Easy Radiation Detector build

After Fukushima nuclear disaster I decided to build my own radiation detector, but haven't any idea how it works I did some google research and ordered russian Geiger Tube STS-5. It have good price/quality value. But you can use any Geiger-Muller Tube for your project. Just take in mind the type of the radiation you want to detect. STS-5 can detect Beta and Gamma radiation, but not Alfa. Technical specification of my GM tube is:

Starting voltage = 280...330 V
Advised working voltage = 360...440 V
Plateau length = at least 80 V
Plateau slope = 0.125% / 1 V
Maximum natural background = 27 pulses / min

Load resistance = 5...10 MOhm
Allowable stray capacitance of input circuites = 10 pF
Transit capacitor = 7...10 pF
Allowable surround temperature -40 ... +50 OC

Its mean I need about 400V DC with very small current to feed the tube. It was the first problem how to get high voltage? All my power supply has maximum 30V DC out. I found 3 options:

1. Directly from AC plug in my house using diode rectifier bridge. But its not safety and not exactly 400V, something about 220V x 1.41=310.2V So I declined it fast.

2. Using small step-up transformer with generator and voltage multiplier. This is very elegant and professional, but I haven't any suitable small transformer in my box. I tried to wound my own on 1 inch toroidal core, but after about 100 turns I was very disappointed. It's pretty hard for the beginner to wound 400 turns, especially if you do not have experience in the winding. So I decided to search another solution and I found it.

3. The circuit you can see is very simple. NE555 is the 14Khz square wave generator. It's drive IRF710 mosfet and you have about 400V after the rectifier D1. Try some values of the inductor L1 and measure the voltage. L1 can be from 1mH to 10mH inductor. I used 8x10mm radial inductor with ferrite core from old TV pcb. D1 should be some fast rectifier diode 800-1000V FR157, BY399 etc.
C4 capacitor can be any value from 2.2nF to 4.7uF but for 450V at least! All resistors 0.125W or 0.25W



The second problem I met was how to measure 400V in this circuit. My cheap voltage multimeter have 1M internal resistance, so I got lower value readings. I did 100M ohm divider to ensure right high voltage reading. I used 10x10 Mega ohm 1% resistors and connected it one by one in serial with the multimeter red test lead. So I got about 4.02- 4.05 V multimeter reading in D1 C4 test point. It's mean the real value was 402-405V.

EP1 is the cheap piezo from the telephone ring. It starting do short cracking sound when the tube detect background radiation - approximately 15-27 times in minute. The number of the pulses increases with the approach to the source of radiation.

This circuit need about 30mA on 9V, so it little big compared to step-up transformer circuit. But it very easy for your first radiation detector building and GM Tube testing device.

Hello,
I've built this electronics but I have some problems with this.

With 10 mH inductor I see 400 V but the current is over 1 A instead of something about 30 mA and so the transistor is heating up rapidly.

Do anyone has any idea what's wrong?

Thanks in advance for any response,
golkaj

Did you used exactly the same components? May be the problem is the inductor type.
I used this types of inductors in my project, dimension 10x8mm



Do you have 14KHz frequency from pin 3 of IC555?

No, I have 10 inductors of 1 mH each and I tried from 1 mH to 10 mH with current from source and outpu voltage increasing with inductance. They look like resistors.

Yes, I have 14 kHz from NE555 and I used linear potentiometer of ca. 20 kOhm as R2 so I can change the frequency (not so much difference in output).

The rest of the circuit is the same as yours.

golkaj wrote:

They look like resistors.

The inductors I think is the problem. You use axial inductors that look like resistors, but I always use in this circuit other type of inductor, see the picture I attached before.

Try to use inductor I suggested before. This circuit have also some variation I found online (see the attachments) Maybe it will be useful.

I'll try to get this kind of inductor.

As I look at your last image I do not know what to do with the leg number 7, not connected? I don't think so...

Before all try anothe inductor. If it will not help we'll investigate it where it may be a problem. Please upload the photo.
I use the circuit from the first message, because it have standard NE555 astable generator.
http://clarkson-uk.com/555-timer/operation/hiframes.html
If you'll not find the inductor it's possible to use small transformer and diode multiplier from the secondary winding to get high voltage.

Strange thing.

Now I used inductor made by myself as ca. 1800 turns of copper insulated wire around a stick which kind of are used for cleaning ears. I've estimated the inductance as ca. 0,5 mH.
This item instead of those inductors from electric store gave me 510 V and current from source 0,16 A. Sounds better but my potentiometer doesn't change the voltage so I have too much now and the current is still to large.

As you said it's about inductor apparently but it's interesting thing, why and how

Tomorrow I will try to get right inductor.

Inductor Q-factor is the reason I guess. Try to find ferrite core inductor.

It's alive!

Thanks for Your advice, I used not-axial inductor of 10 mH (looks like Yours), the current is about 6 uA (yes, microamps) with background radiation and to 30 uA with some kind of source.

The output voltage is easy-adjustable by the potentiometer I mentioned. I must take care with the correct output voltage (I think that I'm working under plateau now) but this is a piece of cake when I have good readings now and working detector

Thanks once more!

PS I'm thinking about using the voltage on working resistance to open a second transistor to make good sound by the speaker.

I'm glad to hear your success
Measure high voltage with 100M resistor. I user 10x10M 1% resistors in serial, so the final reading after 100M you must multiply to 100 to know the real voltage.

What type of tube do you use?

My tube is STS-5. I used 10x10 MOhm but in another way.

If we connect output to the 10x10 MOhm (10xR) resistors and plug the voltmeter with r=R (I have 10 MOhm of internal resistance in my DT-9205A meter on 1000V and on 200V ranges) to one of these resistors we have on it the voltage divided by 2; from source we have current I=U/(9,5R) witch then splits in two ways: through the resistor and through the meter equally 0,5I, than merge together and flows through the rest 9R. It seems to me that the reading is the voltage on internal resistance of course, so it shuold be 0,5I*R=0,5*(U/(9,5R))*R=U/19.

Rather this I used meter as one of these 10 resistors, so I have meter in series (like amp-meter) with 9x10MOhms. Then the reading is simply U/10.

If using 10x10 MOhms in series with 10 MOhm meter then one must multiply by 11 the reading.

This is verified by experiment also of course - so I think that I haven't understand You right

STS-5 is very good and sensitive tube.
I have cheap chinese multimeter and it have 1M internal resistance. It's overload high voltage inverters and because of it I has a lower voltage reading.
For more prescision resistance devider calculation the formula of is:
10x10M + 1M = 101 if somebody use cheap r-meter
10x10M +10M = 110 for your DT-9205A

Oh, You have 1 MOhm of internal resistance - so it's OK, I thought that You have 10M

For better reading in mind I used 9x10M + 10M in multimeter so I multiply by 10 simply.

Now I'm thinking about some kind of quantitative measurement in this system. As I know reading in counts per second require use of microprocessor, am I right?
It would be great if we can measure number of counts in for example 3 seconds and display value in cps. But not in seconds 1-3, 4-6, 7-9 but refreshing on each second, so rather in sequence 1-3, 2-4, 3-6, 4-7 etc. But I'm too short in programming uprocessors

So I thought about analog reading, the signal on working resistance could charge some capacitor and It would be discharging through this resistance so there would be some kind of average voltage which can I measure by voltmeter. But I do not know if this will work. The voltage would be small and internal resistance of voltmeter would be smaller than working resistance...

Please check the link about analog measure:
http://www.techlib.com/science/geiger.html#HV%20Generators
"The pulses may be counted with a digital circuit or may simply be averaged to apply to an analog meter...."
He use 60uA analog ampermeter and two-pole six-position switch for range switch. The problem is still the calibration if you do not have a special radiation source.

Or you can find one of several projects in the web with microcontroller and display. But if you'll meet a real serious radiation level the flash of microconroller will die...Thats way I think a dosimeter with microconroller its just a toy

Professional dosimeters, like Victoreen CD V-777 or russian military DP-5B, is analog reading! Its stable in nuclear disaster condition.

About our simple circuit:
http://fm-pirat.narod.ru/shemi/gmtube
This is my modification with sound and led indication. Of cource it cant measure radiation level, but if its start clicks and led flash faster I know it's time to go away very vast

I have buzzer and LED since my circuit works

The output voltage I connect to the base of BC548 and buzzer is between +9 V and collector, emitter is connected to mass.

Speaker works great, but the diode shines when there isn't any signal also; with registered radiation it is blinking stronger. As some kind of noise, but the voltmeter detects nearly zero.

I am a nuclear physicist on PhD studies so I know exactly that reading in radiation dose rate in uSv/h or other units is ridiculous. Even if "calibrated". This is why I prefer cps or reading proportional to it on analog volt- ora amp-meter.

I've built simple impulse counter on HEF4541, 4543 and 7-segment LCD display.
Unfortunately the noise I mentioned gives "continuous" reading, impulse from working resistance 10M gives from 4 to 6 impulses per one short sound from speaker...

not realy understand your final circuit..
I understand you did cps counter with 7-segment led, but not really understood where you get countiniosly voltage readings
Can you please upload your circuit with jpg or gif format? ]

Another suggestion about your "noise" - is came from NE555 generator? Can you check with the scope?

Do not know why, but now it works

LED is working OK.

I added capacitor 2n2 parallel to working 10M and counting is going also OK, one by one increasing.

In a free time I'll prepare scheme for others, rather to improve than use in this form

I'm interested to see your final circuit

I hope that You won't be angry on me because I used Your scheme

Vx is the voltage that goes into BCD counter HEF4518B http://www.nxp.com/documents/data_sheet/HEF4518B.pdf and from it to BCD to 7-segment decoder HEF4543B http://digsys.upc.es/ed/components/sequencials/latches/HEF4543.pdf and to 7-segment 4 digit display.

Count rate is obtained by dividing (in mind or with calculator) number shown on the display and time of the measurment.

These are pictures of my detector.
On te side-view one can see the hole for the detector tube. It might be useful to make a window of metal mesh but I didn't do this in my first device.

On the scheme above there is one mistake - two BC548 transistor are NOT connected to 400 V but to 9 V of course

Nice work! Thanks for sharing.
What is the function of the buttons?

Black one is simply on/off for power.
Green one right to the black button is on/off for speaker (LED is working always when the detector is on).

Second row, from left to right are used for counter:
on/off (large current through display so it may be on or off), start/stop counting and reset to 0000.

Thanks for the appreciation