Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-L] Half-Life measurement

Probably everybody realizes this, but just to be extra
clear, there are two different conversations in progress:

1) The data is imperfect, so we have to ask:
How could we improve the experiment?

2) Is the data analysis method robust enough
to handle imperfect data?

These are both exceedingly reasonable conversations;
they're just not the same conversation.

On 10/13/21 10:56 AM, Paul Nord wrote:
Students measured 0.3 counts per second for their
background. That's less than half of the 0.7 number you found. So the
shielding is doing something.


Excellent questions.

Or maybe not. My "background radiation" questions were based on a

There are two distinct concepts:
-- Background signal, measured with shielding in place, with
everything in place except the sample.

-- Baseline, measured with the sample in place, at times long
compared to the lifetime of the copper isotopes.

I haven't done a proper quantitative fit, but an eyeball fit
suggests that the baseline is considerably above the background,
up around 0.2 events per second. Or am I mistaken about that???

This is a big deal, because the curve fit routine must cope with
the baseline, no matter where that's coming from, background or
otherwise. And the baseline is high enough to be seriously annoying.

The usual rule is you want the signal to be 10× above baseline,
and we're nowhere near that.

Furthermore, this suggests that fussing over the background is not
a good use of resources at this time. The first order of business
should be to understand the baseline, i.e. the height of the
baseline above the background. I'm mystified by this.

In desperation, let me list some hypotheses, none of which make
any sense, just to indicate how mystified I am:

?? Is some of the ⁶⁶Cu getting promoted to ⁶⁷Cu ??
That has a lifetime of 62 hours, which is long enough to create
a baseline, yet short enough to produce some nonzero activity.
This is nuts because the amount of ⁶⁶Cu is small, and the
neutron flux is not high enough to affect more than a tiny
fraction of it.

?? Is there a tarnish layer of copper sulfide? ³⁵S has a lifetime
of 87 days, which is somewhat too long, but not orders of
magnitude too long.

This is nuts because I assume you would have noticed the tarnish
layer long before now.

???? Have the samples been used so many times that there is a
buildup of Zn in the surface layer? This is nuts for multiple

Do you have a piece of plain old copper that has never been
activated? Does that produce a baseline above background?

Do you have a piece of copper that was activated two months ago
and not since?

Obviously there is some major part of the picture that I don't