I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

I used two caps in case one shorts and for some ground isolation. The diodes
are 5.1 V zeners ($1.29 for 2 at Radio Shack) to limit the clock voltage if
someone unplugs the batteries. This circuit supplies 390 microamps. I hope
to avoid replacing the batteries until their shelf life runs out in 2012 :-)

This could be useful for lots of low-power stuff, eg smoke detectors.

Nick

wrote:
I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump

We used to use a similar circuit for powering electronic thermostats
when I worked for Potterton Boilers here in Blighty. We used just one
Class X capacitor feeding a 24V zener and smoothing capacitor.

I also used to use one for charging the NiCads in my caving lamp. Of
course I made the classic mistake of plugging it in then picking up the
crocodile clips to connect onto the battery (rather than connecting
onto the battery then plugging in). Rectified 240V doesn't half hurt.

--
Malc

I used a ac adapter of the proper voltage to charge a bunch of D sixed
nicads to run my answering machine during power failures. it was like a
UPS uninterruptiple power supply.

my old machine would spaz with any power failure

The batteries went bad after 6 years, my new machine a different
voltage with a battery backup of memory, so I dont use this anymore

I hope you are using rechargable batteries!

Regulart alkaline or others may dry out, overheat and start a fire!

USE ONLY RECHARGABLE BATTERIES IN THIS APPLICATION!

wrote:

I hope you are using rechargable batteries!

NO!

Regulart alkaline or others may dry out, overheat and start a fire!

How much will they overheat if overchaged at 3Vx50uA = 150 microwatts? :-)

USE ONLY RECHARGABLE BATTERIES IN THIS APPLICATION!

NO!

Happy new year,

Nick

I like the concept, but it's hardly worth doing if you value the
portability of the object and batteries can be had so cheaply.

What is your child's life worth when they need to steal
a battery for their latest toy?

That's just idiotic.

"Al Bundy" wrote in message
oups.com...
I like the concept, but it's hardly worth doing if
you value the
portability of the object and batteries can be had so
cheaply.

Rich Komp replies:

Nick,
You are right. I have used non-rechargable batteries for years with
solar battery chargers. The current is controlled by the number of
photons going into the PV cells, not the internal resistance of the
batteries. I once deliberately grossly overcharged a regular alkaline
battery in a solar charger and got it to explode like a firecracker, but
it took almost an amp for a single AA cell, not the microamps we are
discussing. I will admit that after 5 years or so in a solar clock, some
of the batteries have leaked, but the battery contacts, weren't too hard
to clean (and I've never seen a fire from this).
I like breaking rules, especially when they are so FORCEFUL.
Rich

wrote:

I hope you are using rechargable batteries!

NO!

Regulart alkaline or others may dry out, overheat and start a
fire!

How much will they overheat if overchaged at 3Vx50uA = 150
microwatts? :-)

USE ONLY RECHARGABLE BATTERIES IN THIS APPLICATION!

NO!

Happy new year,

Nick

wrote:

I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

I used two caps in case one shorts and for some ground isolation. The diodes
are 5.1 V zeners ($1.29 for 2 at Radio Shack) to limit the clock voltage if
someone unplugs the batteries. This circuit supplies 390 microamps. I hope
to avoid replacing the batteries until their shelf life runs out in 2012 :-)

This could be useful for lots of low-power stuff, eg smoke detectors.

Nick

Caps do have a failure mode, so a series safety R is normally included
with these type circuits.

If your ac supply is polarised, you'd be a bit safer putting one cap in
the live than one in each pole, as the output will then be at apx earth
potential, though not isolated. As it stands its semi-live.


NT

Looks like a shock hazard to me.



wrote in message
oups.com...
wrote:

I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

I used two caps in case one shorts and for some ground isolation. The
diodes
are 5.1 V zeners ($1.29 for 2 at Radio Shack) to limit the clock voltage
if
someone unplugs the batteries. This circuit supplies 390 microamps. I
hope
to avoid replacing the batteries until their shelf life runs out in 2012
:-)

This could be useful for lots of low-power stuff, eg smoke detectors.

Nick

Caps do have a failure mode, so a series safety R is normally included
with these type circuits.

If your ac supply is polarised, you'd be a bit safer putting one cap in
the live than one in each pole, as the output will then be at apx earth
potential, though not isolated. As it stands its semi-live.


NT

Clark wrote:
Looks like a shock hazard to me.



wrote in message
oups.com...

wrote:


I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

I used two caps in case one shorts and for some ground isolation. The
diodes
are 5.1 V zeners ($1.29 for 2 at Radio Shack) to limit the clock voltage
if
someone unplugs the batteries. This circuit supplies 390 microamps. I
hope
to avoid replacing the batteries until their shelf life runs out in 2012
:-)

This could be useful for lots of low-power stuff, eg smoke detectors.

Nick

Caps do have a failure mode, so a series safety R is normally included
with these type circuits.

And at that low a current, why not ferget the caps and just use a couple
of 150K resistors? The additional power loss will hardly spin your
meter off the wall and you'll avoid the possible cap failure problem.

Jeff (Ducking and slinking off...)

--
Jeffry Wisnia

(W1BSV + Brass Rat '57 EE)

"Truth exists; only falsehood has to be invented."


If your ac supply is polarised, you'd be a bit safer putting one cap in
the live than one in each pole, as the output will then be at apx earth
potential, though not isolated. As it stands its semi-live.


NT

I...built a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

Nick (nicksanspam @ ece.villanova.edu)

Caps do have a failure mode,
so a series safety R is normally included with these type circuits.
NT (meow2222 @ care2.com)

It should be noted that such *series capacitor* devices
were outlawed for commercial products long ago in Europe.
As the Big Cat notes, without a resistor to limit things,
failures are usually dramatic.

2 Lithium batteries would be safer, cheaper, less work,
last longer and much more reliable.

Just a moronic idea

"JeffM" wrote in message
oups.com...
I...built a charge pump like this, viewed in a
fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to
clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

Nick (nicksanspam @ ece.villanova.edu)

Caps do have a failure mode,
so a series safety R is normally included with these
type circuits.
NT (meow2222 @ care2.com)

It should be noted that such *series capacitor*
devices
were outlawed for commercial products long ago in
Europe.
As the Big Cat notes, without a resistor to limit
things,
failures are usually dramatic.

wrote in message ...
I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

I used two caps in case one shorts and for some ground isolation. The diodes
are 5.1 V zeners ($1.29 for 2 at Radio Shack) to limit the clock voltage if
someone unplugs the batteries. This circuit supplies 390 microamps. I hope
to avoid replacing the batteries until their shelf life runs out in 2012 :-)

This could be useful for lots of low-power stuff, eg smoke detectors.

Nick

It seems counter-productive to build booby-traps into smoke detectors
and the like.

Unlike a wall-wart, it will require special precautions not to accidentally
touch anything when you go to change the battery or maybe even set the time
on a clock. You will never know whether a capacitor has shorted and whether
you might be killed the next time you provide a path to earth ground.

There's nothing to limit the current when capacitors fail, so that failure
will cause a fire. There's no transient voltage protection, so the odds
of capacitor over-voltage failure is pretty high.

Don

Don K wrote:

wrote:

0.047 uF @400 V
1K || |
----www----||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

Unlike a wall-wart, it will require special precautions not to accidentally
touch anything when you go to change the battery...

And it has battery backup :-)

There's nothing to limit the current when capacitors fail, so that failure
will cause a fire.

I doubt that. The diodes will probably open, and I used really skinny wire.

There's no transient voltage protection, so the odds of capacitor over-voltage
failure is pretty high.

Or diode damage. So... I added a 1K series resistor. This would also work:

0.022 uF @400 V*
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
| |
10K | |
-----------www----------------------------------

*Digikey's $29.95 M400-KIT-ND 400 V capacitor assortment contains
10 of these, along with 10 of 14 other values from 0.001 to 0.1 uF.

Nick

wrote:
Don K wrote:
wrote:

0.047 uF @400 V
1K || |
----www----||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

Unlike a wall-wart, it will require special precautions not to accidentally
touch anything when you go to change the battery...

same precautions as any mains device, all nonisolated parts must be
protected from touch. The danger with this system lies not in the
design itself, but in putting into the hands of people that dont know
what theyre doing with it, and might use it like a wall wart. Since the
safety R has been added, it can be used perfectly safely, but only when
its issues are understood and appropriately addressed.

Smoke detectors could be used hapily with it if a screw were added to
prevent opening.


There's nothing to limit the current when capacitors fail, so that failure
will cause a fire.

X rated capacitors use fusible coatings, so a short causes a burn out
of the conductor around the short. The overall capacitance is not
noticeably affectde, IOW its self healing.


I doubt that. The diodes will probably open, and I used really skinny wire.

Very fine wire sounds like effective fusing to me.


There's no transient voltage protection, so the odds of capacitor over-voltage
failure is pretty high.

Or diode damage. So... I added a 1K series resistor. This would also work:

Yes, though again you need the R and C in the live. The days where you
could scatter proetction devices between L and N lines is long gone -
at least in our part of the world anyway.

If youre using non polarised plugs on this, you'd need an R in both
lines imho. These need to be safety Rs, not the kind that catch fire
and carbonise.

Personally I'd make the R as high as is possible, so it responds
effectively to any fault throughput incease.


NT

wrote in message oups.com...
wrote:
Don K wrote:
wrote:

There's nothing to limit the current when capacitors fail, so that failure
will cause a fire.

X rated capacitors use fusible coatings, so a short causes a burn out
of the conductor around the short. The overall capacitance is not
noticeably affectde, IOW its self healing.


I doubt that. The diodes will probably open, and I used really skinny wire.

Very fine wire sounds like effective fusing to me.


You only need 20mA of AC current to cause ventricular fibrillation.
If the capacitor fails, someone could get killed touching whatever
your circuit connects to.

Don

Don K wrote:
wrote in message oups.com...
wrote:
Don K wrote:
wrote:

There's nothing to limit the current when capacitors fail, so that failure
will cause a fire.

X rated capacitors use fusible coatings, so a short causes a burn out
of the conductor around the short. The overall capacitance is not
noticeably affectde, IOW its self healing.


I doubt that. The diodes will probably open, and I used really skinny wire.

Very fine wire sounds like effective fusing to me.


You only need 20mA of AC current to cause ventricular fibrillation.
If the capacitor fails, someone could get killed touching whatever
your circuit connects to.

Don

It would be a foolish designer that enabled users to touch the output
of one of these, so thats a non issue in a competent design. If for
some reason they did, Y rated caps would be more appropriate than X -
but still, liability like that isnt something I'd want to take on
unnecessarily.

As far as fusing, if a cap shorts, enough i will flow to pop a fine
wire, assuming it really is fine. A single strand of a thin stranded
equipment wire etc. So the fuse concept is good.

I suppose if one wanted to improve the safety of an unsafe device (ie
nonisolated supply with outputs touchable) one could put a cap in both
poles and earth the supply output. Would need to ensure R is
significant to avoid RCD problems.

Note this would work as a 1 cap supply, as the E connection would
effectively short out the neutral capacitor. Ick.


NT

Don K wrote:
wrote in message oups.com...
wrote:
Don K wrote:
wrote:

There's nothing to limit the current when capacitors fail, so that failure
will cause a fire.

X rated capacitors use fusible coatings, so a short causes a burn out
of the conductor around the short. The overall capacitance is not
noticeably affectde, IOW its self healing.


I doubt that. The diodes will probably open, and I used really skinny wire.

Very fine wire sounds like effective fusing to me.


You only need 20mA of AC current to cause ventricular fibrillation.
If the capacitor fails, someone could get killed touching whatever
your circuit connects to.

Don

It would be a foolish designer that enabled users to touch the output
of one of these, so thats a non issue in a competent design. If for
some reason they did, Y rated caps would be more appropriate than X -
but still, liability like that isnt something I'd want to take on
unnecessarily.

As far as fusing, if a cap shorts, enough i will flow to pop a fine
wire, assuming it really is fine. A single strand of a thin stranded
equipment wire etc. So the fuse concept is good.

I suppose if one wanted to improve the safety of an unsafe device (ie
nonisolated supply with outputs touchable) one could put a cap in both
poles and earth the supply output. Would need to ensure R is
significant to avoid RCD problems.


NT

0.022 uF @400 V*
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
| |
10K | |
-----------www----------------------------------

*Digikey's $29.95 M400-KIT-ND 400 V capacitor assortment contains
10 of these, along with 10 of 14 other values from 0.001 to 0.1 uF.

The rest of this stuff is cheap...

$13.65 for 100 1N5231BD1CTND 5.1V zeners,
$15.86 for 1000 P10KBACTND 10K resistors,
and a line cord cut off an old appliance.

Anyone else want to burn their house down? :-)

Nick

nick pine wrote:

0.022 uF @400 V*
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
| |
10K | |
-----------www----------------------------------

*Digikey's $29.95 M400-KIT-ND 400 V capacitor assortment contains
10 of these, along with 10 of 14 other values from 0.001 to 0.1 uF.

The rest of this trickle charger is cheap...

$13.65 for 100 1N5231BD1CTND 5.1V zeners,
$15.86 for 1000 P10KBACTND 10K resistors,
and a line cord cut off an old appliance.

My smoke detector draws 7 uA = 60x370C, which makes C = 315 puffs, eg 2
0.001 uF 400 V caps in series, Digikey's P1058-ND, $11.43/100. They also
sell 12 V zeners suitable for 9 V batteries, 1N5242BDICT-ND, $13.65/100.

Nick

wrote in message
...
nick pine wrote:

0.022 uF @400 V*
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
| |
10K | |
-----------www----------------------------------

*Digikey's $29.95 M400-KIT-ND 400 V capacitor assortment contains
10 of these, along with 10 of 14 other values from 0.001 to 0.1 uF.

The rest of this trickle charger is cheap...

$13.65 for 100 1N5231BD1CTND 5.1V zeners,
$15.86 for 1000 P10KBACTND 10K resistors,
and a line cord cut off an old appliance.

My smoke detector draws 7 uA = 60x370C, which makes C = 315 puffs, eg 2
0.001 uF 400 V caps in series, Digikey's P1058-ND, $11.43/100. They also
sell 12 V zeners suitable for 9 V batteries, 1N5242BDICT-ND, $13.65/100.

Nick


Of course you don't have enough current for the audible device so your smoke
detector should sit quietly by while you toast!

wrote:

My smoke detector draws 7 uA = 60x370C, which makes C = 315 puffs, eg 2
0.001 uF 400 V caps in series, Digikey's P1058-ND, $11.43/100. They also
sell 12 V zeners suitable for 9 V batteries, 1N5242BDICT-ND, $13.65/100.

Nick

You have to supply bleeping power, relying on the battery would be
unsatisfactory. In which case it'll draw bleeping power al the time,
but since its almost entirely capacitive, and a house's total load is
normally lagging, this will in fact reduce overall current consumption.
A tiny tiny bit.

NT

wrote:
I finally got tired of replacing the 2 C-cells in a clock with a moving
pendulum, so I measured the current (3 volts at 340 microamps) and built
a charge pump like this, viewed in a fixed font:

0.047 uF @400 V
|| |
-----------||---------------------|------------
|| | | |
| | 3V
120 VAC --- --- to clock
^ _
0.047 uF @400 V | |
|| | |
-----------||-----------------------------------
||

I used two caps in case one shorts and for some ground isolation. The diodes
are 5.1 V zeners ($1.29 for 2 at Radio Shack) to limit the clock voltage if
someone unplugs the batteries. This circuit supplies 390 microamps. I hope
to avoid replacing the batteries until their shelf life runs out in 2012 :-)

This could be useful for lots of low-power stuff, eg smoke detectors.

Nick



Why not a small dc-dc converter,from your main 12/24/48V battery bank..
Safer,and probably more efficient. (takes the inverter out of the
loop,atleast.)

phatty mo wrote:

Why not a small dc-dc converter,from your main 12/24/48V battery bank..

I don't have a battery bank.

Safer,and probably more efficient. (takes the inverter out of the
loop,atleast.)

If your main inverter doesn't run 100% of the time, you might make the cap
larger to charge the battery more when the inverter does run. This circuit
itself is unlikely to make the inverter come out of hibernation.

Nick

****. get a small flexible PV panel and use the damn
thing.

wrote in message
...
phatty mo wrote:

Why not a small dc-dc converter,from your main
12/24/48V battery bank..

I don't have a battery bank.

Safer,and probably more efficient. (takes the
inverter out of the
loop,atleast.)

If your main inverter doesn't run 100% of the time,
you might make the cap
larger to charge the battery more when the inverter
does run. This circuit
itself is unlikely to make the inverter come out of
hibernation.

Nick

SolarFlare wrote:
get a small flexible PV panel

glue it to the face of the clock
add a series resistor and schottky diode to charge the
battery.

the frugal will steal the parts from a defunct solar
calculator ;-)

Nick, put equal value cap and resistor in series with each
line and your diodes in a small box with ac plug (gutted
wall wart- there's justice?) then run small cable (like the
old earphone cable) up to clock. Paint red insulator over
any "hot" parts in clock. The "power" cable will be about
as dangerous as any current small appliance "leakage" since
most have .045 mF caps from EACH power lead to chassis. UL
specifies an allowable leakage level. 5 ma sounds familiar
since that's the level that GFI's are supposed to trip at.

There's a thought, wire it any way you want to, then plug it
into a gfi ;-) hook all your clocks, smoke detectors, etc
together and use just one gfi. the gfi will use more power
than you're saving.

have a good new year -larry / dallas

Get an old coil for induction and put a series diode to
the battery. Hang it someplace over wiring with a good
load feed.

voila...self contained clock with no dangerous
connection to high voltage.


"larry" wrote in message
et...
SolarFlare wrote:
get a small flexible PV panel

glue it to the face of the clock
add a series resistor and schottky diode to charge
the
battery.

the frugal will steal the parts from a defunct solar
calculator ;-)

Nick, put equal value cap and resistor in series with
each
line and your diodes in a small box with ac plug
(gutted
wall wart- there's justice?) then run small cable
(like the
old earphone cable) up to clock. Paint red insulator
over
any "hot" parts in clock. The "power" cable will be
about
as dangerous as any current small appliance "leakage"
since
most have .045 mF caps from EACH power lead to
chassis. UL
specifies an allowable leakage level. 5 ma sounds
familiar
since that's the level that GFI's are supposed to
trip at.

There's a thought, wire it any way you want to, then
plug it
into a gfi ;-) hook all your clocks, smoke
detectors, etc
together and use just one gfi. the gfi will use more
power
than you're saving.

have a good new year -larry / dallas