Login Join Maker Pro

# External Quenching of GM Tubes

E

#### Ed Rooke

Jan 1, 1970
0
I presently working on a GM detector. Unfortunately the source Im
using has a fairly high count rate and the period between events seen
by the tube is much shorter than the dead time. I thought I would get
around the problem with external quenching; switching the supply to
the anode to below the starting voltage (750 V) of the tube for the
dead time (200uS) this period is given from a monostable triggered by
the event.

The problem seemed to me to be trivial but after trying various shunt
and series arrangements with FETs I have realised it is a bit more
difficult than I originally thought.

One problem is my power supply; it was originally designed for a PMT
so it cant provide that much current. This has forced me to use a 16M
potenial divider to switch down to 750V and the big resistances are
giving me a large time constant even though the capacitances are quite
small. This makes the rise and fall of the voltage too slow (I was
aiming for 20 uS or so).

I cant drop the supply right down to deck because some voltage (not
sure how much - would be interested if anyone knows?) is required at
the anode to let the tube recovery from an event.

The other problem is milli capacitance accross the FETs and emfs being
induced by the large voltages. The most successful circuits I have
made respond to the first pulse but as the tube is switched back on
again after the dead time a pulse appears on the cathode hence
retriggering the monostable and causing the tube to keep switching
off.

I found one circuit using valves on the net but it didnt make any
sense to me. Does any one/ has anyone done this with FETs ?

Any advice would be much appreciated; Im pretty stuck on this one.

Cheers
Ed

R

#### Robert Baer

Jan 1, 1970
0
Ed said:
I presently working on a GM detector. Unfortunately the source Im
using has a fairly high count rate and the period between events seen
by the tube is much shorter than the dead time. I thought I would get
around the problem with external quenching; switching the supply to
the anode to below the starting voltage (750 V) of the tube for the
dead time (200uS) this period is given from a monostable triggered by
the event.

The problem seemed to me to be trivial but after trying various shunt
and series arrangements with FETs I have realised it is a bit more
difficult than I originally thought.

One problem is my power supply; it was originally designed for a PMT
so it cant provide that much current. This has forced me to use a 16M
potenial divider to switch down to 750V and the big resistances are
giving me a large time constant even though the capacitances are quite
small. This makes the rise and fall of the voltage too slow (I was
aiming for 20 uS or so).

I cant drop the supply right down to deck because some voltage (not
sure how much - would be interested if anyone knows?) is required at
the anode to let the tube recovery from an event.

The other problem is milli capacitance accross the FETs and emfs being
induced by the large voltages. The most successful circuits I have
made respond to the first pulse but as the tube is switched back on
again after the dead time a pulse appears on the cathode hence
retriggering the monostable and causing the tube to keep switching
off.

I found one circuit using valves on the net but it didnt make any
sense to me. Does any one/ has anyone done this with FETs ?

Any advice would be much appreciated; Im pretty stuck on this one.

Cheers
Ed

Perhaps the use of the equivalent of a "programmable" or "gated" shunt
regulator, where one can switch the regulating voltage from (say) 1050V
to almost 0V.
The possible problems with this scheme, is that the roughly 400pF
drain capacitance of the regulating FET, coupled with the needed series
resistor from supply to shunt regulator may give a too slow rise time.

Be advised, that once something ionizes the gas in the GM tube, the
time for those ions to dissipate to a low enough level will also cause

I could supply you with a gated 900V shunt regulator, which couuld be
used in series with a 600V (max recommended) regulator - which would
result in two shunt regulated voltages: 1500V and 600V.
If you choose to use two regulators this way, you can specify these
two voltages as you see fit, to a maximum of 900V each, in 25 volt
steps. See:
http://www.logwell.com/products/Codatron_Corotron.html
**
Another possible way is to use a FET to create a negative pulse that
is capacitively coupled to the tube; obviously this can cause risetime
problems related to the series capacitance (coupling in series with FET
drain capacitance).
**
Yet another way is to add a pulse transformer in series with the GM
tube, and pulse it with a FET, so the secondary voltage cancels out the
GM supply voltage. Make sure you terminate the transformer so the
inductance does not ring too much.
One can even use a 555 to drive a step-up transformer, giving a
flat-topped pulse.

R

#### Rich Grise

Jan 1, 1970
0
Why not just move the source farther away?

Good Luck!
Rich

E

#### Ed Rooke

Jan 1, 1970
0
A good suggestion. Unfortunately, within the environment that the
device must work there is a lot of background radiation. The source
must be located by the detector and the high count rate recieved to
get any kind of accuracy. Otherwise Id just be measuring the
background all the time.

Thanks anyway
Ed

R

#### Robert Baer

Jan 1, 1970
0
Ed said:
A good suggestion. Unfortunately, within the environment that the
device must work there is a lot of background radiation. The source
must be located by the detector and the high count rate recieved to
get any kind of accuracy. Otherwise Id just be measuring the
background all the time.

Thanks anyway
Ed

I do not remember if i mentioned that the voltage across the tube
should be decreased to below 32 volts, as that is the typical ionization
potential of most gasses once ionized.
The trick using a pulse transformer may be the best method.

Replies
7
Views
970
Replies
10
Views
863
Replies
15
Views
927
Replies
11
Views
1K
Replies
4
Views
2K