MacGyver (aka JohnyFive) LENR experiment

  • magicsoundAm I right that the paper you tested was 4mm far from the mica window?

    If yes then putting it closer will increase the counts A LOT.


    As was correctly pointed out by Robert Horst, Mica-window pancake GM detectors are very sensitive to false readings from static charge at the mica surface.

    The one I use (LND7317) is expensive ($200) and the mica diaphragm is very fragile. It's also a very high impedance source, meaning the electronic amplifier connected to it is sensitive to any kind of RF that might get into the signal wires.


    To protect it from both these issues, I enclosed the detector in an Al alloy box and used a piece of Al screen over the diaphragm. The detector body carries the positive high voltage, so a 1.5 mm rubber gasket is used to insulate it from the box. The thickness of the box and the gasket together create the ~4mm spacing. Thickness of the screen is negligible.


    From http://physics.usask.ca/~bzulk…s251manual/alpha_2011.pdf

    The range of Alpha particles in air is ~4mm at 500 keV. So IF the particles are Alphas (unknown), they would have to be very low energy to be absorbed by the air gap. EVOs on the other hand, are thought to not be readily absorbed by air, so the 4 mm gap should not be an issue.


    Alphas are highly charged and would probably be blocked by the protective screen anyway. Therefore, what I might be detecting is unlikely to be Alpha particles.


    Here are some images of the detector assembly, disconnected from the electronics and with the back cover removed.



  • magicsound

    Is there higher resolution data available for the active run file? It only has one sample every 10 seconds in the region of interest.


    I have edited my previous comment with a graph with an applied offset and made it clearer that it might not be 100% accurate.

  • Is there higher resolution data available for the active run file? It only has one sample every 10 seconds in the region of interest.


    Nope, due to the length of the experiment, I chose the 10 second interval. So another lesson learned here - do separate high-resolution data logging for the test interval of interest.


    Thanks for adjusting the graph, looks more correct now as far as causality anyway.

  • ... Mica-window pancake GM detectors are very sensitive to false readings from static charge at the mica surface. ...

    The one I use (LND7317) is expensive ($200) and the mica diaphragm is very fragile. It's also a very high impedance source, meaning the electronic amplifier connected to it is sensitive to any kind of RF that might get into the signal wires.


    To protect it from both these issues, I enclosed the detector in an Al alloy box and used a piece of Al screen over the diaphragm.

    Your box with the screen looks like it should do a good job of attenuating the effects of external electrostatic fields and EMI although it would be very difficult to put error bounds on whatever effect remains.


    But there is another possible electrostatic effect pointed out by a friend of mine. Electrostatic charge collects charged radon byproducts from the environment. There is a series of papers on classroom experiments that use this technique to gather radionuclides. See this paper:


    https://www.cns-snc.ca/media/u…rs/2C_nuclear_balloon.pdf


    After rubbing the balloon, in just 20 minutes it collected enough to make the GM counter read 1000x background (40 -> 4000 cpm). This makes 1.6x background sound like a possible artifact of accumulation of radioactive dust over the charging period. Here is a figure from the reference above:



  • This graph shows rad counts from all working sensors in the file. 30 samples here is about equivalent to the 300 samples CPM calculation in the live stream.



    180 samples = 30 minutes CPM. The radiation testing period becomes visible even in the snapshot of the entire run.



    I'm not sure why, a closeup shows that there's an offset between the reported times and the elevated radiaton events. It doesn't seem to be due to the CPM calculation I'm doing. Anyway, more of interest is that there seems to be a correlation with what is called "CounterK". Other data has been omitted for clarity, but no similar increase seemed present there at first sight.



    There's also some weirdness in the timestamps. The timezone isn't UTC. I had to apply a 1-hour offset to the radiation testing times.

  • can  

    The time zone for the lab computer was set to CET for some reason in the past, and not noticed until the test had started.

    Rather than change it mid-stream I left it for the entire run.


    Following suggestions made earlier, I did a null run, using a new sheet of printer paper straight from the package.


    The data file is at https://drive.google.com/open?…6nnnF6WHfxEAJq7-3vjPmZVDM


    and the event log file is at https://drive.google.com/open?…BlhWjQtuqlP5CBPHGy0NSUnxj


    I expect some raised eyebrows!

  • It is possible that there are more particles types. So by shielding one type it could reduce count rate noticeable.

    From my experience there is direct relation between Pancake distance and activated material with respect to counts. Yes, I can also measure them from around 1cm distance but putting it closer is really big difference. The best result is in a direct contact.


    But in case of ZnS(Ag) Alpha detector distance is not that important. There the detector has protective mesh as well in a similar distance.

    And as described earlier increased counts are detected even with range of 2m from the cell. But these are not normal Alphas.


    So to surmise it I am convinced there are at least 3 kinds of radiation:

    - Alpha

    - Beta

    - unknown


    Each is behaving little bit differently.

  • Here is the SEM/EDX report from the very first run Titanium Sponge, D2O, LiOD, Nickel. You can make your conclusions.


    Was this EDS from the LiOD only?


    You should enable cadmium, chromium!


    The data shows a lot of carbon because our air is full of CO2 that goes into solution in water forming CO3H- + p+ and in deep PH CO3-2. This already points to the limits of such an open dish experiment. Dust --> silicon is also seen and no interesting elements measured except once Vanadium, that can be expected from Ti.


    I would try to get a good spectrum if the sheet once delivers 3-5 x background. 320keV, 749keV for Vanadium could be interesting.

  • Yes, there can be contamination with Carbon as it is happening anywhere in the air.

    During analysis peaks from Cadmium, Chromium can overlap with other peaks and create false results. This was avoided during measurement.


    The sheet always delivers around 5x background with Ag foil. But as I said many times, there are no emitted Gammas. This mean that Gamma scintillation detector is useless.

  • Hi Alan,

    Your replication was very well done , your having paid much attention to detail.

    Your results may be significant but I await an LiOD run using a fresh Ti cathode.

    An LiOD run using the old hydrogen filled cathode would also be interesting.

    Put the old cathode in a hydrogen atmosphere to stop it outgassing.

    Your pancake detector is unconventionally wired.

    Convention would have pancake body, box, and coax screen bonded with coax

    inner going to GM inner electrode.

    Your electronics driver/amplifyer polarity probably dictated this unconventional

    setup. Can you post a schematic of the electronics ? Pete.

  • I already identified that the cell is not producing any Gammas as far as I can measure. Background from NaI is still more or less constant.
    I already found that these are not high energetic photons with energy > 20keV.


    Pancake detector is much more sensitive for low energetic particles than NaI. Especially it can pick up Beta radiation.

    And my Alpha detector is picking up according datasheet just Alphas while completely rejecting Beta, Gamma.


    Very soft X-rays with energy under 15keV can't be discounted completely yet.


    magicsound I think that the best would be to measure the paper similarly as I am doing (in close contact) to achieve good results first. Then you can try to exclude any possible errors. This would be easier way I think than adding obstacles before achieving a proper results. So you might get first clear readings and then you can identify nature of it.

  • Please explain your meaning.


    I was motivated to perform a control test of the paper by Kirk's comments on paper chemistry. This test seems to have not produced detectable radiation, using the same measurement process as the the "live"test.


    My expectation was that some reviewers would take this as supporting the evidence for enhanced radiation from the "JohnyFive" replication. There's still plenty of room for skepticism, but one well-founded objection, that the paper itself might be active, seems to have been discounted.

  • magicsound You can still use the cell for few more weeks. It should work well until there is enough water.

    Cell with D2O can yield very likely better interesting results but demineralized H2O is working pretty good for me.


    Side note - From my experiments I've found that just current near or less than 1mA is enough in order to generate required particles.

    This is extremely interesting as you could power it from basically any available source if regulated. For example potatoes or lemons..

  • Your pancake detector is unconventionally wired.

    Convention would have pancake body, box, and coax screen bonded with coax

    inner going to GM inner electrode.

    Your electronics driver/amplifyer polarity probably dictated this unconventional

    setup. Can you post a schematic of the electronics ?


    Yes, you are correct. The electronics package is a GQ GMC-320+, a relatively inexpensive unit I selected initially for its USB interface. It's not open-source so the schematic isn't made available. When I adapted it for the LND7317 tube, I found if it was connected conventionally per your comment, it would not work correctly, with very high noise seen at the output of the detector circuit. It took a while to reverse-engineer the circuit, but I finally saw that the front end FET was connected at the low (cathode) side of the tube, with the anode connected directly to the HV supply through a 3 meg ohm resistor. Here's a simplified schematic of my system:



    The cathode was more sensitive to EMF if connected conventionally through the coax shield to the front end amplifier. That is why the coax is wired with the shield on the anode (HV) and the core on the cathode.