[Swlug] DIY Geiger counter

[James R. Haigh (+ML.LUG subaddress)]

I notice that the electrolytic that you have there is already a low-ESR one, and the way you are using it is only a diode away from being an envelope detector, pretty much.  I've just thought that if you leave the resistor of the envelope detector out for now you will force the circuit into a mode where it will be much more sensitive and much more likely to "get stuck on" as soon as you get some activity.  This could be a useful milestone to know that you are getting somewhere.

    Here's a schematic (in monospace ASCII-art) where the amplifier is a pair of NPN BJTs, like what you have there with the BC547s:-

                    +         +
                    |         |
                    |         >
                    >         < R
                    < R       > 3
                    > 4       <
          fast      <         |      OPEN-COLLECTOR
        Schottky    |    R2   C      AMPLIFIED
CRYSTAL --|>|--o----o--\/\/\--B  C-- OUTPUT (INVERTED)
               |    |         E--B   TO PULLED-UP ADC OR
               |+   >            E   CURRENT-LIMITED LED
          low ---   < R          |
          ESR ---   > 1          |
               |    <            |
               |    |            |
               -    -            -

    I'm thinking that it would be a good idea to leave R1 disconnected for now, because any charge that gets pumped onto the capacitor will be force to drain through your amplifier, thus sustaining each signal for a much longer time, which should help to get it working.  Once you get it working and find that it's on for way too long to be able to count pulses accurately, then you can experiment with the value of R1 until it behaves correctly.

    You may find that it takes multiple pulse events to get enough charge on the capacitor for the amplifier to detect anything.  In fact, yeah, the voltage on the cap has to rise by 2 diode-drops, probably about 1.2vDC, before you get any current flowing through the bases of the NPNs.  Therefore, I have added biasing resistor R4 to the schematic.  It'll raise the resting voltage as a potential divider with R1.  You could even replace R1 and R4 with a trimpot so that you can tune the circuit to the "bite-point" of the amp, so-to-speak, at about 1.2v, or maybe a little more.

    So you could try with neither R1 or R4 and see whether you get anything at all (which should be quite hopeful because if charge keeps punching through the Schottky, it will eventually spill over the 2 PN-junction-drops of the amp), and then you can try with both of them in there or a trimpot to see whether you can fine-tune the sensitivity and hopefully start detecting individual pulses.

    How are you testing the crystal?  Did you crack open that smoke alarm, with all the hazards that that presents if inhaled or ingested, or did you find some other way to test the crystal?  Is there a way to test your circuit without any radioactivity, e.g. by maybe swapping your crystal for a piezoelectric crystal or something that responds to percussion (e.g. from a broken lighter), so you can test how well your circuit works and how sensitive it can be to the slightest tap, and then when all that works well with the piezoelectric crystal, swap-in the Geiger crystal and cautiously test the real thing.

Kind regards,
James.
-- 
Wealth doesn't bring happiness, but poverty brings sadness.
Sent from Debian with Claws Mail, using email subaddressing as an alternative to error-prone heuristical spam filtering.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://mailman.lug.org.uk/pipermail/swlug/attachments/20250309/82d6ede4/attachment.sig>