Bad 3x3? or is it just me?

[Mike S]

Hello all!

I've had my share of issues and questions with detectors, as detailed in my other threads. I've been trying to find "the one" detector that will be the "flagship" for my gamma spectrometry work and the one my "final" shield will be optimized around.

I thought I had found it, a new old-stock 3x3" Bicron at a price I could afford. It arrived shiny, well packed, and had a presence that meant business. I had it on my desk for several weeks as I first selected, and then waited for a voltage divider socket. At last the two met, and I excitedly watched the spectrum develop, and it was...

Garbage. At least compared to my other detectors, one of which (as detailed in another thread) is pretty beat up. Compared to my expectations, it takes more voltage and gain, has a lower count rate, and has very (very) wide peaks.

Since I had a new detector and a new divider base, I wasn't sure where the problem might be. I was able to borrow a different PMT base and that gives similar results, so I'm thinking it's the detector.

Here's a comparison of the same sample (a relatively hot piece of Trinitite) done with 3 detectors. The tests aren't consistent in shielding or length, but they are pretty typical for each detector:

Red: The brand new 3x3" detector
Blue: Old and degraded 5x4" well detector (The crystal is cloudy and has at least one crack and the light pipe is partially delaminated from the crystal).
Orange: A well used Ludlum 44-10 2x2" detector (old and well used enough that that you can actually feel something shift around inside as you move it side to side).

Is there anything else I should check?

• I feel like I've eliminated cables by swapping them between detectors.

• I'm using Theremino MCA and the pulses look OK to me, though there's some noise (probably due to the increased gain).

• I feel like the voltage divider is eliminated due to the substitution.

• I've tried higher and lower voltages (+/-100V from 1050V, which seems optimal)

I contacted the seller and they said that I should make sure I'm using a voltage divider setup for 10-stage PMTs and not 8-stage. I believe that 10-stage is "normal" and most dividers would be setup for that configuration. They said they've sold 30 of these with no reported problems. Did anyone here buy one? How did it work out for you?

Is there some magical setting for the Theremino MCA baseline test parameters that will improve resolution?

Do detectors like this typically have magnetic shielding built in?

Do I need to ground the case somehow?

Have you seen abnormally poor resolution before? Did you figure out the cause?

Are my expectations too high? Is this typical for a 3x3 detector?

Thanks in advance for any help or advice.

Mike S.

[Sesselmann]

Mike,

Just some notes on your post..

Do detectors like this typically have magnetic shielding built in?

Yes, the metal housing on that detector is actually made from Mu metal, so it is itself the shield.

Do I need to ground the case somehow?

Yes, check that the ground pin is connected to the case, else you will get terrible noise, also make sure that the crystal housing (bottom part) is grounded to the shield.

Have you seen abnormally poor resolution before? Did you figure out the cause?

The resolution of a bad NaI(Tl) detector rarely goes below 13%, and yes I have experienced the grounding problem with removable sockets.

Are my expectations too high? Is this typical for a 3x3 detector?

it looks like a nice detector from the outside, but you never know what's inside, don't write it off just yet.

You mention using a hot sample, are you watching your count rate, remember a 3" has a lot more pulses. try moving the source further away.

Steven

[Mike S]

Steven,

"Yes, check that the ground pin is connected to the case, else you will get terrible noise, also make sure that the crystal housing (bottom part) is grounded to the shield."

The voltage divider cap is properly grounded to itself: ground pin on the circuit board is connected to a tab on the body of the connectors and pin 13 on the socket.

If I understand correctly, it sounds like I should somehow provide a signal ground for the detector case. Is there an elegant way to do this? I'm having visions of something involving a hose clamp, since I don't see any other way to connect to the detector body. Neither voltage divider base I have makes a ground connection to the detector body.

--Mike

[cicastol]

Hi Mike,
i have an identical spare 3x3 detector from Bicron, first of all you need a VERY good shield to use this detector with sound card based software MCA like PRA or Theremino, because count rate is reallly too high with such big detectors, IMHO at least 30:1 or better background unshielded/shielded ratio is needed.

Resolution for 3M3 /3 detector was declared by Bicron for 7,5% or better , mine reached 7,3% but with a not so well matched (for my setup) Canberra 2007P base preamp because my main voltage divider is setup for a non standard 9 stage PMT (Luuk 3x3 detector 6,3% res.), probably my Bicron 3x3 with a proper voltage divider could reach easily 7% or even better.

P.S. 3x3 detectors were the standard for gamma spectrometry for many good reasons! :-)

[Mike S]

Hi Mike,
i have an identical spare 3x3 detector from Bicron, first of all you need a VERY good shield to use this detector with sound card based software MCA like PRA or Theremino, because count rate is reallly too high with such big detectors, IMHO at least 30:1 or better background unshielded/shielded ratio is needed.

Resolution for 3M3 /3 detector was declared by Bicron for 7,5% or better , mine reached 7,3% but with a not so well matched (for my setup) Canberra 2007P base preamp because my main voltage divider is setup for a non standard 9 stage PMT (Luuk 3x3 detector 6,3% res.), probably my Bicron 3x3 with a proper voltage divider could reach easily 7% or even better.

P.S. 3x3 detectors were the standard for gamma spectrometry for many good reasons! :-)

Ciro,

I'm getting fewer counts than I expect, so I don't think lack of shielding is the issue. My shielding is about 10:1. The count rate with the sample in the example shown in my first post was only 65 cps, which should be well within the capability of the GS-USB-PRO and Theremino MCA.

What voltage do you use with your 3x3?

Any ideas why I'm getting such poor resolution and low counts?

Thanks!

Mike S.

[iRad]

The voltage divider cap is properly grounded to itself: ground pin on the circuit board is connected to a tab on the body of the connectors and pin 13 on the socket

Mike - This could be your problem if your statement is correct. Pin 13 is not "G" for ground, but "G" for GRID. Pin 14 is ground, and should be connected to ground and if possible to your detector housing/mu-shield for best performance.

Please do not use a hose clamp to connect the base to the housing. Use metal tape to make the connection. Alternatively, just use some thin bus wire to span across the parts and then tape over with 3M electrical tape to secure in place.

[Mike S]

The voltage divider cap is properly grounded to itself: ground pin on the circuit board is connected to a tab on the body of the connectors and pin 13 on the socket

Mike - This could be your problem if your statement is correct. Pin 13 is not "G" for ground, but "G" for GRID. Pin 14 is ground, and should be connected to ground and if possible to your detector housing/mu-shield for best performance.

Please do not use a hose clamp to connect the base to the housing. Use metal tape to make the connection. Alternatively, just use some thin bus wire to span across the parts and then tape over with 3M electrical tape to secure in place.

Sorry, I should have said pin 14. My base has Pin 13 and 14 connected together and both are connected to ground.

The base I borrowed has a focus pot (presumably, it's unlabeled) and a variable resistance between the grid and ground.

i'll rig something for a ground connection to to the detector body.

Thanks for the help!

[iRad]

Mike - Pin 13 should have NO CONNECTION to GROUND.
There should be a resistor from pin 14 to 13 and then 13 to pin 1. The resistors from pin 14 to 13 should be approximately double the resistance of the resistors in the rest of the chain (from G to pin 1... thru anode).

[Mike S]

Mike - Pin 13 should have NO CONNECTION to GROUND.
There should be a resistor from pin 14 to 13 and then 13 to pin 1. The resistors from pin 14 to 13 and 13 to pin 1 should be approximately double the resistance of the resistors in the rest of the chain (from pin 1 thru anode).

The grid is definitely connected to ground in my divider base. Looking at schematics for dividers with focus adjustments, it seems like the range could go between ground and dynode 1 voltage. Would you expect to see the performance issues I'm seeing with the current grounded grid configuration?

The design of the base has 2R between the cathode and dynode 1. I could fairly easily connect the grid between those two resistors so it was at a potential midway between them. In doing so I'd only have 1R between ground and grid and 1R between grid and dynode 1. Would that likely work for testing? Anything beyond that would require more extensive work (which is OK, I would just like to know).

Thanks for your help!

--Mike

[cicastol]

Any ideas why I'm getting such poor resolution and low counts?

Thanks!

Mike S.

With a proper working 3x3 you should expect at least 500cps in normal background, as Tom said check the VD pinout , for sure the is something wrong somewhere !
Voltage for the above spec was set at 800V