Test of SBT-11A anode connections

Okay, here goes again. I will first see if this editor will accept a cut and paste of my data, then write in comments.
Terminal log file Date: 2017-08-07 - 7:05:55 PM
Using old americium-241 smoke detector button against SBT-11A protective mesh
TEST 1:
3 anode resistors, 4.7M 1 series resistor, 4.7M Internal GC anode resistor, 4.7M
Time CPM uSv/h Vcc
19:08 34349 195.7989 5.31
19:09 34962 199.2932 5.31
19:10 34823 198.5008 5.34
19:11 34995 199.4813 5.31
19:12 34850 198.6547 5.31
19:13 34708 186.4447 5.31
19:14 34104 194.4023 5.29
19:15 34400 196.0896 5.34
19:16 34285 195.4341 5.31
19:17 34401 196.0953 5.31
19:18 34454 196.3974 5.34
19:19 34340 195.7476 5.31

Terminal log file
TEST 2:
3 anode pins wired common 2 4.7M resistors in series internal GC anode resistor, 4.7M
Time CPM uSv/h Vcc
19:44 34613 197.5318 5.34
19:45 34711 197.8624 5.31
19:46 34658 197.5603 5.31
19:47 34899 198.9340 5.34
19:48 34659 197.5660 5.29
19:49 34856 198.6889 5.31
19:50 34804 198.3925 5.31
19:51 34909 198.9911 5.31
19:52 34872 198.7801 5.31
19:53 34958 199.2704 5.31
19:54 34913 199.0139 5.34
sbt-11A test
19:55 34846 198.6319 5.31

If the above makes no sense (remember, I can't see the screen format), then both sets of readings average around 34k CPM. Both about the same. Next time around, I'll try just one connection to the centre anode, then two outside anodes, then readings with paired outside anodes and the centre anode.
I had a lot of 4.7 megohm resistors. The first set has three twisted common to a fourth series resistor, and that feeds into my GK4 counter which has an internal 4.7M resistor. The second test was with 3 anode pins jumpered together, then two 4.7M resistors in series.
Values are separated by tabs for import into Excel. I did try to make this file an attachment and got lost in the text editor.

A surprising result perhaps, although I'm not sure what the maximum count rate from the SBT-11A really is - and whether 34K is the saturation point.

It could be that connecting the three segments individually offers some sort of redundancy - if one fails the other two at least keep working. It makes no sense for the manufacturers to go to the trouble of making three separate anode pins available if just one would do the job - cheaper to make it with just the one?

Hi Chris,
The saturation level is actually much higher. I'll push that in a later test, say with two americium-241 buttons.
Here's my latest results. I taped a single, newer 241Am button to the middle of the mesh window. I tried the left anode, centre anode, then the right anode with 1 10M resistor feeding into the GK4 4.7M internal resistor. The results were low for the outside two anodes and saturation for the centre anode. So, I tried this run with all three anode pins common feeding into 1 10M resistor connected to the GK4 with it's internal 4.7M, giving an approximate total of 15M.
Terminal log file
Date: 2017-08-08 - 10:08:48 AM

---

"Time "CPM "Average CPM "Average Microsieverts/h
10:13:02 54074 55,263 173.7
10:14:02 55346
10:15:03 56108
10:16:03 56003
10:17:04 56066
10:18:04 55802
10:19:05 55474
10:20:05 55790
10:21:06 55535
10:22:06 55097
10:23:07 55651
10:24:07 55093
10:25:08 54999
10:26:08 55221
10:27:09 55394
10:28:09 54735
10:29:10 55566
10:30:10 55347
10:31:11 55553
10:32:11 54929
10:33:12 55348
10:34:12 55370
10:35:13 55303
10:36:13 55374
10:37:14 55496
10:38:14 55401
10:39:15 55115
10:40:15 54906
10:41:16 55052
10:42:16 55218
10:43:17 55737
10:44:17 55343
10:45:18 55496
10:46:18 54680
10:47:19 55368
10:48:19 55273
10:49:20 55735
10:50:20 55443
10:51:21 55169
10:52:21 54872
10:53:22 54810
10:54:22 55273
10:55:23 54886
10:56:23 54449
10:57:24 54595
10:58:24 54610

After using one 241Am button in the previous long test run with 1 10M resistor, I put a second button of the same age and type on the mesh, held in place with electrical tape. Same resistor.
First, the americium buttons were aligned along the centre anode axis. Counts over three minutes were around 40k CPM. I re-aligned the buttons cross-wise in the centre, perpendicular to the centre anode, approximately over the left and right anode. Results were 50k CPM over two minutes.
Obviously, the readings dropped from 54k CPM with a single button. I'm guessing saturation was reached, though I will make additional tests. One test will be with two shielded sources and an Excel formula to determine the dead time.

I sure hope the SBT-11A is halogen quenched, else, if they use hydrocarbon vapour, I'm swiftly killing this tube.

Regarding use of computer connector clips, they sure make setting up experiments easy. I didn't know they were designed in a way that would fit until I read your post. Like jumpers and other parts, I've an abundant supply of fans and front-panel leads stripped from old PC's.

Well if the count on the centre segment - the one closest to the source - was higher than the outside two that would support the supposition the three segments operate independantly. It sounds like having two buttons over the centre segment is saturating. It would require an oscilloscope to be connected to the anode pin to confirm that the voltage is not recovering, there may be a lot of low amplitude spikes at very high count rates which are simply not able to trigger a count.

I seem to recall reading somewhere they are quenched with Br.. not certain though.

As mentioned in a previous post, I accidentally ruptured the mica window on an SBT-11 or SBT-11A. The three anode rods are definitely insulated from each other and run the inside length, parallel and spaced a couple of millimetres (can't be exact).
They are also a couple of millimetres above the sheet metal ground which folds up 90 degrees along either side. The anode rods run parallel to the long axis.

Why the Russian engineers designed the SBT this way is a mystery. Would be interesting to find out their reasoning.

I also ruptured an LND7313-3 a couple of years ago. That pancake had an anode of three concentric circles and a small disk centre, each of flat metal about 1mm across, with spokes connecting them to each other and the centre.

After removing the tape and 2 americium buttons from the mesh, I turned the SBT-11A face down on my desk and ran background for nearly two hours. The average count was 22.4 CPM. My LND7313-3 RM-80 is showing 31 CPM average over the same period.

Ran the test to see if there was any residual contamination on the mesh (recoil dust) from the 241Am buttons.