First HPGE run !

[sgt_bear]

Hi Folks,

late last year i was able to aquire a working HPGE through help of a forum member ;)

It is a Canberra GC1519 which is a "Standart Purpose" HPGE detector that has 15% Efficiency (compared to 3" NaI(Ti)) and has 1.9kev FHMW for Co-60. Also i got a 30L dewar.
Since "homebrew" systems usually are not suitable for HPGE Spectroscopy, i've had to look for another solution. I found some Ortec Spectrum Master 92-X devices on ebay, and found out that these are kind of "Standalone". I asked Ortec for a datasheet - and got one. The device is a Standalone-MCA, so you don't need to have a software to collect spectra, it is done on the microprocessor of the device. I first tried to operate it with "Maestro" but sadly this old device (from 1989) was not supported anymore. Since it can be interfaced over RS232, and even better with plain text commands like "SHOW_STATUS", i've decided to build my own software. Its currently very rudimental, but enough to work with. I can set parameters, start the device, download the spectra, and export it to csv file.
The 92-X was in good condition and worked straight out of the box, the only thing that was an issue is the case fan, it made alot of noise due broken bearings. I removed it and use an older computer fan to blow some air onto the voltage regulators.

Here you can see a screenshot of my interface:

Program.png

Today i got Ln2 and cabled up the device in my workshop. The setup is as following:

- HPGE with Dewar and LN2
- Modified Weightscale under the dewar, lets me estimate remaining LN2 by weight (1L = 0,8kg)
- Ortec 92-X with RS232-USB
- Digital Scope to check the "Amp Out" signal
- Laptop for software

Here you can see my setup:

Setup2.png

I cabled up the detector and made a quick and dirty test run to see if the detector is working. I used a 0.25uCi Cs-137 Source directly on the detector. The Resolution is wonderful and the detector is still in good condition. The Peak does only fluctuate by around 10 Channels. Also i have not enabled any compensation or stabilization on the 92-X yet, maybe i can push the resolution even more by enabling spectrum stabilization.

Spectrum.png

Magnified.png

Last but not least, i've placed a breadboard Oxygen-Sensor in the room that uploads the current oxygen level to a website and has an alarm, so i can make sure the air is safe to breathe.

O2Sensor.png

In the next days i will run a background spectrum collection, a long term Cs-137 run (to see if anything influcences spectrum, eg. 92-X warming up etc), and some of the "usual" spectrums of items that i have. If im satisfied with the results, a shield for the detector will follow!

[Sesselmann]

Jonathan,

Congratulations on the acquisition, nice gear indeed.

What kind of pulse are you getting from the detector, i would not exclude the possibility of using PRA software, the count rate is low, and the resolution is more than good enough.

The pulse may just need a bit of shaping.

Also what is the bias voltage?

Steven

[GigaBecquerel]

I am glad to see the detector up and running, and I'm impressed with your setup!
Now you only need a lead castle for your new toy ;-)

I'm impressed with your MCA setup, and would love to see a more complete documentation of it, if you're willing to share that.

Steven,
HPGes output the typical preamp "sawtooth". While PRA might give a rez similar or better to HPGEs under the right conditions and with a pulser there is no way to get the best rez out of a HPGe with a soundcard MCA. Soundcards just don't allow for the proper shaping times.
Trust me, I played around a lot with my HPGe before I got a "real" MCA.

Lukas

[sgt_bear]

Steven,
The bias is 4500V positive.
The main problem with the signals is, that they are rather high, the small signals are around 2V, and large ones reach up to 10V, so it would need a op-amp circuit with negative gain to supply a soundcard.

Attached is a scope signal of a random pulse:

scope.png

@Lukas,
Now that i know all parts are working, i will spend some time on the MCA interface and then plan to upload on github. Surely a detailed post and some spectrum will follow here too. I can send you the ortec manual if you want

[Sesselmann]

Lucas,

I don't see any problem attenuating the pulse from the HPGE, essentially you could just put a potentiometer inline with the signal and lower it so that the maximum pulses are < 1 volt. Just plug it into the signal port on the GS and that should do it.

Steven

[GigaBecquerel]

@Steven,

There is no issue with attenuating the pulse from a HPGe, and I got very nice specra with soundcard MCAs.
But you cannot go below ~2.5 keV FWHM with a soundcard, because it cannot properly digitize the 4-5 µs FWHM pulses that you need to get the best possible resolution from a HPGe.
It's not about the amplitude, but the shaping time.

@Jonathan
I am looking forward to the documentation!
I don't have a 92-X, so I don't really have a need for the manual, maybe you can just add that to the github?
I have had very different experiences with Ortec, if they allow you to share stuff or not.

[Sesselmann]

@Steven, There is no issue with attenuating the pulse from a HPGe, and I got very nice specra with soundcard MCAs.
But you cannot go below ~2.5 keV FWHM with a soundcard, because it cannot properly digitize the 4-5 µs FWHM pulses that you need to get the best possible resolution from a HPGe.
It's not about the amplitude, but the shaping time.

A raw 5 µs pulse should be just fine, that's about the same as you get from a PMT, once it passes through the GS opamp it comes out the other side at around 100 µs. The only thin g you need to do is attenuate its so that the large pulses don't blow up the opamp. Probably a good idea to have a spare LM6142 opamp or two on hand.

Steven

[luuk]

Hi Jonathan,
That looks great, what beautiful resolution that brings me back to the days when I build and tested HpGe and Ge(Li) detectors , nothing can beat that resolution!
Have fun with your great new toy.
Luuk