Dear Hamed Sabouri ,

you are right, there is a lot of noise, which shows up as an increase of counts when going to the low channnels.

But I can reassure you, there is no real additional peak.

The step*) in the counts at ch120 is due to the setting of your low level discriminator. The region up to ch120 is intentionally set to 'zero' in order to keep the dead time of the dataprocessing low.

Removing the noise in that region is not an easy task.

a) First of all you should ground all your equipment which is involved in your set up (such as: HPGe-detector, its power supply, high-voltage generator, computer system (incl. monitor etc); LN-liquid level detector). But please ground them at the same point. Avoid any circles/loops in the net of grounding lines...

b) Furthermore please connect all the power cables of your components to the very same phase of your mains power supply line and use a filter in that power line.

c) Additional EMI shielding of the detector window by a thin Al (household) foil can also be helpfull...

*) you mentioned that that 'peak' position is independent on the gain setting; it is just the LLD setting of the amplified signal...

Good luck and

best regards

G.M.

Dear Martens,

I am grateful for your guidance and will adhere to the aforementioned steps. How should LLD and gain settings be done in the respective devices. Do you think my settings are correct? Because to improve the signal-to-noise ratio, the optimal amount of these quantities should be applied.

Dear Hamed Sabouri ,

thanks for feedback...

your LLD and gain settings do not affect the noise level in the spectrum.

The LLD setting governs how much channels are set to zero at the beginning of the spectrum and the amplifier gain governs the position of the 241Am peak.

The noise is due to the setup and its receptivity with respect to any EMI from the neighborhood.

Let me add some further precautions:

Let the signal line from the HPGe output be as short as possible, having the right Ohm value of that cable according to the HPGe and amplifier specifications.

Please keep a large distance of that cable to any power line; especially to high power lines (mains) and motor lines...

Remove any defect arcing Neon lamps from your lab and any the neighboring rooms...

Remark I: I wonder why the onset of the (noise) background starts at the Am241 peak when going the low channels....

Thus a question:

is there any significant amount*) of low atomic number material in the beam just in front of the detector window or just in front (or close to) of the Am241 source; you might suffer a bit from Compton scatter in such a case...

Remark II: I wonder why you have such a 'poor' spectral resolution; it is about 2.2keV; quite a lot...

Question: Which shaping time do you use and what is the current countrate?

*) other than simple air

Best regards

G.M.

Perhaps you should take natural background without the source for reference. Also, note there are lower energy x-rays coming from Am241 ...

Dear Martens.

Thank you I will comply with the said items.

From my understanding, increasing the gain may result in overdrive in the detection system. This is why I suggested there might be an optimal value for the gain setting.

Dear Pavlov

I am aiming to reduce the generated noise. I also obtained the background spectrum, which was particularly strong. It would be reasonable to expect that X-rays from americium would form distinct peaks, but in my spectrum, there appears to be a growing continuum to small channels.

Dear Hamed Sabouri ,

with increasing gain you will also increase the noise. Thus you will have a very lot of counts to be processed and the datacquisition will go into saturation....

Thus I will repeate my question: what is the countrate in counts/sec of your current set up; it might be too high.

You may bring down the countrate (intensity) by using larger distance between the source and the detector...

Dear Nikolay Pavlov ,

taking the background is a good idea.

As I deduce from the previous answer of Hamed Sabouri the background spectrum was 'particularly strong'. So there will be some pick up of EMI.

With respect to the contribution of other Am241 lines in the spectrum, please have a look at a spectrum of an Am241 source, which I have taken long time ago (see the attachment).

The other lines are very tiny and can only be seen when the y-axis is appropriately scaled.

Thus, from my point of view, there is only a very tiny contribution of the low energy Am241 lines to the background of Hamed Sabouri 's spectrum.

Best regards

G.M.

Gerhard Martens , Hamed Sabouri

I would expect the typical EMI would more likely to smear existing peaks than to add continuum of counts.

The spectra gradually growing toward lower energies and limited by threshold looks more like natural gamma background. That background dominates if source is weak.

One thing to check is if the shaping time and gain is set properly, one can simply try alternative settings.

To note, the lower x-ray lines of Am-241 might be quite visible - e.g. in Si detectors (low absorption at 59.5keV of course) with minimal entrance window (so welcoming all low energy x-rays) the Am-241 may look different (pic on the link below).

https://www.researchgate.net/profile/Oleksandr-Starodubtsev/publication/228567911/figure/fig1/AS:393601596051483@1470853341762/Am-gamma-spectrum-obtained-with-the-small-4-mm-2-silicon-pad-detector-at-room-temperature_W640.jpg

Strangely, your spectra looks more like CdZnTe spectra suffering from a bit incomplete charge collection /escape ( the main peak is more stretched to the left):

https://www.researchgate.net/profile/Matthew-Veale/publication/260936019/figure/fig1/AS:296789937803268@1447771643978/241-Am-595-keV-spectrum-at-100-V-10-20-30-V_W640.jpg

Dear Hamed Sabouri ,

the term 'incomplete charge collection' mentioned by Nikolay Pavlov, triggers the idea of a non optimal application of the High Voltage (HV) level onto the HPGe detector in your case. You should increase the high voltage in small steps (e.g. a few 50 to 100V steps; but do not stress the detector too much) and see what's happening with your background and the width of the 60keV peak...

Best regards

G.M.

I see some tiny peaks at around channel 950 ... Can it be some ~84..90 keV characteristic X-rays from the lead shield, if it is used?

Here is some example https://www.researchgate.net/publication/312348866/figure/fig3/AS:11431281208799805@1701442457548/XRF-spectra-used-in-the-experiment_W640.jpg

dear Gerhard Martens

I have used a 1 micro-Ci source placed directly on the detector, equipped with an aluminum window, and recorded a counting rate of 1800 counts per second. The sensitive part of the detector is enclosed in a two-layer housing made of lead and copper.

The EMI is undeniable, and I intend to address it using the mentioned techniques. The spectrum you provided is very suitable and exhibits minimal noise. I am looking for a settings that allows me to capture the spectrum with a resolution comparable to yours.

Despite the high noise levels, my spectrum shows a significant full width at half maximum (FWHM) of around 5keV for the main peak. Additionally, any peaks from americium that may be present in the spectrum are not discernible due to this noise.

Do you mean by incomplete charge collection, not applying working voltage to the detector?

Dear Hamed Sabouri ,

thanks for feedback.

You still apply voltage to the detector; otherwise you would not see any peak; but the voltage may be too small.

Please have a look at the manual and/or the specifications of the detector for the HV value. Such values are in the range of about 1500 to 2000V.

So, please increase your HV in a few 50V steps, and see whether the background gets smaller and the 60keV peak gets narrower.

With regard to the peak width: this quite large width may be due to pile up effect with respect to the large noise peak, which you have cut via the LLD.

Regarding the background: your source is very close to your detector and thus a lot of photons do not enter the detector crystal at about normal incidence. Please keep some distance to ensure nearly normal incidence for all of the incoming photons. In addition, please make use of a (Pb or W) pin-hole diaphragm in front of the detector in order to ensure that the rims of the detector crystal do not see any photons. Here, as well as at the detector surface, most propably, one will have charge sharing and subsequent insufficient charge collection due to the inhomogenous electric field there.

Such experiments can be simply done...

Good luck and

best regards

G.M.

Dear Gerhard Martens

Thank you for your reply.

Your comments were helpful.

I will apply.

Dear Nikolay Pavlov

Thank you for your attention.

The sensitive part of the detector is located inside the lead shield, and the peaks observed in the channel range of 950 must be due to this case. On the other hand, due to the protection, the noise contribution should be low, but apparently it is not. To improve the spectrum, I believe we should minimize the EMI from the peripheral equipment of the detector using the techniques suggested by Gerhard Martens. Additionally, the gain settings need to be properly adjusted, although I'm not sure how to achieve that.

You’re correct that my spectrum closely resembles that of CdZnTe. In your opinion, could this similarity be related to the application of improper working voltage?

Hamed Sabouri

>>.. correct that my spectrum closely resembles that of CdZnTe. In your opinion, could this similarity be related to the application of improper working voltage?

I guess so, i suspect that might be the bias voltage (too low) or shaping time (too fast) but i am just guessing, i did not have much experience with Ge detectors.