Current sensor

[mmcnees]

I would like to know where I can find schematics to build a very cheap current sensor.
The reason it needs to be cheap is because I'll probably only use it once.
What I'm tring to do is, monitor my electric hot water heater(thanks to high energy cost). I'm trying to figure out how long it is running per day, to see if a water heater timer would be worth paying $40.00 for.
I would like to have a sensor that I could hook up to a relay that drives a small battery clock(non digital, the type you get at walmart that takes one AA battery..the two dollar type).
Or something like that. If there is a better way thats ok also.
I have found some by searching the internet but the ones I have found are all basicly outdated. Like you can't find the ICs anymore or you have to buy 50 of them.
If anyone has any ideas for something I could build from parts at Radio Shack please let me know.

Thanks
Mike

[jimandy]

I'm not going to take credit for this although I would like to, but in an earlier thread involving hot water heaters, stephen posted this. (You can search the forum using the phrase "hot water heater".

stephen

I would consider using a current transformer with an indicator light. A pilot lamp, such as the #47, is rated at 6 volts and 150 milliamps. With a 1:100 ratio current transformer a primary current of 15 amps would light the bulb to full brilliance. The primary voltage drop would be about 60 millivolts.

A common method of construction for a current transformer is to wind the secondary turns on a toroid form, usually a tape wound core. A large hole is left in the center, and the primary consists of a single turn of wire passing through the center. This has the advantage that the insulation does not have to be removed from the primary wire.

The core does not have to be a toroid. An E-I core could also be used with one turn of wire looped around the center leg as the primary.

A current transformer should not be operated with an open circuit in the secondary circuit. The voltage can easily rise to levels sufficient to break down the transformer insulation. Since incandescent lamps tend to fail as an open circuit, I would place a pair of zener diodes in a back-to-back connection across the indicator lamp. This would limit the voltage if the lamp burns out.

[Chris Smith]

Another cheap and simple way is to form a piece of 10/12 gage solid copper feed wire [un-insulated romex wire] around a pencil or a dowel, about 8 or 10 turns of which ever size wire feeds the circuit already.

Make the coils have a slight spacing so no two coils touch each other

Then glue a magnetic reed switch [usually made of glass] on the out side of these coils parallel to these winding [axial] with silicon glue.

Then feed the power through this coil and onto the heater element.

When the power is in use to the element, the reed switch closes from the magnetic field produced by the heater consuming power and the electromagnetic field you have formed.

Then, this reed switch runs your low power circuit to the clock.

This should cost about a buck to build with no electronics involved, just wiring, soldering, and a activated switch circuit.

As long as you place the 120 volt exposed parts into a simple box to keep the voltage away from shorts and people, you should be safe.

A spare gang box and wall receptacle with an short extension cord would also allow you to make this device small, simple, and reusable for its parts or other sensing projects.

If your magnetic field strength is too low to pull the reed switch together, you can insert a insulated ferrite slug or core into the hallow.

My water pump uses this method as part of the start up circuit.

<small>[ February 20, 2006, 04:14 PM: Message edited by: Chris Smith ]</small>

[Bern]

What I think I would do, is bring out the wires from the two heating elements. Depending on where you are, these will probably be 220 V. Dig around and find a couple of relays that will work, either by themselves or a resistor in series if needed. Maybe use transformers, or what ever you can find in your “Goodie” box.

Connect to an old fashion mechanical clock, and you have it. Be very careful with this power. If you are not comfortable working around this kind of power, remember it is nonforgiving. Make sure the breakers are off when making any connections.

If you have two clocks, wire one to each element, with appropriate voltage in mind, and it should do your job.

A quick and dirty method, but for a one time, temporary application, should do the job.

[MrAl]

Hi there,

Actually, you don't need a current transformer,
all you need is a clock that runs on the same
voltage as the Elect Heater element. When
the element is energized, the clock runs.

If you have 220v and cant find a clock to run
on that then you can use a step down trans
or else use a diode and resistor to drop volts
and convert to dc (filter cap too) and drive
a small dc clock (like you talked about).

Either should be easy enough.


Take care,
Al

[Dean Huster]

Trouble is, MrAl, electric water heaters have two elements and two thermostats and whether the top element or the bottom element is on depends upon the internal water temperatures, so you won't get a true indicator of time. The top element is a SPDT thermostat that also determines whether or not the bottom element can operate as the top element takes precidence (that's the "quick recovery" part of an electric water heater). Your idea will easily work on a small single-element water heater, of course.

I've not worked with reed switches at all, but I was under the impression that they required either a permanent magnet or a DC coil and that an AC coil, because the reed switch has a pretty fast response time, would "chatter".

Dean

[rshayes]

A reed switch would almost certainly chatter. Tektronix used to use mercury wetted reed switches to generate fast rise time pulses. Some operated at 60 Hz, but I believe some were operated at about 1 KHz by a transistor inverter.

A shading coil of some type might reduce the tendency to chatter. It works on normal relays, but I don't know if it would be enough with a reed relay.

<small>[ March 04, 2006, 04:25 PM: Message edited by: stephen ]</small>

[Dean Huster]

I believe the shading coil's operation relies upon the inclusion of an iron core in a traditional electromagnetic relay and would only act as a shorted transformer secondary without a core. But then, I may be wrong about that. I guess that's why we experiment, isn't it?

I think that Tek's only pulse generator that used the mercury-wetted relay was the Type 109. After that, the Type 284 using tunnel diodes was the pulse generator workhorse for the company. Pulses generated by the 60Hz-driven 109 were a bear to see at high sweep rates without killing the room lighting and/or using a hood on the scope. I don't ever remember one operating at 1KHz.

Dean

[rshayes]

Hello Dean:

Several of the letter series plug ins also used mercury wetted relays.

Type P was used for calibrating the 540 series scopes. It used a mercury wetted relay driven at 60 HZ.

Type R was for measuring transistor rise time. This was also a mercury wetted relay at 60 Hz.

Type S was for measuring diode turn off time. It used a transistor multivibrator at about 250 Hz to drive a mercury wetted relay. This is where my memory slipped, it was 250 Hz rather than 1 KHz.

[Chris Smith]

If any chatter occurs with a reed switch, just saturate the field to stop the chatter.

My water pump doesn’t even do that with the controller.

Chatter only occurs if the magnetic hysteresis is too slow to hold the relay contacts closed [one direction] while a over powering strong magnetic field will simply pull it in one direction and time will do the rest.

If your field strength is too small for your relay switch, and chatter occurs, simply saturate it with a simple core of Ferrite, a simple metal core, or anything magnetic in nature.

This will delay the push pull [hysteresis] and it will simply be on.

[Dean Huster]

Subject #1: Yes, saturation could be done, and I thought that considering the present application target it would be a bit impractical/impossible to do lots of turns on the coil. However, considering the current through the "coil" with a heating element turned on, a single turn may set up a massive field if you add an iron core somewhere to saturate -- after all, you can't saturate the reed switch itself or the winding.

Subject #2: I forgot all about the letter-series calibration plug-ins, Stephen. There were some pretty obscure ones out there that weren't well-advertized as were the 7K and 5K plug-ins.

Dean

[Chris Smith]

Dean My water pump only uses around seven or eight windings next to the reed switch with no help from a core.

Anything slightly metallic [a metallic mix] or ferrite will delay the pulsating effects of the 60 Hz.

Instead of a electromagnet, a metallic core becomes a temporary strong magnet and even holds its magnetic charge after the power is down for a very short period afterwards.

So chose some thing that can be magnetized, yet not become a permanent magnet [for too long] , which then releases slightly after the power is off.

[MrAl]

Hello again,


Dean:
"Trouble is, MrAl, electric water heaters have two elements and two thermostats and whether the top element or the bottom element is on depends upon the internal water temperatures, so you won't get a true indicator of time"

Oh ok Dean, i didnt know that they sometimes have
two and they dont turn on at the same time.
This would mean the two element power feeds
would have to be "or'd" together. Perhaps two
diodes driving a dc clock...either element turning
on then starts the clock...just looking for the
simplest idea here. You know: resistor, cap,
zener diode, two 1N4006 diodes, low volt dc
clock, done.
But wait, what if the two elements turn on
both together...in order to calc total power
useage we'd have to know the power run time of
EACH element individually, or else have a clock
that could be run twice as fast if both are on
(assuming both elements draw same current when on).
This would REQUIRE two clocks anyway so the
original idea would work, but only with two clocks.
Then at the end of the month both clocks time
is added together after multiplying by the
relative power draw of each element (if they
are not the same that is).


Sound better?


Take care,
Al

[Dean Huster]

"But wait, what if the two elements turn on
both together..."

Al, they never are both on together. It's either the top element, the bottom element or no element but never both elements.


As a bit of additional non-trivial trivia, it should be noted that with high lime/calcium water, you need to periodically turn off (electricity and water) and drain an electric water heater, pull the bottom element and clean all the calcium out. I find that running a long screwdriver into the bottom element hole and breaking up the calcium and then using a shop vac to suck the stuff out as you turn the cold water supply back on for short intervals. It'll take several cycles of this break-it-up-suck-it-out stuff, but you won't have to totally remove the water heater to do it. The calcium collects in the bottom and insulates the lower element from the water, making it less efficient and sometimes causing it to burn out from excess heat. The accumulation of calcium also reduces the tank capacity. I removed about 3 gallons of calcium from a water heater in a house we had just bought many years ago.

Dean

[MrAl]

Hi again,

Dean:
Do both elements always have the same wattage
rating?

Take care,
Al