Joule Lamp

[SeaMonkey]

Hi Lynxsteam & people,
...

I have tried to put a 1N4007 diode (your initial CCT) between the transistor base
and the transistor emitter. The amp consumption is divided by about 3 but
my 6 watts fluorescent tubes bright less.
...
Very Best,
Jean

You may want to try three or four diodes series connected across the base-emitter
leads.  The purpose of those diodes is to limit the amount of reverse bias which
is applied to the base-emitter junction by the feedback winding during the time
the transistor is being turned OFF.  A certain amount of reverse bias at that time
is desired since it will aid in turning the transitor off rapidly and completely.

However, if the reverse bias pulse is strong enough (More than about 7 Volts) to
take the base-emitter junction into breakdown then the transistor can do unpredictable
things and its power dissipation will increase.

Limiting the reverse bias feedback pulse to about 4 to 5 Volts will expedite
turn OFF for increased efficiency and prevent it from taking the base-emitter into breakdown.

[JouleSeeker]

These AirCore coils will oscillate when the output is placed in H2O.  You can vary the distance of an aluminum rod and a zinc rod until you match the Hz that will match the load you are looking for.  Then measure the ml or moles of H2O disassociated into hydrogen and oxygen.  You could compare the power required to do this to the watts at input.  5-20 watts isn't much so you probably have to do this for a while. 
I ran mine for about two minutes and bubbles do form.  I could hear a slightly audible tone in the glass of water and the frequency was 6.87 khz.  The transistor started getting warm.  6.87 khz would be similar to about ten - twelve bulbs on the circuit.
I don't know if the power equivalent to disassociate water and emit light are comparable.

This is very interesting, Lynx...
But I have a question about Water electrolysis and heating with AC...

Let's start with 60 Hz 120 V from the mains, two electrodes in tap water.  How much electrolysis into H2 and O2 will there be?  how much heating?

Now I've done numerous experiments with "Davey sonic bells" just like this -- two bells, one connected to the hot line and one to return/ground.  I observe slow heating from room temp -- not many bubbles indicating electrolysis.  There are some small bubbles, but I think this is due to dissolved air coming out of the water... not sure.

Further, the heating agrees with what I expect for simple resistive heating of the water -- watts in, hot water as the result.Q (Joules) =  Cp * mass * (change in temperature)The specific heat (Cp) of water is 1 calorie/gram °C = 4.186 joule/gram °C.

Distilled water heats very slowly -- very high resistance.


Next let's go to higher Freq, say 20 KHz.  I haven't done the experiment with water; but I expect no significant electrolysis again.

I'm here to learn -- is this right,  no significant electrolysis  compared to resistive heating, with AC in tap water?

I'm particularly curious about using this method to measure output power in a blocking-oscillator circuit, with the output AC used to directly heat water... I think Lynx was suggesting something like this, with electrolysis...

Comments welcomed.

[Lynxsteam]

If you use aluminum for one plate and galvanized steel for another (or other metal except al) it forms a diode in water.  By adding baking soda to the water the aluminum becomes an even better diode after an hour or so with AC.  I am not sure what you are trying to do.  Can't you just measure the AC amps to the bulbs?  Is it too imprecise with the high frequency?

LEDs don't really exhibit resistance, they have an I/V characteristic.  So you can't really calculate V2/R for power.  It just so happens that the voltage drop is about 3-5 volts so its perfect for reverse biasing the transistor.  Its also tough to calculate resistance for an incandescent bulb because resistance increases as the bulb gets hotter.

Are you trying to determine the efficiency of the DC converter?  Aren't we mainly concerned with power in/light energy out?  Are you dissatisfied with the light box measurement?

This evening I will do a video showing my garage lit with LEDs utilizing grid power and then with the same bulbs utilizing the LJL DC converter.  That's where "the rubber meets the road".

Also, not sure how many people are reading this stuff, but I have a question.  Would people be more interested in mounting the converter near the battery, and have a remote switch near a wall plate, or have the converter mounted with integral switch near the wall plate?  Would it be more convenient to have one large converter, or a couple small converters?

[JouleSeeker]

  While I'm consistently seeking new and better ways to measure power output, I should note that my light-box has been checked with a number of lumens-rated bulbs, and it shows considerable consistency and reliability -- as demonstrated in the graph of the data I obtained with it, attached. 

  I do think it is important to use more than one method of measuring power output; lumens measurement can be an effective tool as we seek improvements.

[Lynxsteam]

Joule Seeker,

I agree it is important to find out how efficient this circuit is.  Verifying a couple different ways is smart.

Here is another way to measure.  It isn't quantitative, but rather it is the "qualitative" measure some of the forum members here have wondered about.  The one thing I could have done but didn't do would have been to install equivalent incandescent bulbs.  We pretty much know that the watts would have been about 140 and much of that power is creating heat. 

I was impressed not only with my device keeping pace with Detroit Edison but the quality of the light from these bulbs.  Its very pleasant, especially compared to the glare of CFLs.

Here's the video.
http://www.youtube.com/watch?v=LLqJ3CVVEQc