COP 17 Heater | Rosemary Ainslie | Part 2

[witsend]

Gad, if Bart recommends this so do I. And I love simple solutions.

Let us know if all's good and thanks for rallying Bart.

[gadh]

On second thought , and after consulting with an expert in physics and electronics, he suggested that using a simple voltage divider wont be enough for such spikes. in order to get the accurate divider with such spikes, we'll need to add some capacitors to the divider circuit, which will make it complicated (rather than a simple divider which uses only 2 big resistors) and i think it'll affect the entire circuit and also will cause us to re-calibrate the scope.
therefore i do not see any solution (for now) other than using a special high voltage probes (which also have a tuning mechanism).

Bart/Rose, can you recommend on a voltage divider circuit that wont affect the basic circuit (on getting the required resonance) ? please add the divider schematics if you can.

[witsend]

On second thought , and after consulting with an expert in physics and electronics, he suggested that using a simple voltage divider wont be enough for such spikes. in order to get the accurate divider with such spikes, we'll need to add some capacitors to the divider circuit, which will make it complicated (rather than a simple divider which uses only 2 big resistors) and i think it'll affect the entire circuit and also will cause us to re-calibrate the scope.
therefore i do not see any solution (for now) other than using a special high voltage probes (which also have a tuning mechanism).

Bart/Rose, can you recommend on a voltage divider circuit that wont affect the basic circuit (on getting the required resonance) ? please add the divider schematics if you can.

Gad - I only know of the probes we used with our Fluke and that came with it's own variety of probes. Essentially you need one with a higher attentuation with a bandwidth equal to or greater than 1 MHz. I've just spoken to Donny and will email you his Skype address. He'll give you all the advice you need here - but has just assured me that the divider may not be good precisely because, as you mentioned - it may interrupt the required resonance. But look out for my email. He's at home and expecting you to call.

When the two of you have got to a solution please post it here. I think there are others who may also be asking the same questions.

Thanks Gad.

[gadh]

About solving the high voltage probing in the scope

You can build a home-maid high voltage probe, using the following guide:
How to make a 100X oscilloscope probe - a great resource for How To's from WikiaTESTEC|HV250|PROBE, HIGH VOLTAGE, 2.5KV | Farnell Israel

The HV probe (see previous link) impedance is rated at 100M, with a BW of 300MHz which should allow the resonant harmonics to feed through mostly undisturbed.
The one thing to remember is to adjust the scope probe compensation (screw on the probe body) to allow a test square wave signal of say 100KHz to have the best edge transition,
i.e the waveform edges should be close to square as possible. See waveform examples in the first link (wiki.com) I posted here.

Many thanks for Donny for his assistance on this issue !

[witsend]

You can build a home-maid high voltage probe, using the following guide:
How to make a 100X oscilloscope probe - a great resource for How To's from WikiaTESTEC|HV250|PROBE, HIGH VOLTAGE, 2.5KV | Farnell Israel

The HV probe (see previous link) impedance is rated at 100M, with a BW of 300MHz which should allow the resonant harmonics to feed through mostly undisturbed.

The one thing to remember is to adjust the scope probe compensation (screw on the probe body) to allow a test square wave signal of say 100KHz to have the best edge transition,
i.e the waveform edges should be close to square as possible. See waveform examples in the first link (wiki.com) I posted here.

Many thanks for Donny for his assistance on this issue !

Many thanks for this Gad. Much appreciated.

[witsend]

Guys, Just a quick update.

There's been delays with the cylinder. The supplier was trying, unsuccessfully to source some thicker tubing to house the resistor. None available so will be using a slighter thinner type as a 'starter'. Therefore delivery is now detailed for Friday - God willing. I'll keep you posted.

Also design modifications to accommodate safety concerns so he's included a pressure valve.

I'll photograph all when I get receipt. Probably only at the weekend.

[witsend]

Guys - just for the record.

I have been knocking on the campus doors for many years now. But was always met with rank scepticism to the claim.

Times have changed and - as a tribute to this change - a local university has agreed to use this project to test these claims. Thanks to an invester in this project we intend puting these tests directly onto an application. And thanks to the impartiality of the university, we will be able to get some accurate measurements and, if results show this, some really weighty accreditation. But equally, if the results fail, then it will put paid to this technology. Only these results will decide the matter. And, in my book, that's fair and impartial and scientific. I don't think that there's been any kind of endorsement of the claim - certaily none that I know of. At this stage it is simply a question of 'let's evaluate this from the evidence'. Which certainly has scientific merit. But that evaluation depends on the outcome of the tests. And those tests still need doing.

[witsend]

Here's the hot water cylinder guys. Finally arrived.

[witsend]

Gad, I'm now ready and able to take that call if you're up for it. Apologies for not being available earlier.

Kindest regards,

[gadh]

finally - my experiment results

Hi all.
finally i got the load resistor built and bought a high voltage probe (specs - as i wrote before).
Experiment description:
load resistor (see image) - NiCr 80/20 wire, gauge 20, spacing of 1-1.5mm between each wire. using teflon rod, diameter 34 mm - but with a carving of 1mm depth - so inner diameter is 32mm. 48 turns = 10.5 ohms resistance.
MOSFET - i glued a heatsink to it (not shown here) - taken from an old pc motherboard.
Wiring - Power circuit : 1.2m x2 between 24v batteries to load resistor and to the shunt, approx. 14 gauge wire, single copper thread (standard power electricity wire used for 220v in your house, i think). other wires are 10-35 cm (not shown in the images).
555 timer circuit: 24 gauge wire, single thread.

scope channels (see in the images):
ch 1 = MOSFET Drain leg (load resistor spikes).
ch 2 = gate pot. leg to 555 (the 555 output signal)

experiment process: first of all, in order to get the required 3.7% duty cycle with 2.4khz base frequency, we disconnected the power circuit from the 555 circuit. otherwise - we could not reach under 8% duty cycle !
after disconnecting , we managed to get about 3% duty cycle and 2.4 khz freq., using the following potentiometers values:
2k ohm pot. = 194.5 ohm (called the "duty cycle pot.")
10k ohm pot. = 9.87 kohm (called the "frequency pot.")
100 ohm pot. = (called the "gate pot.") - value does not matter at this stage (since the gate is not connected to the 555 timer circuit) ! we'll deal with it later on.

after getting the required freq. + duty cycle, we connected the power circuit to the 555 circuit, and adjusted the 100 ohm pot to its lowest value (almost 0 ohms). the spikes on the load resistor were the biggest. after tuning the gate pot. to 0.8 ohm, we got about 420 volts (see in images - ch 1) and we got an instant self oscillation of ~318 khz !! we also got a lower spike in the same freq. of about 150 volts (see in images - ch 1 too).
the 555 signal was distorted due to the oscillation - we think (see ch 2 in the images).
so thats it i think. but i did not get any "aperiodic oscillation" (to my knowledge...).
what is the next step in order to get ready for the energy measurement and expecting COP > 1 ?
how to get the aperiodic osc. ? is it needed or what we got is enough ?

thanks to you all for your kind support,
Gad

[witsend]

Hi all.
finally i got the load resistor built and bought a high voltage probe (specs - as i wrote before).

Gad - very well done.

Experiment description:
load resistor (see image) - NiCr 80/20 wire, gauge 20, spacing of 1-1.5mm between each wire. using teflon rod, diameter 34 mm - but with a carving of 1mm depth - so inner diameter is 32mm. 48 turns = 10.5 ohms resistance.
MOSFET - i glued a heatsink to it (not shown here) - taken from an old pc motherboard.
Wiring - Power circuit : 1.2m x2 between 24v batteries to load resistor and to the shunt, approx. 14 gauge wire, single copper thread (standard power electricity wire used for 220v in your house, i think). other wires are 10-35 cm (not shown in the images).
555 timer circuit: 24 gauge wire, single thread.

All the above is good. Not sure that the glued fet will hold under operating conditions. Our experience is that this can get quite hot.

experiment process: first of all, in order to get the required 3.7% duty cycle with 2.4khz base frequency, we disconnected the power circuit from the 555 circuit. otherwise - we could not reach under 8% duty cycle !

after disconnecting , we managed to get about 3% duty cycle and 2.4 khz freq., using the following potentiometers values:
2k ohm pot. = 194.5 ohm (called the "duty cycle pot.")
10k ohm pot. = 9.87 kohm (called the "frequency pot.")
100 ohm pot. = (called the "gate pot.") - value does not matter at this stage (since the gate is not connected to the 555 timer circuit) ! we'll deal with it later on.

after getting the required freq. + duty cycle, we connected the power circuit to the 555 circuit, and adjusted the 100 ohm pot to its lowest value (almost 0 ohms). the spikes on the load resistor were the biggest. after tuning the gate pot. to 0.8 ohm, we got about 420 volts (see in images - ch 1) and we got an instant self oscillation of ~318 khz !! we also got a lower spike in the same freq. of about 150 volts (see in images - ch 1 too).
the 555 signal was distorted due to the oscillation - we think (see ch 2 in the images).
so thats it i think. but i did not get any "aperiodic oscillation" (to my knowledge...).
what is the next step in order to get ready for the energy measurement and expecting COP > 1 ?
how to get the aperiodic osc. ? is it needed or what we got is enough ?

thanks to you all for your kind support,
Gad

I'm still not sure how you adjusted the duty cycle but what's wonderful is to see that it went into it's own required resonating frequency. Very well done indeed Gad. That's the whole object of the switch is to let the system find its values. You can't really impose this.

Now you need to do detailed measurements across the shunt to determine the amount of energy delivered by the battery. And you need to assess the amount of energy dissipated at the resistor.

I should be available most of tomorrow. Well done Gad. If you can call me tomorrow I may be able to get a conference call going.

[b4FreeEnergy]

Hi Gadh,

Nice to hear you made it too. I guess you now have about the same situation I have. The only big difference is that you got it into oscillation with that base frequency of 2400Hz. I did not succeed at all with that low base frequency and had to go much higher before I got my circuit swapping over to that other much higher frequency 'automatically'. (After adjusting the pots a little) The biggest difference between your and my setup is the diameter of the coil I think. I only have 28mm. I hope I can find some time this weekend for a few measurements.

Cheers,
B

[gadh]

Bart, maybe another thing you need to check - first disconnect the 555 circuit from the power circuit, adjust the base freq. and duty cycle as Rosemary suggested, only then connect the 555 circuit to the power circuit - and adjust only the 100 ohm (gate) pot. to some low value (closest to 0 ohm).

[witsend]

Guys - just to let you all know. Preliminary results indicate that Gad may have managed a successful replication of this circuit. We're awaiting finer details of measurements but his mean average voltage across the shunt is negative notwithstanding which there is clear evidence of heat dissipated at the MOSFET and the resistor. We're awaiting fine tuning and some more detailed measurements.

Very well done indeed Gad. It's not an easy waveform to find and I think it's a tribute to your talents that you've managed this much with all those difficulties you're having with your measuring instruments.

[witsend]

Just to remind our viewers and our readers here - the shunt is positioned at the negative rail of the battery. The thesis is that energy that is first delivered by the battery is then returned to the battery to recharge it. Effectively therefore one would have some variation of a sine wave but with a DC setting the difference between these to values would then represent the actual energy delivered by the battery. If the net result is negative, then one may conclude that there is more energy returned to the battery to recharge it than was first delivered. This flies in the face of mainstream prediction.