ImpulseQt 3.1.2

[Sesselmann]

Impulse Users...

It has been a few weeks since the last update, since I was busy working on other things, but here it is..

ImpulseQT v3.1.2
In this version I have made the following improvements..

The heatmap waterfall spectrum now works much better without slowing down. It now saves and appends the spectrum to json every 30 seconds, while retaining only 1 hour in memory. The screen only shows the last 60 intervals, but I added a new button that opens the complete recording.

The package is also a bit smaller as I excluded some unwanted imports.

I have tested the app on Mac ✅, and Windows 11 ✅, but need someone to test on Windows 10 for me 🙏

PS: This version probably won't work on Windows 7 as it uses PySide6 ❌

https://github.com/ssesselmann/ImpulseQt/releases

Testing stability with a two hour recording of Cs137

[Rob Tayloe]

I have attempted to download the Windows version using several browsers (Brave and Chrome) and the "dot-zip" version is terminated with the message "virus detected". The mac and source code versions download successfully.

I used a macbook pro with safari and was able to download the windows, mac, and source versions successfully. I was able to transfer the pc zip file to a win10 computer and extract the executable. A windows defender scan of the zip file shows a virus (trojan). The same virus warning is given when the exe file is scanned. Scan image shown below -

Screenshot 2025-10-24 091043.jpg (55.71 KiB) Viewed 1097 times

[Rob Tayloe]

I submitted the ImpulseQt.exe file to an on-line virus scanner and only 1 of 69 scanners detected malware. Screen print shown below -

[Sesselmann]

I submitted the ImpulseQt.exe file to an on-line virus scanner and only 1 of 69 scanners detected malware. Screen print shown below -

It is a common problem with apps compiled with PyInstaller that they can triggers a false positive.

I will publish a paid version on the Microsoft store and leave a free Github version for anyone who wants to live dangerously.

It is increadibly complicated and time consuming to set up a MS developer account, spent 8 hours today and still not done, hope it brings in some revenue....

Steven

[Rob Tayloe]

I did install ImpulseQt v3-1-2 on a Win10 pc. I used an older HP mini-PC with an i5 processor. Initially I installed Beqmoni as it seems to install drivers during the program set-up. Then I was able to download ImpulseQt, unzip it, and do the install.

I am using the GS Max 8000 with a 1.5-in by 1.5-in NaI detector which wants 750 V (single wire). With some iteration I settled on -U042, which gave 751 V.

The shape of the pulse with a Cs-137 source looked ok (didn't capture a screen grab). With the 2D tab I started acquiring a spectrum. I am using 2048 channels. The 662 keV peak seemed to be just beyond the last channel. It is not obvious how one adjusts the "gain" or energy per bin. And is there a way to set the lower discriminator so as to avoid the low energy x-ray noise? Perhaps it is a terminology thing. I thought that I'd figured out how to do this with Impulse, but I can't seem to do this with ImpulseQt.

One of your excellent demonstration videos would be most helpful. Thanks much!

[Sesselmann]

Rob,

Good progress..

If your peak is too far to the right you just need to lower the voltage a few steps.

An when you do a pulse view, count the dots from the baseline to the peak (-ris) and dots back to the baseline (-fall).

The Microsoft app store is currently reviewing the app, might take up to three days.

Making videos is way outside my comfort szone, but yes I will do...

Steven

Tab 1

[Rob Tayloe]

Steven,
I'd like to offer some comments / observations that come from using ImpulseQt v3-1-2 with a GS-Max-8000 for a bit.

As mentioned previously, there is no gain control to allow for placement of photopeaks within a desirable bin space. For example, I start with Cs-137 and try a get the 662 keV peak just a bit less than the mid-point. So if I am using 2048 channels for NaI, then I'll try and get the 662 keV peak around bin 800 to 900. Without a gain control, the means of "adjusting" the "scaling factor" is through the high voltage setting. I end up with a high voltage that is a bit different from that used with other hardware / software MCA units. I am still looking into the effect of changing the HV, rather than a gain setting, on things such as linearity. My preference would be to set the HV at some desired setting and vary the gain to get the desired "scaling factor".

At the conclusion of acquisition of a sufficient amount of data, I press the "STOP" button in the 2D Histogram tab. I then save the csv file which has the designated filename (which for me typically includes the detector, source, and perhaps some configuration designation). I will also do a screen capture of the 2D tab page, saved with a similar filename, but "jpg" extension. I can read the ASCII csv file with other software for subsequent processing. I will replace the source with another and revise the designated filename. I then press the "START" button. I have observed that the bin buffer does not always clear. I have taken to pressing STOP after a few seconds, then pressing START again (and get a message regarding over-writing the data). This second START results in a cleared bin buffer.

In the Device Setup tab, the Pulse View button can allow one to view the pulse plot. I have made adjustments to the number of '-fall' samples (with a CsI detector) to fully capture the pulse. I do not see a utility for the Oscilloscope function. All I observe are seemingly random pulses. Without a "trigger" for the oscilloscope to capture a pulse (and how would this differ from Pulse View?), I am not seeing a means of making use of this function. I do miss being able to connect an actual external oscilloscope to the MCA, as one could do with the GS-PRO devices.

I do not see where I could make use of the Waterfall function. If you could offer further explanation, that would be useful.

I have not needed to make use of the Count Rate tab (count rate vs time?). When I did demonstrations for students at the OSU nuclear reactor, I would activate a small sample of Al using the "rabbit" (a pneumatic means for quickly introducing a sample into the operating reactor). I would demonstrate the beta and gamma decay of the Al-28, using a plastic cover over a GM detector. I would have students measure the half-life of the decay, using a stop watch and time the change of the count rate (on the GM detector) to half of some chosen starting value - the results were quite close to the actual Al-28 half-life (~2.24 min). I would then place the Al sample beneath an NaI detector, using a Spectrum Techniques MCA and laptop. The gamma spectrum would be displayed on the screen. Then I would switch the MCA mode into a count rate logger (1 Hz data collection rate). The decay curve would be displayed, which would be a straight line with a log display setting on the Y-axis. That would end my 15 min time allocation and the group of students would move onto the next location (which were top of pool to observe the blue glow [Cherenkov radiation]) and then into the reactor control room. While I enjoyed these demonstrations for groups of students (and high school science teachers in the summer time), I no longer do this. Most of my sources have sufficiently long half lives that I don't see a significant change over a measurement period. Never-the-less I do like this feature, please keep it.

I see that measurement data is also saved in a ImpulseQt-data file with a json extension. I have no means of making use of this data at present. I usually clear this file after a set of measurements (sometimes saving the data in an alternate location).

[Sesselmann]

As mentioned previously, there is no gain control to allow for placement of photopeaks within a desirable bin space.

Actually there is, but it is not so obvious, it is called "max integral" and the command is "-max nnnnn", setting this higher will shift the peak to the left, but should not exceed 25,000.
The default setting for max integral is 18,300, but lowering the voltage is the recommended way to reduce gain.
The GS-MAX-8000 could benefit from a little less gain, more often too high than too low, but that will require a hardware update.

I then press the "START" button. I have observed that the bin buffer does not always clear. I have taken to pressing STOP after a few seconds, then pressing START again (and get a message regarding over-writing the data). This second START results in a cleared bin buffer.

Every time you change a setting in ImpulseQt, it updates global variables and saves the settings file, this includes the filename. The program automatically retrieves the last saved settings when you open the program, this includes all the plots. To start a new plot you only need to enter a new name and click start, no need to clear the memory.

I am not seeing a means of making use of this function. I do miss being able to connect an actual external oscilloscope to the MCA, as one could do with the GS-PRO devices.

Note that pulse mode and oscilloscope function are not software functions these are modes activated by the firmware in the microprocessor inside the GS-MAX (not my software). The technician uses these function to adjust the baseline and pulse length settings before first use. It is not likely that the end user will need to adjust the baseline.

I do not see where I could make use of the Waterfall function. If you could offer further explanation, that would be useful.

Spectrometers are used in many different ways, one client category that might use this function are geologists, they might be surveying an area with a vehicle mounted or drone mounted spectrometer and want to know the ratio of Uranium, Thorium and Potassium, another use case might be for security, monitoring a port or entrance and wanting to know when someone passed with a source.

In any case I hope the function will be useful for someone. I am thinking of enhancing it with a plugin USB - GPS this would allow for geolocation while using a laptop.

I would demonstrate the beta and gamma decay of the Al-28, using a plastic cover over a GM detector.

Yep, that's why the count rate histogram is there, I must enhance it with a log function and I also need to add the averaged count rate as a number, not just a red line ...

I see that measurement data is also saved in a ImpulseQt-data file with a json extension. I have no means of making use of this data at present.

The json file is actually written in NPESv2 format. This is a format created by the open gamma project, the files can be opened in this web based program for further analysis.

https://spectrum.nuclearphoenix.xyz/

This idea came about because there was no standard format for gamma spectrum files (other than csv).


This site also has other interesting information...

https://nuclearphoenix.xyz/

Hope you find it useful.

[Rob Tayloe]

Steven,

Thank you for your replies. I will have to investigate further the use of the json file, but at present my method of doing a screen capture and saving the ASCII format csv file works well for me. I will also play with the max integral setting and see how this can affect the need to alter the HV setting to achieve the desired "scale factor".

I do not reload ImpulseQt v3-1-2 between measurements. After saving the csv file and screen capture, I enter a new filename, replace the source, and hit START. The bin buffer does not always clear. However, my work-around of hitting STOP after a few seconds, then again hitting START, over writes the json file, and restarts counting with a cleared bin buffer. I can continue this simple process easily enough.