Still not desired resolution (new scint from GEO)

This area is for discussions involving any fusion related radiation metrology issues. Neutrons are the key signature of fusion, but other radiations are of interest to the amateur fusioneer as well.
myID
Posts: 139
Joined: Wed Jun 25, 2008 6:16 pm
Real name:

Still not desired resolution (new scint from GEO)

Post by myID » Wed Jun 25, 2008 6:41 pm

Hello-

first a big thank you to Geo: had trouble with german customs when I ordered the scint from him and he was VERY helful!
So this is the device I am using with my MCA:
http://cgi.ebay.com/3-NaI-Tl-sodium-Iod ... 0069886481
My problem now is: I still do not have the spectrum I d like to see. I attached some pictures. I measured a radium source (old watch), my settings were:
850V, LLD almost 0, amplification really low, 1024 channels, no offset
I used a rolled 4.6 nF cap and 990k for decoupling (see drawing)
Now:
1. is this the max I could get from this scint when it has a bad resolution (I know- squared ones not too suited...)
2. are cap and resistor value OK?
3. Would shielding make a big change?
4. perhaps settings in general wrong?

Thanks- any help would be great because it really gets frustrating....
Attachments
CCF06252008_00001.jpg
CCF06252008_00001.jpg (41.63 KiB) Viewed 4021 times
CCF06252008_00000.jpg
CCF06252008_00000.jpg (70.76 KiB) Viewed 4021 times
cap and resistor.JPG
cap and resistor.JPG (22.77 KiB) Viewed 4021 times
square pmt 2 radium.JPG
square pmt 2 radium.JPG (32.6 KiB) Viewed 4021 times

User avatar
Richard Hull
Moderator
Posts: 12851
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Still not desired resolution (new scint from GEO)

Post by Richard Hull » Wed Jun 25, 2008 8:17 pm

I can see what looks like goldilocks and the three bears and the Bi214, (radium C), peak near 609kev. The low energy separation 180kev - 400kev doesn't look all that great. At least it does resemble a real radium gamma spectrum.

The calibration would demand a pulser to allow you to set the gain though the 609 peak by itself would help

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 11:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Still not desired resolution (new scint from GEO)

Post by Carl Willis » Wed Jun 25, 2008 8:20 pm

Roman,

These things are impossible to troubleshoot effectively from a distance. I get the impression that the best advice is for you to become familiar with the principles of your equipment so that you'll know what's going on. A good book like Glenn Knoll's "Radiation Detection and Measurement" will be enormously beneficial. You have a bench full of useful, high quality test gear and are inherently in a better position to determine your cause of difficulty than some people just looking at a couple photos, so it would make the most sense if you went ahead and learned some useful background on what you're doing.

Anyway, for what it's worth, here are a few things to check.

1. The PMT should be supplied with positive HV. Last we heard from you, your power supplies were said to be negative supplies.

2. The output pulse will be a negative pulse if the HV supply is a positive supply like it should be. Can your MCA's amplifier handle this? If not, the system is just analyzing the "undershoot" from the AC coupling, which will indeed give a pulse-height spectrum albeit much degraded from the ideal.

3. The last advice you were given included using a pulser to check the MCA. That is a very good idea. With it, you could eliminate the electronics as the cause of poor resolution in one simple experiment.

4. How long is the PMT cable, and how long is the cable from your capacitor to the MCA? Both should be very short. No 50' rolls of cable right from the store.

5. The values of R and C look usable to me. C should ideally be silver mica or an NPO ceramic cap. In the interests of eliminating noise, these components should be placed in a metal project box. You can also place a filter capacitor across the HV input to this circuit.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277

myID
Posts: 139
Joined: Wed Jun 25, 2008 6:16 pm
Real name:

Re: Still not desired resolution (new scint from GEO)

Post by myID » Thu Jun 26, 2008 8:29 am

Hello-

thank you very much for again looking at my outcome.
I guess I have a basic knowledge on what is going on (confused that supply thing...)
My supply is POSITIVE, the pmt is set up for that, cap is highest quality I had because I do not want it to eat my tiny charges, resistors are not metallic film so kind of noisy but the only ones I had on hand....
I used short high quality BNC cables (and did not connect my multimeter like on the picture) I will order better parts now because I guess the values are ok and should not make that big of a difference (from my understanding the cap is definitely big enough because I only need a small amplification to have counts on all channels?)
I understood that a pulser would be a wonderfull tool- obviousely- but getting this stuff over here in germany is expensive and time consuming. I am building the photodiode pulser described in the files section but I guess it is not really suited and I am not sure how to modify it- just use a variable voltage devider at its output to check different channels??
Sure I understood a shielded case would be better because we are talking about small fast pulses but I was wondering if it lowers resolution in general (from my understanding I would not expect this too much).
One big problem indeed is documentation- I still do not have a manual for my MCA and figured out how to use it so far by experiments. I THINK the input amplifier inverts the signals when set correctly and "looks" at negative pulses and my experiments so far seem to proove that.....
Looking at Richards "bears" (even though I know he is probbably one of the most experienced persons in this forum...) I was a little diappointed about "all bears in one" and was wondering if there would be something simple I did not try or look into...
Thanks for the book tip- I will try to get it somehow because mo knowledge mo better

Greets
Roman

homebrew
Posts: 4
Joined: Tue Apr 14, 2009 12:00 am
Real name:

Re: Still not desired resolution (new scint from GEO)

Post by homebrew » Tue Apr 14, 2009 1:00 am

hi Roman,

I bought several of those same 3 inch Bicron detectors and found out
the hard way what was wrong. first thing is that you need a dual-tapered
divider rather than the one supplied. The divider as supplied causes
compression of pulse height vs gamma energy because the last few
stages have too low a voltage to overcome space charge. Because
NaI has a decay time longer than typical pmt interelectrode transit time,
space charge limiting can be a problem as energy increases if voltages
are too low. The original design I think starves the last few stages and
especially the anode in an attempt to get semi-constant pulse amplitude
for counting applications. At the other end (cathode) the voltage is also
too low for good resolution.

My cure was to 1) cut the metal tube 2.75 inch from the cable/pot end -
(the dead end) and 2)replace the divider resistors with 1/4 watt resistors
as follows starting at the photocathode: 5.1,3.3,2.2,1.5,1,1,1,1,1,1.5,2.2,
and finally 3.3 Meg. That final resistor is from last dynode to HV, and 3)
loose the cable capacitance by putting a buffer or preamp right at the
assembly. The final 3.3 Meg to HV needs to have 1000p-0.01 u added
across it. Leave the other three original caps in place across the last few
resistors. When all is done, the last four divider resistors will have caps
across them. Also leave the original 4.7 u supply bypass. Loose the sens
pot or leave it tuned full in direction that puts it at a dead short = best
sensitivity. (cw??) If you have to leave the pot there put a 0.01u cap
across it that can withstand the voltage.

My load was 1.5 Meg anode to HV w 8pf in parallel to bring the total cload
to 12 pf ,3300pf 1.5 kv from anode to preamp input.

My preamp was an LF356 with: 330k from 3300pf to -in, 1 Meg from out to
-in, 6pf from out to -in, and +in went to a pair of 2.2k divider resistors off +12
and gnd and a 0.01u to gnd. Lf356 powered by +12 with output leading to
MCA (4.25 v full scale) thru 0.1u. Be sure to bypass power on the LF356.

HV was +510 and scaling was 5.5 Mev = 4.25 volt. Values above will give
a nice gaussian positive output pulse to MCA with 2us risetime and 3.9 us
time to peak value and about 10-20us decay time.

Compression went away and PHR at 1.45 Mev was very good - as I
remember it ended up at 5.5% . This is now the best gamma det I have
and I especially like the 3 inch depth for stopping power but the smaller
dimensions the other way for fewer backround counts.

wk

User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 11:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Still not desired resolution (new scint from GEO)

Post by Carl Willis » Tue Apr 14, 2009 4:59 am

Hello wk,

Good information.

I recently replaced the dead factory crystal in a TSA-460 portable MCA with one of these portal monitor 3" jobs from Geo, complete with its as-supplied dynode chain minus the gain pot, and I have noticeable compression at the high-energy end of the pulse-height spectrum. I don't care too much about it, because with only a 250-channel memory, there's limited ability for quantitative energy measurement on those old units anyway.

But if I had to suggest a solution to improve the linearity, it would be to decrease the resistance of the entire chain while holding the ratios roughly constant and adding more capacitance on the last couple stages. The problem with that in my application would be an increased current demand from the battery-hungry HV supply. So I may reconsider your re-proportioning idea.

I don't do too much with surface-mount parts. Do you know if the parts on that dynode chain are "805s" or some other series? If you know what size they are then I'll have an easier time ordering replacements. Thanks for your contribution.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277

homebrew
Posts: 4
Joined: Tue Apr 14, 2009 12:00 am
Real name:

Re: Still not desired resolution (new scint from GEO)

Post by homebrew » Tue Apr 14, 2009 9:16 am

hi Carl,

I think there were two sizes, 805 and another. I bought surface mount
resistors but it was so tight I could not use them and I didn't want to
unsolder the board from the pmt leads. I ended up using 1/4 watt
leaded resistors standing up. good news was that I kept the center
of the string at 1 meg so 5 of the resistors could stay the original
surf mounts that were on the board, and only 7 had to be removed
and replaced w leaded parts. As I mentioned previous msg I think
I added 1 more string bypass so that the last 4 were bypassed
instead of the last 3.

The lack of compression might be nice even if you only have 256
bins because for 0.5 mev trig and 5.5 mev full-scale that would be
abt 20 kev per bin where the inherent scint/pmt avg analog reso
mid scale is about 80-100 kev so you still have several "stairsteps"
up and down each side of a peak. Ironically, even tho I could have
as many as 4096 bins to cover 0-5.5 or 0-11 mev, I settled on 256
as the best compromise between counts per bin, how long it took
to gather enuf counts for good statistics, how good the spectrum
"looked" on the pc screen in mathcad, etc etc.

wk

homebrew
Posts: 4
Joined: Tue Apr 14, 2009 12:00 am
Real name:

Re: Still not desired resolution (new scint from GEO)

Post by homebrew » Wed Apr 15, 2009 2:28 am

hi All,

Looked inside the box today - replacement resistors in divider string
for 3 inch Bicron were 1/8 watt - not 1/4 watt

myID
Posts: 139
Joined: Wed Jun 25, 2008 6:16 pm
Real name:

Re: Still not desired resolution (new scint from GEO)

Post by myID » Wed Apr 29, 2009 11:53 am

Sorry-

I do not have much time to work on my fusor due to my move- and I have not much time to follow the forum due to my new job
Anyways- this is great info! I am really looking forward to modify my detector and get the desired resolution- finally!!
I love this forum and all the great tips I get from the experts here!!!

THANKS

hjerald1
Posts: 54
Joined: Mon Feb 25, 2008 4:28 pm
Real name:

Re: Still not desired resolution (new scint from GEO)

Post by hjerald1 » Thu Mar 04, 2010 3:01 pm

I also have one of these Geo probes and would like to improve its performance.

Questions:
Has anyone by now (e.g.Carl or Roman) tried wk's modification? How
did it work in your hands?

For folks like me (electronically challenged) could someone (wk,
Carl, others) supply some more details (pictures, components
used etc....sorry I am not even familiar with what "805s" are for example)
of the mods so that my chances of doing it correctly are improved?

Did I not see somewhere that Geo provided some info on opening the
probe (mild heating to soften the thermoplastic adhesive?) or am I having
a senior moment?

If you wizards of all things electrical will humor me in this project, perhaps I can help someone in the future with questions on chemistry, or the like .

Thanks,
Jerry

Post Reply