[Return to Device Update] [Return to Feature Articles]

Infinite Energy Device Update
Published in IE Volume 7, Issue #38, July-August, 2001

New Energy Research Laboratory Device and Process Testing Update
Ken Rauen

Several sonofusion test runs here at New Energy Research Laboratory with a titanium target were reported in Issue No. 37 to produce excess heat in the 3 to 4 watt range. One produced 3.1 watts after the reactor joule heater had been turned off, which was a surprise to us, because Roger Stringham's experience showed that a reactor temperature around 100°C was necessary and a colder reactor would quench the reaction. As of the last issue, we did not know what to make of this apparent excess heat output.

A D2O run at 1500 V peak ultrasonic exitation. A portion of one excess heat run is shown from it's data aquisition computer database file. On the graph, excess heat (the noisy trace) has a vertical scale of 2 W per division and the reference center line is 0.0 W.

We now accept the results of that run to be genuine excess heat. Many more tests have been conducted under many different conditions in an attempt to kill the excess heat phenomenon, and it took many tests to find that out. We want to know which conditions produce excess heat and which do not. We also want null tests to be reliable so we can identify the on/off controlling parameters. The more conditions we can identify which do not produce excess heat, the more confident we are that our methodology does not have a systematic error creating an artifact, a false positive signal. A balance is necessary; repeatable zero tests are needed, and of course, repeatable excess heat events are needed also. We now have many of both.

NERL's sonofusion research has been perplexing. Cold fusion categorically has been extremely difficult to reproduce, and that had been NERL's experience until recently. Since Issue No. 37 appeared, and for a short time, we were faced with the novel experience of always finding excess heat, and we could not find conditions that would not produce excess heat except by turning off the ultrasonic oscillator. It was a very strange feeling. We were searching for systematic errors for a month. Finally, we found some conditions which do not produce excess heat and could verify the system calorimetry with more than the joule heaters in the oscillator calorimeter and the reactor calorimeter.

It turns out that excess heat in the 2 to 4 watt range can be produced not only by Roger Stringham's method of ultrasonic activation with applied argon pressure of up to 50 psig and operating just over 100°C, but also by: 1) no applied joule heat and temperatures close to room temperature; 2) argon gas pressure from 0 to 50 psig; 3) ultrasonic power driven with transducer voltages from 600 VAC peak up to 1,500 VAC peak; 4) target or no target in the reaction chamber, and 5) heavy water, normal distilled water, or deuterium depleted (3 ppm deuterium) water. See the display of part of one test run showing excess heat. There seems to be no significant difference in the excess heat by any method, though we have not statistically analyzed any data. Roger believes these all on results demonstrate what he wanted with the ultrasonic ring-down pulse; that sweep of operating conditions runs through the unknown but expected "sweet spot" in every test, and in every ultrasonic burst. What puzzles us most is the apparent lack of correlation between excess heat and deuterium concentration. One of our peers in the scientific community thinks we might have a systematic error, since there is no apparent correlation of excess heat with deuterium concentration in the water.

We searched for good zero conditions just as much as for the esteemed excess heat. Nulls were found with: 1) an empty chamber (though not necessarily perfectly dry), and 2) replacing the piezo transducers with dummy resistor loads. The empty chamber has shown us that the Ohio Semitronics wattmeter, which monitors the input electrical power to the oscillator (inside its own separate calorimeter) has a 0.9 watt error, which makes it read the real power too low by that amount, so the computer which monitors the entire measurement system calculates a false excess heat of about 0.9 watts, which has ranged from 0.4 to 1.0 watts excess heat. This means our 2 to 4 watts of excess heat results should really be 1 to 3 watts of excess heat. The numerous empty chamber tests were supported by tests with the resistive loads on the oscillator. Now we are sure that excess heat is occurring reliably, and that our test apparatus is performing well, except for the OSI error, which we do not want to tamper with; we know its error and it is consistent. We want to leave it where we know it is.

As we go to press with this issue, we are testing a new wattmeter that is custom-made by Pioneer Microsystems, Inc. of Pittsburgh, Pennsylvania. So far, the new wattmeter reads the same as the OSI reading minus the oscillator calorimeter heat loss. Piezo loading is the next level of testing.

[Top of Page]