good point xee about th heating and change of resistance.

EMD, suggest using at least a 10W power resistor in there. if  the transistor doesn't get too hot let it run for a bit and  check the load resistance (disconnect it from the circuit) right after you take your voltage measurement.

If the load resistor was 190 ohms instead of 150 ohms due to heating, then you would no longer be over unity:

Just make sure resistance does not drop It may drop with higher temperature as well, depending on resistor's material.

Hi!
Sorry to say, but this simple circuit is not, and will never be an  "OU" configuration.... A transistor and a few components - I think this is one of the most researched circuits in past century, wouldn't you think?

Like some people said before, the faulty measurement procedure and consequent errors are to blaim...


Seeing all those different "OU el.circuits" all around the web - not even a single el. component / circuit (being that passive or active) has ever been scientifically recognised as an option to invalidate CoE...

The man-made electronic components are by it's origin and definition just a conservative devices. The passive ones, like C and L are the ones which can sparckle some ideas.... Let's say, in a resonant condition. Which would be exactly unity under ideal conditions....
Yet you all know that there's actually no ideal capacitance or inductance... There's always a resistive component, which is an equivalent for a friction in mechanical devices. Having an oscillating circuit, there are radiated losses, too.

The main goal for engineers wrt. electronic devices is to come as close to an unity as possible (in energy terms). But semiconductors are 'f...ed up when it comes to energy efficiency'.

Unless someone defines where the additional energy is comming from (sucking energy from e.g. the "surrounding heat, Aether, ZPE, Dark energy,...  "), there can be no valid claims about 'OU'.   

To tell what I've mentioned exactly: "standard model" (which is based on laws of conservation among other things) has no gravity in it, and has some deviations with real world data. You just can't be serious telling overunity does not exist in this universe, especially considering some people achieved it. Do not underestimate "filters of perception" in human brain. Many "unwanted" or "unimportant" things are usually left unexplored. Simply remember who you was at your younger ages: you "saw" basically nothing, because you had no idea about what you saw. In an older age you have a lot of ideas, and you "see" things, but unlike in a younger age, you are stopping new ideas from entering your brain. That's what "filter of perception" is. Your younger state is what brought you to where you are now.

@ xee,   I used a 1/4 watt resistor because that's what I have.   I'll go buy some higher values ones.  Also,  I don't worry about the resitance GOING UP WITH TEMPETURE,  since that's not what happens in practice.  As the temperature heats up, the resistance goes down, not by much in this case.  For example lightbulbts can drop from 130 ohms to 30 ohms, I've measured it first hand.

@grumpy,  the battery does not heat up.  I understand those battery charts, its the mAh, a measure of energy, so if you drain more you depleate faster, that's all.  But you can draw two amps if you really wanted.

@aleks,  yes the voltage is more like 6.5, but I like to err on the side of caution.  I did some rough calculations of my margin of error and the COP can varry from 1.35 to about 1.20.   

@willitwork,  yes I've been thinking and planning to do this ultimate test to prove OU without a shadow of doubt.  However, it's a bit trick, and the reason is that the oscillator and the output power depends on the load resistor.  I tried different resistors and I get different performance.    This is not uncommon, most stages in an electronic system need matching between stages.  So, I am researching and thinking how I can do this effectively.  OU engineering is not as easy as it looks.

EM

@ EM

I suggest you measure your VBE and you will probably find it very high and almost certainly above the typical SOA limit of 5V. This will degrade the junction and bring the transistor gain / hfe down a peg or two until the transistor eventually destroys. Strange things happen during this process and I have experimented with a few configurations without a base emiter diode as per your circuit.

Hoppy

....

@willitwork,  yes I've been thinking and planning to do this ultimate test to prove OU without a shadow of doubt.  However, it's a bit trick, and the reason is that the oscillator and the output power depends on the load resistor.  I tried different resistors and I get different performance.    This is not uncommon, most stages in an electronic system need matching between stages.  So, I am researching and thinking how I can do this effectively.  OU engineering is not as easy as it looks.

EM

Hi EM,

First I would suggest trying a simple linear regulator (I think of the adjustable LM317) set to  7 or 8V DC output just to match your present battery voltage  and load the regulator output with 100mA  to see how your oscillator behaves with that type of load (of course your output coil is lifted from gnd and drives a full wave bridge (4 x 1N4148 or 1N914 or similar HF diodes) with a puffer electrolytic capacitor as earlier was already suggested). 

And if your oscillator still consumes around 100mA from the battery while your regulator reproduces this power input then you could loop back its output to replace the battery...

I know the linear regulator surely wastes power which may amount to even loosing the higher than one COP margin but this would be the simplest first try to investigate and get a further insight on your circuit.
Next step would be to look for a simple DC-DC converter which can have over 90% efficiency that surely maintains the higher than one COP of your oscillator circuit.  Here is some switching type ICs for this job (of course there are many other types too): http://www.siongboon.com/projects/2005-08-07_lm2576_dc-dc_converter/   and maybe the LM2575-ADJ type (its output voltage is adjustable, not fixed) can be a simple candidate for this test.

Wish you good luck and experimenting.

Couple of questions:  is the capacitor in the order of several nF of value between the gnd and the transistor base? Also the value of the capacitor for the tank circuit that affects the running frequency?

rgds,  Gyula