Claimed OU circuit of Rosemary Ainslie

[fuzzytomcat]

Post something complete please, with connection points to the circuit indicated as well. A full schematic would be very helpful for them, as opposed to kluged-together web pages.

.99

Hi .99

Thanks for the input here .... what your proposing from Paul is a good idea because I know for a fact there are people replicating the RA COP>17 circuit that have no scope or other fancy equipment that would need other methods of testing the circuit and referencing generic diagrams without component values or no connection point locations replicators would have difficulty coming up with the correct items or even to make this possible for "ALL" open source experimental use with confidence.

This is one reason why I modified the Revised 08-12-2009 circuit "GATE" potentiometer that went from a 100 ohm 10-turn in the original Quantum article to a 5K ohm 10-turn that made the proper adjustments for any gains to be seen almost impossible, when the proper area for gains is in the 5 to 7 ohm range as indicated in my posted adjustment procedures. This is the reason why I changed the schematic to a 10 ohm multi turn potentiometer or possibly use the original 100 ohm 10-turn to make adjustments easier for any replicator to do without the "luck" factor being involved.

Fuzzy
     

[poynt99]

Rose,

Unfortunately there is an incredible and fundamental flaw to this approach.

In order to properly determine the circuit COP, the power dissipated and power supplied must be measured for ALL the components in the circuit that either source or sink a significant amount of power.

This mainly includes: 1) the power supply, 2) the load resistor, 3) the MOSFET, and 4) the low-side shunt.

Taking into consideration only the power delivered by the supply (POS) and the power measured in the load resistor (PIM) does not illustrate the overall full picture, nor allow for the complete accounting of power, especially when the power measurements for those two components is highly suspect.

The CONTROL measurements on the load resistor taken by Glen are solid and there is no concern regarding them. That is a given.

However, as per my test plan, and good testing practice, all the components as per above need to be measured and thus accounted for. In addition, when a particular measurement is clearly suspect, it is paramount to find a means of double-checking this measurement by utilizing at least one other different method.

The results for the POS in Glen's test#5 are clearly suspect, as anyone can see from the table below:

Battery Power AVG (POS) from Glen's test#5

Hour      2µs                  20µs              40µs
1    1.514827571   -2.479456000    1.500064000
2   -3.205999153   -4.045888000   -5.239488000
3   -5.533148312   -5.207520000   -1.636576000
4   -2.350759808   -1.969696000   -3.160768000
5   -2.070294717   -3.493984000   -4.054848000
6   -1.257904431   -4.098176000   -2.052608000
7   -2.550643907   -4.765152000   -3.565344000

AVG   -2.207703251   -3.722838857   -2.601366857

To review: the highlighted value in the table of 1.5W was chosen by the Ainslie team because it was a "worst case" measurement in terms of showing the lowest COP when compared to the 4.5W Control test for the equivalent TRAA. This results in a COP of 3.

Why not choose the -5.5W figure? That would yield a COP of more than infinity, and would indicate that not only is the COP=infinity, but that a net charge is going back into the supply with a net average current of roughly 220mA. In simple terms, this means that on average the supply is not sourcing any power at all, but is in fact gaining charge. Running the circuit at this particular setting, it ought to be quite evident that the battery is readily charging up, yet the overall profile of the battery voltage in Glen's test clearly shows the battery voltage falling off over time. Glancing once again at the above table of results, one can see that in fact 19 out of the 21 measurements for POS indicate a net charge going back into the battery, so why did the battery voltage continue to fall over the duration of the experiment?

I challenge anyone, academic or not to justify, rationalize and validate this procedure in light of the wildly varying (and particularly the negative) results shown in the table. It is difficult to believe that any academic would condone and stand by this table of results as being usable. However, this question, along with the filtered-shunt question will be asked of the academic-types I am able to find.

Getting back to the thermal method of measurements, yes it is currently being done on the load resistor. What I, Paul, and MileHigh are suggesting is that this be followed through with the MOSFET and shunt resistor as well, and compared to and used as a check against the oscilloscope measurements being obtained.

For those that want to test this circuit for its OU potential can easily do so using this method, and no high-end oscilloscope or expensive exotic probes are required to do so.

.99

[fuzzytomcat]

Hi .99

I did do some thermo profiling in my TEST #2 when I had the improper 5K ohm 10-turn gate pot and was unable to get below 140 mv "mean" on channel-1 until I changed the potentiometer out in a later test to get into the 40 to 80 mv range.

Unfortunately there were many members that was saying that I was wasting my time with such measurements on items like the Mosfet, 555 timer and Potentiometers so I stopped publishing this information but continued to measure these for my own benefit and the temperatures still are in direct relationship with the rise and fall of the prototype "quantum" load resistor I'm using now.

Here are the links to this and attached images for those all those that may have missed my attempt of recording these values ....

TEST #2
http://www.energeticforum.com/69966-post2890.html

ORIGINAL HAND WRITTEN DATA
http://i276.photobucket.com/albums/kk15/fuzzytomcat/scan0004.jpg
http://i276.photobucket.com/albums/kk15/fuzzytomcat/scan0005.jpg


Fuzzy

[poynt99]

That's great Fuzzy

I would encourage you to perform a formal thermal test once you have a setting you are happy with and take a measurement of the load resistor, MOSFET, and shunt resistor temperatures.

Then perform a control test on the MOSFET and shunt to establish the power required to obtain the same TRAA in each case.

Now when combined with the load resistor control Wattage you previously determined, you have the total RMS power dissipated or consumed by all 3 of these components.

What is left is to obtain the battery power being supplied when the circuit is in operation. I have provided a reliable method to easily do this using the filtered shunt. It is easy to build, and if made with suitable connectors, can be inserted or taken out of your circuit at will for comparison purposes. Use it and see if it affects your measurements and scope shots. If it does not, then why not use it? At the moment, using the oscilloscope along with the shunt wave form and probes you have is not providing a useful or accurate measurement, so I suggest the alternative.

Now with the accurate power output from supply measurement (POS) and the 3 dissipated power measurements obtained using the thermal method (PIL, PIM, and PIS), you can readily see if the POS is less than the 3 dissipation powers combined. If it is, then you have OU.

If POS < (PIL + PIM + PIS), then COP>1
If POS > (PIL + PIM + PIS), then COP<1

.99

[fuzzytomcat]

That's great Fuzzy

I would encourage you to perform a formal thermal test once you have a setting you are happy with and take a measurement of the load resistor, MOSFET, and shunt resistor temperatures.

Then perform a control test on the MOSFET and shunt to establish the power required to obtain the same TRAA in each case.

Now when combined with the load resistor control Wattage you previously determined, you have the total RMS power dissipated or consumed by all 3 of these components.

What is left is to obtain the battery power being supplied when the circuit is in operation. I have provided a reliable method to easily do this using the filtered shunt. It is easy to build, and if made with suitable connectors, can be inserted or taken out of your circuit at will for comparison purposes. Use it and see if it affects your measurements and scope shots. If it does not, then why not use it? At the moment, using the oscilloscope along with the shunt wave form and probes you have is not providing a useful or accurate measurement, so I suggest the alternative.

Now with the accurate power output from supply measurement (POS) and the 3 dissipated power measurements obtained using the thermal method (PIL, PIM, and PIS), you can readily see if the POS is less than the 3 dissipation powers combined. If it is, then you have OU.

If POS < (PIL + PIM + PIS), then COP>1
If POS > (PIL + PIM + PIS), then COP<1

.99

Hi .99

I'm sure you as being one of the replicators agree the RA COP>17 circuit to make heat is a easy task but the hard part is to do it the most efficient way which is how I approached this project. I cannot say how important it is for anyone wanting to get results in any gains is by "fine" adjustments to the "gate" potentiometer in the 5 to 7 ohm range monitoring the battery voltage with a DMM to get the "HIGHEST" 24Volt battery bank voltage possible, watching the gate potentiometer backlash and getting the "LOWEST" mV reading from 40 to 80 mV on the oscilloscope probe tip between the Mosfet source and the shunt resistor.

I use the 100ns scope shot as my "road map" getting the Mosfet source plus 555 timer off and the "drain" spike four (4) divisions apart, getting the highest drain spike plus lowest Mosfet source mV reading as possible and the best results happen there using the 32mm prototype "Quantum" resistor. The store bought resistor is impossible to get to these low Mosfet source pin readings and I will be making a even larger diameter 10 ohm prototype resistor to see if the possible gains can be greater yet.

I have ordered a "non-inductive" 0.25 ohm 1% resistor for the shunt from "Caddock"  http://www.caddock.com/  it's the same resistor that "Groundloop" is using and shown in his assembled PC board http://www.energeticforum.com/67009-post2474.html  it appears to be a P/N:  MP930-0.25-1%  http://www.caddock.com/Online_catalog/Mrktg_Lit/MP9000_Series.pdf

It appears Groundloop may also still have the 5K gate pot installed, and I think you also mentioned there may be a actual ground loop in the PCB that may have some bearing on his test results.

There is also some problems using the PCB mounted cermet type trimmer potentiometers as the tolerance is between 10 to 20% and the "Vishay" 10-turn SP534 potentiometers that I use http://www.vishay.com/docs/57065/533534.pdf  are a 5% tolerance which should minimize the readings jumping around as they do in the gate ohm ranges I've indicated.

As far as the "filter" you suggest ....... as long as it is a circuit that most members agree upon and documented properly I see no problem with me trying it after the changed "non-inductive" shunt resistor testing happens. I know that you are, but possibly other members are not aware of the possible first "open source" paper that will be submitted to academics for review so any tests that may be required that can be done and relevant, with good or bad results should be tried and posted such as I have.

In the future prolonged testing I will include if I can the temperatures of the components again as  they appear to possibly be of interest and could be used in further testing if actually required.

Fuzzy