way
will not.

Addition:
When the phone line is ringing, there is a lot more than 3 volts
across the 330K resistor. The voltage may reach 7 to 10 volts peak.
Since a reverse biased emitter to base junction may avalanche at a
voltage lower than this, the circuit should have something to limit
the voltage. A 1N4148 diode should be put across the E-B junction
of the 2N3906, cathode to the emitter so that is reverse biased in
normal operation. So then if the positive voltage gets higher than
.6 volts more than the +3Volts, it will conduct and clamp the voltage.
This will also help protect against transients. Another 1N4148
should be put across the 330K resistor, cathode to base, so it is
reverse biased during normal operation but will protect the base in
case the tip lead goes in the negative direction.

Optional addition:
There is no R.F. interference suppression in this circuit. It should
have a capacitor across the 330K resistor to bypass any stray RF to
ground and prevent any false lighting of the LED. I would guess an
adequate value would be 1000pF (or .001uF) at 50 volts or more,
ceramic capacitor. This is in case there is a ham or CBer in the
neighborhood running high power.

Another addition:
A blinking LED would get the user's attention a lot better, and it
could cut the drain on the battery substantially. Most of the
current used by this circuit is the LED current. If the LED is made
to blink and draw current for half the time, then it would cut the
total battery current about in half, and extend the battery life to
double. If the LED blinked at less than 50% of the time, it would
be even less current drain. Addendum: I bought a red blinking LED,
and put it in. With the DVM on the current range and a 6800 uF cap
across the 3 volt leads to the flasher, it looks like the current is
well under 1 mA. This should make a set of batteries last a long, long
time. (The DVM still jumps around alot, tho.)

Circuit Operation:
With all phones on the hook, and 48VDC at the phone line input, the
330K resistor has about 3.3 VDC across it. Therefore the base to
emitter junction of the 2N3906 has a slight reverse bias on it, since
the emitter is at +3VDC. So neither transistor conducts current,
and the whole circuit draws very little current from the battery
when all phones are on hook.

As the voltage drops below 48V when any phone is off-hook, the
voltage across the 330K drops, and the emitter to base junction of
the 2N3906 becomes forward biased, conducting current. The amplified
but still small current from the collector goes through the 33K
resistor to the 2N3904, and it, too, conducts enough current to light
the LED.

The resistors across the phone line add up to 4.73 megohms, so at
48V there are about 10 microamps drawn from the phone line, when
all phones are on hook. There is no current flowing in the 2N3906
because it is reverse biased, and also no current flows in the 2N3904.

The phone line voltage at which the LED starts to light is somewhere
below 40VDC. This voltage will decrease somewhat as the batteries
discharge with age. The current drawn when any phone is off-hook is
much less than 10 microamps, and is dependent on the voltage across
the phone line.

When the phone line is ringing, the line voltage is over 100 volts
AC, at around 20 Hz. There is a lot more than 3 volts across the
330K resistor. That's a good argument for some form of voltage
limiting here.

I built the circuit with the resistor values as shown, and it does
indeed work. When the phone is taken off hook, the LED lights up.
I used a yellow LED, with a 300 ohm resistor in series with it
instead of 220 ohms. The total current drawn by the circuit with
the LED lit is less than 4 mA. Using a pair of C or D cells, the
circuit should operate for a long time at this current level. At
this low a current, it doesn't pay to buy alkalines, when
zinc/carbon ones will work about as long and are considerably cheaper.
Addendum: as mentioned above, I replaced the LED with a blinking LED
(available at Radio Snack), and changed the resistor in series to
270 ohms. It draws less than 1 mA from the battery, which will give
even longer life to the batteries, even for long-winded habitual
phone abusers. Actually, I qualify for that, except that I do the
abuse with a modem. ;-)

Just remember that people will still pick up the telephone without
looking at the light, just from force of habit. So don't expect this
circuit to work 100% of the time. It just won't, I know from experience.

: Thanks
: Jon Jenkins

: --

Warren Block

unread,

Feb 18, 1996, 3:00:00 AM2/18/96

to

Boris Mohar (bor...@interlog.com) wrote:

: Hi, we have one of those remote elctronic door chimes in our house.
: Problem is the batteries in the receiver always go flat "silently". We
: never know until someone banging at the door tells us that our doorbell
: doesn't work.

: What I need is a low voltage detector.

: Can anyone think of a design which draws only miniscule current while

: the voltage is above some threshold value and then intermittantly "chirp"
: as the voltage goes low ??

In the new issue of Circuit Cellar Ink (#67), on page 9, there is a kind
of neat new product announcement: an LED and low-voltage sensor combined
in an LED package. The note says "Although designed to detect low power
in two 1.5-V AA batteries, the LEDs also accomodate packs of 3-6 1.5-V
cells by using external resistors."

Standby current is given as 5 uA (that's microamps). Cost is $1.50 for
singles. The address given is:

Lumex Opto/Components
290 E. Hellen Rd.
Palatine IL 60067
(847) 359-2790
(847) 359-2867 Fax

--
.----------------------------------------------------------------.
| Warren R. Block * New EMail Address: wbl...@rapidnet.com |
| Rapid City SD USA * Brought to you in majestic INEMASCOP! |
`----------------------------------------------------------------'

Sam Goldwasser

unread,

Feb 19, 1996, 3:00:00 AM2/19/96

to

In article <31233f66...@news.interlog.com> bor...@interlog.com (Boris Mohar) writes:

> Smoke detector chips have low bat chirp function.

> Boris

Excellent idea. And since 'they' recommend replacing your smoke
detectors every 10 years, you should have a free source of chips
which will work just fine. The chirp should be adequate to light
an LED even if it does not have a flashing LED capabiilty built in.

--- sam

John Lundgren

unread,

Feb 20, 1996, 3:00:00 AM2/20/96

to

Sam Goldwasser (s...@stdavids.picker.com) penned:

: > Boris

: --- sam

The crazy smoke det in my place already blinks the LED. I think it's
meant to be used for a test of whether the thing is working or not. Every
minute or so it flashes. This goes on continuousley as long as the
battery is OK. If the battery gets low, then it chirps. CHIRP! Loud
sucker! So it draws enough power to blink an LED at long intervals. It
started chipring late last year, and I changed the battery. I guess I
was getting a little slack. I mark on the battery the installation date,
and this one had been installed nearly five years ago(!)

If the smoke det is self checking, it may not be easy to disable or
remove the smoke det sensor or fiddle too much with the circuit.

| Most FAQs are available through Thomas Fine's WWW FAQ archive: |
|http://www.cis.ohio-state.edu:80/hypertext/faq/usenet/FAQ-List.html|

Paul Grohe

unread,

Feb 22, 1996, 3:00:00 AM2/22/96

to

On 20 Feb 1996 13:19:01 GMT,
in sci.electronics.misc
jlun...@delta1.deltanet.com (John Lundgren)
from Delta Internet Services, Anaheim, CA thoughtfully posted:

> Sam Goldwasser (s...@stdavids.picker.com) penned:
> : In article <31233f66...@news.interlog.com> bor...@interlog.com (Boris Mohar) writes:

> : > Smoke detector chips have low bat chirp function.

Actually, smoke detectors achieve long battery life by turning
themselves on at short, regularly spaced intervals. They are not on
all the time, they just seem like it.

After a smoke detector in my house decided to dine on new 9V batteries
every 3 weeks, I *had* to investigate!

The good smoke detectors turned themselves on for about 1/10th of a
second every 5 or so seconds. They wake up, take a reading, then go
back to sleep until the next time.

During this "sleep" time, they drew about 4uA, and during the
"measuring" time they would draw about 50uA. The lamp would flash
every minute or so.

The "bad" smoke detector drew 1mA continuously. So, back to the store
it went!

If you could figure a way to add a similar duty-cycle to your battery
monitor, you could save a little of the batteries life.

Cheers,
Paul Grohe
---------------------------------------------------------------
Paul Grohe National Semiconductor Corp.
Sr. Electronics Technician 2900 Semiconductor Drive
AMPS New Products Eng Group Mail Stop C2693
Email: gr...@galaxy.nsc.com Santa Clara, CA. 95052-8090 USA
(408) 721-7389 Tel (408) 721-2513 Fax

For technical assistance, literature, or samples call
Canada & US: (800) 272-9959 Europe: (49) 814 110-3720
Email Support: sup...@tevm2.nsc.com
NSC datasheets online at: http://www.natsemi.com
Usenet Disclaimer: Any opinions expressed are mine, not NSC's
---------------------------------------------------------------

Jon Jenkins

unread,

Feb 24, 1996, 3:00:00 AM2/24/96

to

Many thanksa to all who replied. I will see if I can get a
sample of the custome chips or build some of the
circuits myself. Which ever works best/cheapest
I will report back here somewhere.

Thanks again for all the ideas, Jon