Hi Svilen,
Nice results and nice setup:) But I have one suggestion for eventual improviement. I see many DIY shields are 3PI configuration, that means bottom, sidewalls, but no top cover (for easy sample change), and detector from the top. This orientation is exposed to cosmic-rays from above, so background reduction is not optimal. We must know, that cosmic-ray flux is angle dependant: minimum comes from the horizon (angle 0° relative to the horizon), maximum comes from the zenith (angle 90°). So orientation like described above is the worst choice. The amount of vertical cosmic-ray "leak" depends on the solid angle to the sky, which detector "sees". If the hole (well) to the shield is deep and narrow, solid angle will be small (so do cosmic-ray flux), but if the well is wide and shallow, this influence will be higher. The same is valid for the detector position in the well of shield - the higher would it be, more cosmic rays would the detector "see". Try to lay all the setup horizontaly (lean on the desk for 90°) and measure if it is any difference. Of course this position is not always convenient for samples. Optimal would be if you would turn the whole setup for 180° (upside down), but then you must solve problem with holding of detector on place and, the most important, supporting of heavy shield, which must not press with all its weight on the detector. Bottom (now top) should be detacheable, side walls simply fixed to the bottom (on desk , e.g.), sample is then put directly on the detector from above and replace the cover. If you have the possibility to try that, try it and notice if it is any improvement in background reduction.
From high K-40 peak in your background spectra I assume, that you have fairly good shield, or you have much glass in your detector, which is not low-K, or the building material affect this spectra. K is especially high in bricks, made from clay, so in my house, which is made entirely from bricks, the background inside the building is 2-3 times higher than outside and spectral analysis reveals, that K is responsible for that. Because K-40 has so high energy, fighting his line is hard only by making the shield thicker and thicker. It's better to simply escape from it somewhere else;)
Regarding sensitivity: not many people know that you can make a crude caesium "standard" just from common food, which can be found in a market store. Check my post:
viewtopic.php?f=5&t=127&p=1970#p1970
Of course, that specifical activity is only approximate, for the very beginning and orientation "where you are", not to be used as an exact reference for every other sample. And it is valid only for this matrix (dried blueberries), density and sample geometry. But it is some estimation anyway, for the beginning, for not being "totaly lost". For potassium it is quite easy: if you have some common potassium compound (KCl, KOH, K2CO3), specific activity can be simply calculated according to molar chemical composition (K content). The woodash is also interesting stuff for calibration. It has both caesium and potassium (at least in my place). If you make a water extract and dry it (so called "potash"), you reduce the mass by factor of 10 (volume even more) and get decent calibration material, which is water soluble, so you can further dilute it and make lower specific activity of the solution at your wish. With this solution you can "spike" other "clean" matrixes and make other check sources (or "standards", if you want). And this way you can test the capability of your detection system, which is measuring its MDA (Minimum Detectable Activity).
If you will need any ash in the future, I can send you. It has approx 1000-2000 Bq/kg Cs-137 (I can measure more accurate value of each "batch", prior its shipment).
Regards,
Alain