After going over the data, there is good correlation between diode and any capacitor placed across it. In short, it appears the DC voltage produced by the diode is greatly effected by long term capacitance. IOW, if you place a capacitor across a diode, the DC voltage produced by that diode will slowly decay over time.
Data suggests that an undisturbed diode with high Rz might last up to 10 hours, but this appears to be an unusually long time, as most diodes may become starved of noise within minutes of being connected.
For best results, do not connect anything to the diode while it is recovering/resting. Use an electrometer to measure the DC voltage, an electrometer that has less than 10 fA input bias current. I use the INA116PA. If you build the INA116PA circuit, then you'll need to measure the Ib before using it on a diode. Get the entire circuit ready for the diode. Before placing the diode in the input, place anywhere from 470 pF to 0.01 uF low leakage capacitor across the electrometer input. Turn it on, let it warm up a few minutes, carefully discharge the capacitor, and then time how much the electrometer charges the capacitor. The Ib is C * V / t, where C is the capacitor capacitance in farads, V is the change voltage across the capacitor in volts, and t is the time in seconds. If you are unable to obtain less than 10 fA, then try removing everything on the electrometer input and solder only the capacitor, and repeat the test. Remember, the capacitor and the electrometer should all be in mid air!!! Also, there *cannot* be any appreciable breeze because that will blow ions from the battery to the capacitor. You may struggle with this for a few days, but soon enough you will begin to understand the physics behind this in that what's bad and what's good for lowering Ib. Once you get that down, then you can connect the switches and redo the tests. For lowest Ib, the amount of metal surface connected to the electrometer input should be kept to a minimum. The more metal surface = more ions it will pick up. Such ions do not make any measurable difference with diodes, unless you're measuring nanovolts, as such ions even from a bad setup will be on the order of femto amps, but you will want to have the bias current less than ~ 10 fA so that you can place a capacitor of low capacitance on the electrometer. Such a capacitor is not need when the diode is producing relatively high DC voltages, and therefore you may not want to use a capacitor since diodes don't like capacitance. Although, for demonstrating to a notable scientists, you might want to have some capacitance to put the scientist at ease.
All of this may sound tedious and difficult, but it's not that bad. In a few days you'll probably have it all working well, maybe a week at most. If it takes you longer than a week, then please by all means make an anonymous post to the electronics newsgroup/Usenet at groups.google.com, or send one of the EE's a private message. Someone will most likely help you. Or send me an anonymous email.
Once you get the electrometer Ib to less than 10 fA, then you can place up ~ 470 pF low leakage capacitor directly on the electrometer input. So if your diode Rz is 100 Gohm, and the capacitor is 470 pF, then it would take the diode 141 seconds to charge the capacitor to 95% of whatever voltage the diode is producing. See the online RC Time Constant Calculator for further calculations. This means you would leave the diode on the electrometer for ~ 141 seconds before obtaining the DC voltage, and then you would *immediately* reverse the diode input (example, if you're using Mercury tilt switches and immediately take the DC voltage. Remember, the Mercury tilt switches that reverse the voltage should be connected before the capacitor, as this will allow you to obtain a difference in output voltage *immediately* since it will be reverse the capacitor as well across the electrometer input.
Created on 2009-07-09 17:13:52 by EnergyMover
FE diodes, FE Misc devices, Free energy, Free energy devices, Science, Testing procedures, Diode, Free energy, Testing procedures