[Sneap] High radiation at low terminal voltage

[Dr. Pertti O. Tikkanen]

Hello fellow sneapers,

In our vertical 5-MV-, N2/CO2-insulated tandem we are having severe 
problems in reaching any HV higher than about 2 MV. Good ideas for 
locating the problem are welcome!
Here is a description of how the symptoms developed:

-After a tank opening in September 2007 we conditioned the machine up to 
4,3 MV and did run some experiments where the tandem performed perfectly.
There wasn't any need to go to higher TV's.
The just-installed green Siegling belt ran also well. This time we had 
left out the inner equipotential bars. A bad mistake, perhaps? 
-The problems started to appear after a few days AMS run at 3.2 MV. In 
that run, the charging current was relatively high, as we used about 
50-microamps of corona probe current. It wasn't exceptionally high, 
however.
-Subsequently, when starting the tandem for another experiment, much 
higher radiation levels, tens of micro-Sv/hour (compared to what we 
usually have, only a few uSv/h) were observed already at 3.5 MV. It was 
impossible to go any higher than about 3.8 MV because of high radiation
levels and very unstable TV.
-As the insulating gas was a bit more humid than usual, we dried it once 
more to reach a dew point of about -69 Celsius. However, this made the 
problem even worse; the radiation started to inrease heavily at voltages 
lower than 3 MV's.
-There were almost no tank sparks. At least none were related to increased 
radiation level.
-Before opening the tank, we monitored the radiation to find the location 
of its source. All that we found was that it was concentrated near the 
terminal.
-The tank was opened, but we didn't find any apparent reason for the extra 
radiation. The only difference compared to normal was that there were a 
lot of sparking tracks at the edges of the terminal shield. That's the 
place where the electric field is the strongest, I suppose.
-We thought that there are some extra particles in the insulating gas and 
installed new filter elements into our pumping lines. Measured also the 
particle concentration in the gas. It was very small.
-The resistors were checked carefully, a few were found defective (in 
random places), all surfaces were cleaned carefully, the belt was found in 
good condition. Looked for extra items like loose wires etc. in the 
column, found none. So everything looked just OK and we closed the tank, 
only to find out that
-The high radiation level was still present!
-We got an idea that there was some chemical contaminant in the gas 
because we had recently installed new absorbant in our gas dryer. 
-Replaced the insulating gas with a new mixture of N2/CO2. Furthermore, 
the gas dryer was regenerated carefully at higher temperature.
-After all this we are in a situation that a lot of radiation is generated 
at about 2.2 MV or even at lower TV's.

What did we miss?
We are anyway going to open the tank again and try to locate the source of 
the problem.

Best wishes and best regards,

Pertti

% Pertti O. Tikkanen, PhD         email: pertti.tikkanen(at)helsinki.fi
% Senior Laboratory Manager

% Accelerator Laboratory                        TEL:   +358 9 191 50006
% P.O. Box 43 (Pietari Kalmin katu 2)           FAX:   +358 9 191 50042
% FIN-00014 UNIVERSITY OF HELSINKI, FINLAND     GSM:   +358 40 72 65386