Fusione fredda Renzo Mondaini—trascrizione

I just found that the latest glow discharge electrolysis tests with the 0.45mm tungsten wire are causing my DSL internet connection to fail at the router level, probably due to the strong RF interference. The DSL router is two rooms away from the testing location.

Unfortunately, since it's a shared Internet connection, this is going to limit the amount of testing I can do, unless I add a Faraday cage around the experimental setup.

I used 38g tap water + 1.5g acetone + 2 ml 0.1M KOH (pre-prepared) and 750V (open-circuit voltage; it decreases under load but I haven't measured it yet). KOH was added in two 1ml batches and RF noise (audible from loudspeakers I have in my proximity) ramped up substantially after the second one, although the first was apparently already sufficient to cause issues with my router, judging by its event log.

Wyttenbach

Sorry, I don't understand. Very different from what?

For what it's worth, I tried looking at the router log and marking the times when the internet connection failed in the Geiger counter graph below. The events, which are correlated to the testing performed, do not seem to be associated with clear changes in recorded radiation level.

Wyttenbach

Here is a less compressed view of the period between 12:45 and 14:15 UTC, during which I had some DSL failure events following plasma discharge testing with the freshly-arrived tungsten wire. The main issue here is that there were no statistically significant changes, even if some of the events could be speculatively attributed to peaks in the data.

This is all the Geiger data recorded so far.

I cannot afford putting 2.5 k$ into a gamma spectrometer unfortunately, although I do imagine it would be way more sensitive than the GM counter I'm using. Besides, so far these have been ultra-low-budget tests made out of curiosity just to see if anything unusual could be observed with cheap equipment and materials.

EDIT: as an additional data point, below is a graph showing DSL line statistics. Today there were many "synchronization events" in roughly the period of testing. In the past few days I haven't made any testing. Last time I made them was last Sunday (I wrote a report in a previous comment).

Just thought it would be interesting to point out that the as-received tungsten wire looks smooth and dark, while the cathode piece previously immersed in the acetone+KOH aqueous solution looks dull and metal-bright.

Wyttenbach

Those are the normal fluctuations I'm getting with standard background readings at my location. The recorded radiation level is even slightly lowered than usual (roughly about 10 CPM) because I added a small depleted VRLA battery behind the Geiger counter.

I just tried adding some more shielding in the form of steel plates but it did not seem to affect the signal too much (perhaps about 5 CPM? It's the data to the right of the red dashed line in the graph below), although more time is needed to make sure.

In any case, even without shielding, if there is any significant short-term emission from the reaction, it should at least be visible in the data, as the counts would add up to those of the background. That basically nothing is being observed makes me think that either the signal is too small or too brief to be properly recorded by my Geiger counter (or that there is no reaction to be recorded with it).

If there was some sort of guarantee that ionizing emission is being emitted together with RF emission, it would be much easier to just measure the latter—as I've mainly been doing so far—to crudely gauge reaction intensity.

Wyttenbach

About a year and a half ago I asked for a quote by a local vendor for 100x100x50 mm lead bricks and the price was 35 euro/brick, which I deemed too much given that I would have needed 5-6 bricks, at the least.

Steel/iron might be significantly less expensive than copper while only being 10–15% less dense. Perhaps I might end up getting some steel bricks/ingots for shielding purposes, but I highly suspect that it would be wasted monetary effort, given that the rather intense RF produced by the plasma reaction would prevent strong conclusions on the origin of any observed signal.

This being said, I tried adding some more shielding material in the form of obsolete computer equipment (steel, etc), and this is as low as I can get with reasonable effort. The vertical lines denote when I added shielding (I did it in two batches):

As for the actual tests, now I have to wait for the next occasion when I can potentially cause issues with nearby equipment without complaints by other people.

Wyttenbach

Getting lead sheet and rolling it into a cylinder around the Geiger tube might be simpler for "pancake" -type GM tubes, but not with cylindrical GM tubes like the one I have.

You meant doing something along these lines, correct?

I made a brief test using previously prepared electrolyte solution but this time RF noise did not seem to be particularly intense, judging by the response from nearby loudspeakers. I am wondering if the several Kg (25–30) of computer parts put in front of the setup are also shielding the RF emissions to some extent. They still caused disturbances strong enough to affect my DSL internet connection in some way, however ("unrecoverable errors").

There did not seem to be strong peaks standing out from the average in Geiger data. The colored sections denote the testing periods:

Quote from Wyttenbach

You can put the entire board inside. You can even flatten the lead cylinder a bit if you want to save space!

At that point I think it would not be wrapped particularly efficiently. A "sarcophagus" made of steel bricks/bars might do a better job and easily allow to add several centimeters of shielding material without too much monetary commitment. I'm still not sure if it would be worth the expense, however, given the results observed.

In the image below the bricks are 2.5 cm thick.

I found a more detailed way to assess the effect on my DSL connection of the plasma reaction. It looks like it might not be simply affecting the router, but the phone line itself.

When the plasma reaction is on and producing significant noise, there's an increase in noise mainly in the 22–30 MHz DSL band (more information here), which causes a loss of SNR. This is visible in the SNR graph provided by my router. See the red line in the upper graph here, showing the minimum SNR values reached:

This was the RF spectrum in the 24–1000 MHz range measured with a whip antenna close to the jar, first extended to 30 cm, then fully extended. I think the antenna response plays a significant role in the overall signal shape, so it's difficult to say if the peaks observed, in isolation, could be interesting:

Still no peaks standing out from the Geiger counter. I forgot to take note of all the short testing sessions here:

Alan Smith

I'm using the software-defined USB RF receiver (Nooelec NESDR SMArt V4) at the minimum gain setting and the signal is already sufficiently strong to almost saturate it at that range, at least at relatively close distances. The dynamic range with it is only about 35–40 dB, far from that of a professional/stand-alone RF spectrum analyzer, and the signal will not get much higher on the Y axis than what is observed in the screenshots posted.

Will keep in mind about the tip when I'll try other antenna types in the future, though.

Another very brief test using new/clean electrolyte solution (46 ml tap water, 1.55g acetone, 1 ml 0.1M KOH) and a new 0.45mm tungsten cathode wire with 750V (OCV) immediately caused my DSL connection to fail from the EM noise emitted by the plasma or induced/injected in electric conductors nearby. A second, less intensive attempt repeated briefly after that allowed to capture the effect on the connection/line signal-noise-ratio, which again seemed to affect mostly the 22 MHz and upwards range. It seems repeatable.

Calculating the max-min SNR difference (after digitizing the data with WebPlotDigitizer) in the spectrum provided by the router immediately before/after the plasma reaction should give the spectrum of the RF signal affecting the line, which is what I tried in the second graph below:

I just wonder what would be the result with a more serious power supply with similar characteristics; the reaction could end up causing significant disruption also on equipment in nearby buildings.

For what it's worth, to make sure that the reaction was not solely caused by the 12V ATX power supply I used for powering the high-voltage converter (Antec EA380, a basic but otherwise standard PC power supply providing stable 12V), I tried powering up the converter with a 12V VRLA battery and it seems that much of the noise is still present, although at a lower intensity. I did not measure whether battery voltage sagged under load.

Noise was still induced in the telephone/DSL line, although not as much as before, and it still affected mostly frequencies above 20–22 MHz (this time >24 MHz). However it was a brief test.

The SNR min-max difference from the DSL router was actually only loosely similar to the actual RF spectrum measured (in the 24–85 MHz range; gray line) in proximity of the setup with a whip antenna extended to about 35 cm. RF noise intensity seemed still large, overall, but nearby speakers did not seem to get affected noticeably as with earlier tests.

It's possible that a large portion of the noise observed with a power supply connected to the mains is actually injectedinto it (and/or the power supply) and not just directly due to RF emissions. A while back I tried measuring with a clamp meter DC input power into the HV converter, and when the plasma reaction is producing strong RF emission, it shows highly oscillating values that the clamp meter sees as some sort of AC.

Quote from can

It's possible that a large portion of the noise observed with a power supply connected to the mains is actually injected into it (and/or the power supply) and not just directly due to RF emissions. A while back I tried measuring with a clamp meter DC input power into the HV converter, and when the plasma reaction is producing strong RF emission, it shows highly oscillating values that the clamp meter sees as some sort of AC.

I made another test with the 12V battery, this time connecting a multimeter (the multimeter function of the same clamp meter of the test quoted above) to its +/- terminals. It appears that battery voltage fluctuates between 12V and >20V when the strongly RF-emitting plasma reaction occurs, so I might have actually been correct in that noise could have been injected into the mains. The previously used PC power supply might have been sinking power from the HV converter! (or: )

Whether this is actually an anomalous effect—even though something along these lines has been suggested by other groups—is a different story.

The screenshot below shows the highest voltage peak I saw on the battery terminals.

Curiously, at some point (after I inadvertently eroded the tungsten cathode, turning the solution cloudy from tungsten residues) it became considerably more difficult to cause the same reaction, but the effect was renewed after adding another 1 ml 0.1M KOH solution.

The disturbance recorded by the DSL modem/router was higher than earlier, but still not as much as with the PC power supply. I did not/forgot to make RF measurements with the radio receiver.

Below is a video of the multimeter from which I took the screenshot above, during the RF-emitting plasma reaction.