38B57 Scionix Technical Data & Calibration

Scintillation crystals, PMTs, voltage dividers etc...
Post Reply
Springdream
Posts: 14
Joined: 14 Nov 2017, 10:32
Contact:

38B57 Scionix Technical Data & Calibration

Post by Springdream » 13 Nov 2018, 08:00

Hello community,

I see that many of you (and myself) are using a 38B57 from Scionix (I got mine from Tom), but there is not too much spec sheets or technical data out there (I'd say none actually).
I am wondering where to get some technical data about it.
Data like
- PMT: do all devices use a Hamamatsu R980 Photomultiplier Tube? Hamatsu Spec sheet (http://dtsheet.com/doc/749852/hamamatsu-r980) shows 350...1250V
- voltage range: experience for good results. I encountered problems when I go over 750V. There shows up a false peak instead of 1460keV...)
- typical max. keV achievable (seems to be a function of the dimension of the scintilator). As the length is 57m(?) it shoud be quite high (>2000keV)
- total efficiency, i.e. number of counts/number of gamma rays passing the scintillator(keV, supply voltage, location of point source). I try to make a calibration for quantitative measurements.

Does anyone has some information available?

Greetings from Germany,
Martin
Last edited by Springdream on 15 Nov 2018, 06:22, edited 1 time in total.

luuk
Posts: 174
Joined: 01 May 2015, 21:00
Location: the Netherlands
Contact:

Re: 38B57 Scionix Technical Data

Post by luuk » 14 Nov 2018, 08:38

Hello Martin,

The 38B57 detectors were made with R580 and R980 pmt’s that are basically the same only the R580 is out of production. There is no certainty to tell what pmt is in your detector but I expect a R580 at that time the most used one.
There is in principal no limit in what you can detect but efficiency for high energies will be low of course, but no problem to detect e.g.Th-228 at 2.5MeV if you take your time.
These pmt’s are 10 stages so most detectors will work with a relative low hv from about 400-700V but it all depends how the detector was used or misused during his productive life.
The original vd was custom made and contains active elements (diodes and transistors) it is not linear special when the hv is raised the linearity is getting worse, you often see these detectors offered with a resolution from 5.6-5.9 % @662KeV that is not the real resolution the practical resolution is >6.0%, most >6.5 % up to 7.5%
If you detector is still good working and linear all depends on how the modification was done on your detector you have to ask Tom how he did the modification on these detectors.
There are no original data sheets from these detectors available because they were tested a long time ago on an old pc system and the hdd is crashed.

Luuk

User avatar
iRad
Posts: 85
Joined: 01 May 2015, 12:27
Location: Stuart, FL USA
Contact:

Re: 38B57 Scionix Technical Data

Post by iRad » 14 Nov 2018, 11:44

Hi Martin / Luuk
for most of the The 38B57 detectors I sell, I strip out all original active components and replace the entire passive resistor network with new metal film SMD parts. Unless another value is specifically requested by a buyer, a 120M network is installed so that the detector will work with a low power device (like a Theremino PMT Adapter and Ludlum survey meters). Three SMD caps are also kept for buffering the last three dynodes.
It is disappointing to see that some sellers use misleading resolution numbers (and other unrealistic statements) when trying to sell their listings, however these detectors do offer very good "realistic" resolution for the price.
Cheers, Tom
Please check out my eBay Store: http://stores.ebay.com/The-Rad-Lab

Springdream
Posts: 14
Joined: 14 Nov 2017, 10:32
Contact:

Re: 38B57 Scionix Technical Data & Calibration

Post by Springdream » 15 Nov 2018, 06:45

Hello Luuk and/or Steven,
regarding supply voltage: it is not that easy as described in other posts here.
At the beginning I made the mistake that the line in voltage of the card (I am using a fireface UC interface that provides superior signal/noise level) was too high/saturated and hence the spectra showed a false peak that basically came from the clipped voltage of the input.
Now I was trying to identify the maximum possible energy with my 38B57 (which can be quite high as it is 57mm long).
I used a Th232 rod sample that goes up to 2600keV.
Unfortunatelly, by matching the peaks I was terribly wrong because of the max. energy is varying a lot as a function of the supply voltage (I did only see a small part of the spectrum).
The higher the voltage the lower the max. energy in the spectra.
I guess reason for that is the limited output voltage of the GS1100 Pro itself.
I added two osci plots that show long term amplitudes ot Cs137 pulses (130 cps) and every few seconds there are some full scale spikes (maximum the line in allows).
@550V those peaks are appr. 8x higher than Cs137, @650V the peaks are only 4x higher. I guess that reflects the increased sensitivity of the PMT by increasing the voltage.
The amplification of the GS1100Pro seems to stay constant and output voltage seems to reach its limit.

Anyway I see no reason for those "max energy spikes" and consider them a flaw, but I don't now if they come from the 38B57 or the GS1100Pro.

At the end the result leads to (false) peaks at the varying max. energy depending on the voltage.

I added a combined spectrum plot with all voltages used (600, 650, 690, 700, 710, 750) and @750V max. energy is already down to 1230keV (obviously far too low for all Th232 peaks)

I did not modify the internal sensitivity screw of the GS1100Pro. Would that help to achieve better energy max at different voltages?
Anyway there is no description and some recommendations would help a lot.

Best,
Martin
Attachments
Spurious Peaks @650V.jpg
Spurious Peaks @650V.jpg (54.59 KiB) Viewed 564 times
Spurious Peaks @550V.jpg
Spurious Peaks @550V.jpg (57.38 KiB) Viewed 564 times
Limit kev vs Voltage part low V.jpg
Limit kev vs Voltage total.jpg

User avatar
Steven Sesselmann
Posts: 513
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: 38B57 Scionix Technical Data & Calibration

Post by Steven Sesselmann » 15 Nov 2018, 07:49

Martin,

Thanks for the questions, not sure if I have interpreted your question correctly, but here goes..

The typical line in audio level is around +- 1.4 volts in most cases, I believe some professional audio equipment may have +- 1.7V, and as we are only using positive/negative pulses we need to make sure our volume is set to take advantage of the dynamic range.

The software (PRA) divides the audio input range into +-100 arbitrary units and sometimes you will see a false peak in the bins just exceeding 100 arb.u., these are usually high energy cosmic rays for which we have no other solution than to accept when they overshoot our scale.

If you can't get your output volume set correctly by tuning the HV bias, then it's perfectly okay to make adjustments to the preamplifier trimmer potentiometer inside the GS-1100-PRO, alternatively play with the microphone volume settings in the PC audio controls.

Pay attention to the large spikes, any signals coming from natural phenomena should be random, but if you at any time start seeing regular interval pulses of similar amplitude , then it might indicate a high voltage arc in a cable of a connector somewhere. These "false " counts tend to be regular as the constant capacitance charges up and discharges.

Your spectra don't look too bad.

Steven
Steven Sesselmann | Sydney | Australia | gammaspectacular.com | groundpotential.org | rephopper.com | beejewel.com.au |

Springdream
Posts: 14
Joined: 14 Nov 2017, 10:32
Contact:

Re: 38B57 Scionix Technical Data & Calibration

Post by Springdream » 16 Nov 2018, 09:18

Hi again,

Ok, again my fault: the spurious peaks resulted from limitations of my sound card input (depending on the gain needed which depends on the voltage) and not the detector or GS1100Pro.
I need to go lower than 650V to get the 2614.8keV peak Ti208 properly with enough headroom to clipping due to max. voltage.
Also I can go down as far as 550V and still get good results. I was not expecting that such low voltages might be useful.

I hope that helps others to avoid confusion.

Best,
Martin
Attachments
Limit kev vs Voltage part low low V.jpg

User avatar
Steven Sesselmann
Posts: 513
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: 38B57 Scionix Technical Data & Calibration

Post by Steven Sesselmann » 16 Nov 2018, 14:35

Martin,

Looks good, glad you got it sorted out.

Steven
Steven Sesselmann | Sydney | Australia | gammaspectacular.com | groundpotential.org | rephopper.com | beejewel.com.au |

Post Reply

Who is online

Users browsing this forum: No registered users and 5 guests