Re: Peak shift in NaI spectrum of an Oxford Instrument MCA

[Chris Alston <alstonc@odrge.odr.georgetown.edu>]

Sir

If I understand correctly, the problem is a "wandering" gain. I.e., it's
not that you're increasing the amp gain by the same amount each day, to
restore the 662 centroid to the channel you've chosen, but that you find it
low, jack up the gain to correct it, but then find, after a couple of
hours, that you've effectively overshot, and now are too high. You don't
say whether or not you have a preamp (maybe an FET) in, or at, the PMT
base. Those things can be a pain, in NaI spectroscopy, because even the
small amount of current generated by counting a not-especially-hot sample
(e.g., a patient's thyroid, in uptake studies), can warm the transistor,
thus changing the gain.

Regardless, my first guess would be a bad capacitor in the HV power supply
to the PMT. I had just this problem with a Canberra MCA, a few years ago.

Then, I hope that you're giving your baby a clean, stable, AC supply.
Feeding it electricity, straight out of the wall socket, is like giving an
infant dirty water to drink. If not, I highly recommend investing in, at
least, a computer-grade line conditioner/filter (e.g., a Tripplite LC1200,
or an APC Line-R 600). We're talking < US$200. Better yet, buy a small
uninterruptible power supply (UPS). I suggest using one that produces a
true sine-wave voltage (e.g., an APC Smart-UPS), rather than one of the
lower-priced units that generates a stepped-wave approximation of same.
You're still looking at < US$300, because you don't need a lot of volt-amps
(VA).

Giving an MCA (or any computer) clean power is really the most basic sort
of maintenance, like keeping it dust-free, dry, etc. It would pay for
itself many times over, particularly if there were to be a surge of
catastrophic (instrument-killing) proportions. In such an event, if the
power protector were not able to limit a spike, a good one would sacrifice
itself, instead of passing it on to the instrument. You could then, likely,
return it (the UPS) to the manufacturer for a replacement. Also, the good
power protectors usually come with insurance policies, such that, if your
instrument were damaged by bad power, while connected to the protecting
device, the manufacturer would pay for repair or replacement, up to a
specified maximum figure. Again, American Power Conversion (APC) is notably
trustworthy, in this regard.

Good luck
cja
alstonc@odrge.odr.georgetown.edu

P.S. Just out of curiosity, how many keV/channel are you using?

At 08:44 PM 9/21/99 -0500, you wrote:
>Ask for advice for an old piece of MCA:
>I'm puzzled what problem is inside my Oxford Instrument MCA (version PCA3).
>Every time when I check the spectrum energy with Cs137(662keV) , the energy
>reads 600~630keV, the same percentage drop happens with channel readings
>too. When the hardware fine gain is increased by 5% to 10%. The peak
>becomes 662keV again but after 2 hours, the peak raises to 680keV. This
>pattern has repeated for many weeks. I have excluded the room temperature
>effect to the device. This MCA does not have any abnormalities in FWHM
>shapes and efficiency counts. What could possibly be the problem inside the
>hardware ?

************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html