View Single Post
Old 07-07-2009, 09:27 PM
witsend witsend is offline
Gold Member
Join Date: May 2009
Posts: 1,881
One thing you really should know is that DSOs like the FLUKE 199 are subject to some well-known errors, particularly with complex and spiky signals, and you should not always trust their readouts. TinselKoala

Spescom - one of our accreditors - got fluke to write a letter to guarantee that the results on our meter were accurate within the frequency range tested. That's in the paper. But the point is that the more accurate the instrument the greater the evident gain. We've tested on Tektronix which in SA costs upwards of R250 000.00 and got the same results. We also went to the trouble of getting a calibration certificate for all the instruments that we used.

But the actual final proof of the pudding is in comparing the draw down rate in controls. The control - in the test result depleted within the time that the test batteries hardly showed a drop.

The other thing is that you do not need that oscillating frequency to prove a gain. The point about that oscillation is that it is well known. It is a problem when used for signalling. It is clamped out through the simple means of applying pressure to the wire. We show - through the published test - that that effect exponentially improves the efficiencies - well above any other frequency tested.

But the paper refers to various other circuits, using nfets, inductors, resistors at different points in the circuit. Whichever way the circuit is configured - provided that there is a switching cycle and that the counter electromotive force is able to be applied back to the battery supply source - the gain is inevitable.

Therefore do I wonder at your measurement of the power from the battery. If you indeed take a series of waveforms and multiple samples of each waveform - you need do nothing more than sum the voltages, divide it by the number of samples, and then do the analysis as mentioned. There is always a clear and evident gain.