| T O P I C R E V I E W |
| Benezi |
Posted - 12/11/2014 : 13:27:27
Hello friends,
in last days i tried to assemble a new tube to the GMC.
It was a big SBT10a which i installed in a new case.
I used BNC-connectors to connect the GMC-320+ with the new tube.
You can see it on the picture.
It works great due to the big tube, but i have a little problem:
if i approach the tube and the case of the GMC-320 the counts increase about 30%. If i take it away, the additional counts are gone. The same happens if I measure pitchblende. By approaching, it increases about 30%.
I suppose that it is a problem of shielding the GMC. Therefore i used a coaxial cable even inside the GMC.
But still the same problem.
Does anyone have a idea?
Thanks for your help,
Regards from Germany,
Benedikt
Image Insert:
163.88 KB
Image Insert:
183.54 KB |
| 4 L A T E S T R E P L I E S (Newest First) |
| Alchemy2 |
Posted - 12/21/2014 : 08:24:09
Benedict,
I also see that you have failed to shield off the cable and are using a jumper. Nope! Carefully remove a small piece of covering off the cable to expose the shielding. Don't cut too deeply or you'll damage the cable. Then, carefully tin with solder the exposed shield. and solder to the board near the clip (negative clip) like you did for the positive lead. This grounds the shield more directly. Otherwise a lot of the shield is floating and acting as a variable parasitic capacitance. This is why your hand affects counts. Do same with tube as well. Shield to connector with only a small wire to the BNC ground.
This should help.
Darren
|
| Alchemy2 |
Posted - 12/21/2014 : 08:19:18
Also, Bendict:
If your cable is too long, parasitic capacitance will diminish the count frequency (Max CPM) or even fail to count at all. Try making connections as short as possible. RG174U mini cable is fine for this application, keep length below 400mm (I'd say no longer than 250mm to start)
Good luck!
Darren |
| Alchemy2 |
Posted - 12/20/2014 : 01:24:36
Benedict,
try using 10M resistors off ALL anodes, tying them together where the BNC central post (signal connection) is.
So you go anode-10M-BNC center off each anode. cathode goes straight to shield.
I haven't tried my SBT-10A tube yet, but my 320+ with SBT-11A can count to over 40000+CPM no problem and the dead time evaluated using a two source method came out to 11us.
Darren |
| Benezi |
Posted - 12/12/2014 : 13:36:51
Here (h**p://hardhack.org.au/book/export/html/44) i found the following:
"The 4 10Meg Resistors (Note experiment with values) also play an important role by reducing parallel capacitance, preventing the voltage across the tube C1 to run away due to the high impedance of the supply causing oscillation and also ensures a zero bias to the filaments of the fluorescent tube reducing the production of spurious pulses and discharges within the tube."
I think that could be the problem. I tried to set Resistors parallel to the tube, but then then the LED is red, does not register anything.
Does anyone have a idea? |
