Geiger Counter Basics

[Richard Hull]

Geiger counters are a form of gas amplifier tube in which ionizing radiation triggers an avlanche type discharge in a low pressure gas already under electrical stress. Amplification factors within the detector often exceed a factor of 10,000!

The geiger tube is traditionally evacuated and then backfilled filled to about 50 torr pressure with air, argon or neon.

The normal geiger circuit might contain a Geiger tube and a single current limiting resistor of between 3 meg and 6 megohms.

The geiger tube's conductive outer shell (cathode) is normally at ground potential. The Anode or central wire is supplied with about 700-1000 volts through a current limiting resistor. The signal (large negative going pulse) heralding a nuclear particle detection is usually retrieved directly off the anode via a 2 kilovolt rated, 100pf disc capacitor. This, now isolated, output pulse can be applied directly to a small audio amp to hear the clicks. It can also be applied to a simple integrator circuit to drive a meter to indicate the rate of incoming pulses. A simple counter could also be attached to count incoming pulses.

A GM tube is in discharge mode at the moment of detection and wants to draw a lot of current, but the limiter resistor prevents this and after about 3ms the discharge fades away due to the current limiter action.

To be able to count beyond about 300 counts/second, the tube's conduction must be stopped more rapidly. The discharge condition must be QUENCHED.

In the early days 1925-1950, this was accomplished by electronic vacuum tube quench circuits. The normal method was to use the output pulse to cause a vacuum tube in shut with the geiger tube to short the tube out, instantly removing all energy from the device. The resultant pulse was very short and allowed the counter to exceed 100,000 counts per second in well designed systems.

During this time frame there was no other device capable of counting nuclear particles individually at rapid rates.

Shortly before WWII it was discovered that chemical quenching was possible within the device which would easily allow 10,000 counts per second and avoid the electronic circuitry. Introducing about 5-10% alcohol or a halogen such as Iodine or Bromine would perform this task.

Common rate meters and health physics counters rarely needed to count beyound about 2000 counts/second. So the simple internally, or "self-quenched" tubes became the norm.

With the advent of the scintillator and the silicon detectors which could easily exceed a million counts per second, any form of high count rate geiger circuits were abandoned. This, inspite of these latter systems costing many times more to produce.

The average amateur wanting to count alpha particles at high rates has only the geiger tube to turn to. A scintillator will not count alphas in normal inexpensive setups. A PIPS or Surface Barrier silicon detector is about 10 times more expensive than a good geiger tube and much, much more difficult to impliment electronically. Silicon detectors can be readily harmed by constant exposure to extremely high count rates of alpha particles.

So, at the dawn of the 21st century, we are currently only using internally quenched geiger tubes in ultra simple counter systems in low to moderate count situations.

How can we do modern electronic quenching for geiger tubes?

The ultimate limit to count rate on a superlative modern self quenched tube is called the dead time. This is related to the RC time constant of the capacitance of the tube ( a few picofarads) and the current limiter resistor (megohms) and in the best systems is about 50us. This will ideally limit the count rate to less than 10,000 counts per second, with a safe zone of one dead time. ~600,000 cpm.


We can still shunt the tube out after only a couple of microseconds, but need a low capacitance fast switch located near the device to attack the dead time issue.

Another method which can be applied is to bias the tube to about 100 volts below its geiger region (~600 volts) and then pulse in extra voltage to place it in the geiger region. A monitoring circuit waits for the first pulse detected, increments the counter by one and then drops the voltage back below the geiger region. This quenches the discharge faster than a chemical agent, but not quite as fast a the good shunt arrangement. Once the circuit senses the extinction of current flow in the tube, the voltage is pulsed back up to the geiger region again and the sequence repeats.

Thus far, I hae achieved 15 us geiger pulses allowing me to count to about 66,000 counts/second or 3 million CPM.

Any other ideas?

Richard Hull

[guest]

Two questions: how far does one have to yank down the geiger tube voltage for effective quenching, and for how long? If I get those answers I think I can provide a solution. If it is sufficient just to yank the tube out of the geiger reigon, the solution becomes even simpler.

[Richard Hull]

The tube geometry has a lot to do with how far out of the geiger region you have to drop the tube voltage. Fill gas, internal quench agents, etc. also have an effect. By doing this, you will speed up quench but nothing will replace shunting the tube for ultimate speed. As a rule, I might say 200 volts below the geiger region would do it. Remember, when you are yanking, the tube is conducting and the thing wants to keep going. You are essentially just relieving a potential stress here, not removing it as in shunting. Only lab bench tests with a specific tube would be definitive. I would say that halving the normal quench time would be a significant step.

The tell tale failure of quench in high speed detection is a staggerd two pulse peaking of the output waveform. This means you were only partially quenched when a second event came along. The peaking of the normal pulse is often only 25 us wide in a normal GM tube, but the long slow tails can be over 100usec. Stamping out these burning embers is what we are after. Beyond this is defeating the dead time with more amazing antics.

Richard Hull

[guest]

Humm... I was calibrating my chicago nuclear 2605
when the probe cord turned to dust. The Geiger stopped dead in its tracks. Any sugestions on cord type(s) ,it does not have to be the coiled cord I'm replacing.

Larry Leins
Physics Teacher

[Richard Hester]

You could use coax . This will support the voltage and also carry the signal well. You probably want to put a MHV conntector on the counter, so that you can easily change probes. Good old RG-58 should do just fine.

[DaveC]

With the coax, the current limiting resistor, would need to be at the G-M tube end of the cable in order to prevent the cable capacitance from lengthening the conduction pulse... although, the lengthening is really not much ( about 1.5 ns per ft).

Dave Cooper

Dave Cooper