Implications of Signal, Seeing into the Cat with X-Rays - Video from MFMP

HI Thomas, normally I respect your critical approach, especially when you obviously take time to Analyse some results and data and make good arguments. Even if I do not always agree I respect your approach and find your thought through criticisms interesting. I think your point about TC noise sounds interesting and am quite interested what you find about that. However...

In your above statement you mentioned that the curves characteristics are not bremsstrahlung like and even in fact "have completely the wrong characteristics". I'm quite confused by that as it looks very bremsstrahlung like to me for typical emission of beta say with a Q value around 1400 keV, and fermi and plasma frequencies in the metal below the sensitivity of the device? Could you clarify what you meant by that? Perhaps I misunderstood your point as you seemed quite certain in your statement

Also were there any elements present that would have characteristic X-Ray emissions with in the sensitivity of the Scintillator? I don't think many heavy elements were present. I suppose if tungsten or lead or something is In thedevice it could have characteristic X-rays about 67 keV or 80 keV or so. But I think characteristic X-Rays for nickel and aluminum would be below the sensitivity of the Scintillator could you clarify this for me too as I think you were also quite certain about this too as you said it is "obviously wrong"

From the cut off around 50 keV to 80 keV the signal is much more than 6 times. Closer to 50x, I think, but I would need to go back to the data for a better look and more accurate multiple.

Did you catch my Potential damage pun?

Actually what the signal looks like is close to what happens when the attenuation filter is removed. What is interesting is that the low keV cut off is intact, which I would tend to believe would be not be in the case of an internal electrical problem or some sort of channel counting error glitch.

Quote

In your above statement you mentioned that the curves characteristics are not bremsstrahlung like and even in fact have completely the wrong characteristics.. I'm quite confused by that as it looks very bremsstrahlung like to me for typical emission of beta say with a Q value around 1400 keV, and fermi and plasma frequencies in the metal below the sensitivity of the device? Could you clarify what you meant by that? Perhaps I misunderstood your point as you seemed quite certain in your statement

I deleted that comment (sorry, overlapped with yours) because not wanting to get into complication. Inner Bremsstrahlung would maybe fit the curve - external Bremsstrahlung not (it is linear count vs frequency).

Re charactersitic x-rays if they are below the sensitivity then my comment stands, you are thinking I'm saying something that I'm not.

Thanks Thomas for you quick update and reply apparently we had some overlaps with editing

for information here is a simple model plot from wiki of the bremsstrahlung curve.

https://en.m.wikipedia.org/wik…Bremsstrahlung_power2.svg

I like this one better.
https://en.m.wikibooks.org/wik…ia/File%3AXrtSpectrum.jpg

new video, haven't watched yet

@'Paradigmnoia and @'Thomas Clarke',

thanks for for the graph it is also a good one. I think this one however comes from a beam of electrons with fixed energy 100keV. It is indeed linear which is interesting in this case.

In beta decay it is not so simple. In this case the beta electrons are not all not emitted at the maximum Q energy but over a range of energies peaking at a much lower frequency. As some of the energy is taken by the neutrino.

Here is a typical beta emission curve:

https://en.m.wikipedia.org/wik…ay#/media/File%3ARaE1.jpg

so the bremsstrahlung in the case of beta emission is depends on the full range of beta particles energies not just the Q Value

perhaps this leads to the more curved shape in the profile.

@StephenC
The beta emission curve is a Landau distribution. I can't remember if we went over that here.
As I mentioned earlier, you can see the 1.31 MeV area beta hits in the filtered spectra from MFMP. It gets hidden a bit in the natural background bremsstrahlung.
I'll see if I can dig one out in a moment.
Edit: I seem to have deleted the one I had.
Edit2: I might have been mixed up with the Compton edge from the 1461 keV peak. There was however in one of my versions of the MFMP data a small cluster at 1311 keV that might have been from direct hits of betas from the K40 decay. Below is another handy image.
http://nsspi.tamu.edu/nssep/co…amma-ray-spectra-page-two

In astrophysics, the photometric H band, centered at 1.65 µm, falls in a very special place, at or near the flux maximum in the energy distributions of nearly all stars cooler than the Sun. The reason is that this feature is caused by the wavelength dependence of absorption by the negative hydrogen ion (H− , being a proton with two electrons). Photons in the part of the spectrum shortward of 1.65 µm have enough energy to knock one of the electrons from an H− ion, being absorbed in the process while photons at longer wavelengths can be absorbed by a free free process (here a passing electron happens to be close to a neutral hydrogen atom when the photon comes by). The total absorption by H− is the sum of these two processes, which has a minimum value at 1.65 µm.

At high temperatures, the the H- ion decomposes and therefore cannot be a cause in the LENR reaction since reactor meltdowns produce temperatures on the 4000 K range where H- ions would surly decompose.

Stellar Spectra in the H Band
https://www.aavso.org/media/jaavso/2016.pdf

@ Thomas Clarke and others

Cannot many of the issues with ground loops and USB glitches be remedied, or at least altered, by using all and separate battery supply voltages? Just a thought and suggestion, based on long ago experience. Even though that might be somewhat of a pain, it has the promise of removing some of those possible issues from the interpretation. Worth the pain in view of the potential importance (no pun intended there!).

I don't think there is too much involved in getting rid of electrical problems or USB problems, overall.

There unfortunately isn't some library out there of artifacts to look at and see if anything looks like a funny signal one might get from a problem.
Experience helps. Repetition helps.
Wiggling around cords and popping USBs and other cables hot out of ports to see if a problem can be caused can be a rather expensive learning curve, though.

Ground loops particularly may not respond at all to "wiggling" cables. But the low voltage DC of most instrumentation supplies today does allow full float with batteries. Isolation transformers can be used in other circumstances.

Quote from Longview

@ Thomas Clarke and others

Cannot many of the issues with ground loops and USB glitches be remedied, or at least altered, by using all and separate battery supply voltages? Just a thought and suggestion, based on long ago experience. Even though that…

Given that the sense voltage from the scintillator is referenced to ground and sent to a PC ADC I'd reckon ground isolation is important there. You can deal with it by using a separate ADC with an optical interconnect (for example). It is not trivial. Otherwise yes It is sort of important not to have ground loops and isolating all equipment is therefore important. You need separate isolation for each sensor to be safe. Although in theory single-point ground and all else connected direct is OK, practically that still injects some noise.

Personally I've had trouble with this, so although you can get things to work by connecting all grounds together and hoping it is bound to be flakey.

Anyways, I have yet to see an example of an electrical problem injecting something like 140000 to 185000 extra counts smoothly into the low keV end of the spectra but not leak into the cut off zone below 50 keV or make big spikes elsewhere in the spectra.

@Paradigmnoia Thanks a lot for the link regarding the gamma spectra. I like it a lot its a very good summary of the different features of the gamma spectra resulting from gamma emission and I hadn't seen this particular link before.

It is true the beta emission curve looks like a Landau distribution. I suppose for a Bremsstrahlung radiation spectrum from beta decay we would need to integrate the linear electron beam type spectrum over all the energies and intensities of the beta emission spectrum.

In the case of K40 we would get only a few betas with energy around 1311 keV and much more at the lowered of the spectrum with a peak around 100 or 200 keV with an average around 560 keV, so I would not expect a cluster of X-rays around the 1311 keV value from the beta.

I think this would lead to a curve up towards the lower end similar to what has been seen in spectrum 7. It would be interesting to see if someone with better maths skills than me could derive this kind of integration over beta emission energies and see if the curve followed the same rate of increase and over all profile as seen in particular at lower energies.

There may be some interesting deviations at the lower energy. From the beta decay energy graph we see a maximum intensity in the landau type graph at a certain value after which we see it drop quite rapidly as the energy decreases further. From the beta decay example from wiki that I sent I think for a maximum energy in the 1.3 MeV region we would expect a peak intensity around 100 keV or so. I suppose the effect of this on the bremsstrahlung curve would be for the rate of change of increase in intensity to decrease from below the maximum intensity beta energy downwards. Interestingly in the curve for spectrum 7 it almost seems to continue increasing in rate up to the minimum sensitivity value of the spectrum, but maybe any change in rate is difficult to see as it is already increasing very rapidly with lower energy in this region.

I was also curious about the K40 decay… Its one of the few nuclei that have a ground state Q value for beta decay in the 1.3 to 1.4 MeV region. I'm not sure there was any extra Potassium in the device though?

I suppose a similar change in bremsstrahlung curve profile would occur from thermal electrons with a Boltzmann distribution of energies, but I think in this case we would probably see a different profile than that seen at much lower energies and I would be very surprised if we had thermal electrons with a maximum energy of around 1.3 or 1.4 MeV or so.

@Paradigmnoia@Thomas Clarke I'm also curious what Thomas comes up with regarding any possibility of an electrical problem causing the spectra we see. It is certainly true that we seem to see a broad effect over almost the complete spectrum but increasing very much in the low end. Although it looks like Bremsstrahlung of beta to me I think it is certainly very important to be sure there are no other possibilities causing some kind of systematic deviation to the spectra.

@StephenC
The K40 signal is natural terrestrial background. If you see the 1461 keV bump in the spectra, the background has not been completely removed.

The nasty thing with high energy betas is that they need to be stopped in low Z materials, or they will initiate X rays by collision bremmstrahlung. Lead is a bad shield for high energy betas, whereas paper, acrylic, and wood are good.

My understanding is the 1460 keV emission is from Gamma emission following electron capture. Not from Beta decay.

Edit: Ahh it think I understoood your point better. I agree with that it looks like the background is not completely removed.

Its strange though as I think they calculated the spectrum quite accurately based on the duration of the sample and took into account thermal variations of the scintillator in the spectra quite well. Doesn't the spectrum 7 also include the background though? I remember there was a plot somewhere with the background removed but I can't remember if it still showed excess bump around the 1300 to 1500 keV? This would be near the low accuracy part of the curve though due to the lower number of counts at these higher energies.So perhaps the noise at these energies looks like a signal.

I do wonder if the real Q value is lower maybe around 1000 keV or something and we are seeing elevated counts at higher energies for some other reason.

Edit 2: Im also clear about the normal K40 spectrum being background. Its curious though that the apparent Q value for the spectrum 7 seems to be close to that for K40, which made me wonder if there is also some in the device which is somehow being stimulated into beta decay. I don't think there is though. I'm still trying to identify other Beta sources that maybe generated in the device that may have a similar Q values in the 1300 to 1400 keV range.

Internal/Inner Bremsstrahlung may also be a possibility as it also generates a very similar curve to that seen in Spectrum 7 but again it still has maximum energies at the associated Q values for Beta emission or electron capture.

Its an interesting puzzle.

Edit 3: Regarding shielding the Scintillator from Beta to ensure we only get Bremsstrahlung from the device. I have seen some setups mentioned in papers where a magnet is included to deflect the beta away from the sensor. This can be difficult to set up but perhaps if a sufficiently energetic temporary source was used during set up it might be possible. Probably we would want to be careful of the effect of the magnet on the device though. Probably paper or wood is better.

I have fiddled around with the spectra files, even last night a bunch. The background doesn't remove very neatly at all. I tried time periods that were roughly equal to 7 and ones twice as long as background.
I selected and compared ROIs between the cut-off and 69 keV, which is the shoulder of the bismuth peak. The extreme number of anomalous counts is pretty clear when in multiple consecutive background removals to wipe out almost all background (just a few dots here and there remain, and the majority of the channels are empty), there are still around 200000 counts left in the signal area.

Quote from Paradigmnoia

Anyways, I have yet to see an example of an electrical problem injecting something like 140000 to 185000 extra counts smoothly into the low keV end of the spectra but not leak into the cut off zone below 50 keV or make big spikes elsewhere in the spectra.

Not enough info. The software does a lot of filtering followed by pulse counting. The filter (should) be adjusted to the expected tube sensitivity. I have not looked at the VB code - there is quite a lot of it and reverse engineering it would take some time. When I'm next at a loose end...