[Swlug] DIY Geiger counter

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

At Z+0000=2025-03-12Wed22:58:11, Alan Gray via Swlug sent:
> [...]
> It isn't just the material properties at extremes, but the variation between visually similar products due to manufacturing choices. A good example  are ceramic disc capacitors. Some are very poor at RF frequencies. You cannot tell by looking.

    All of my childhood RF circuits used ceramic disk capacitors from a hand-me-down Radio Shack 30-in-1 Electronics Projects Lab from probably circa 1980s before I was born.  Maybe back then, there were not so many cheap knockoffs of even the cheapest passive components.

    What makes them poor at RFs?  Is it parasitic inductance?  Where does that come from in a disc?  I've heard of parasitic inductance in rolls and spirals, but not discs.  I don't understand how a ceramic disc capacitor could even be made bad.

    There'll be some way to test a capacitor to check it, but an ordinary capacitance tester will only tell you its capacitance.  Conceivably some testers could also tell you its parasitic inductance, but I'm not familiar with the specialist gadgets.  Peak Electronics Design Ltd. make all sort of testers.  I wonder whether they have a way to test the RF performance of a capacitor.

> These matters are important throughout, but at the beginning of the chain, with such weak signals, they are critical.

    Same for loop areas of the paths of changing signal currents.

> Look at any front end RF circuit and you will find mica capacitors dominate especially in the signal line.

    I'll keep an eye out for mica capacitors when I next buy some, because my digital logic circuits are aiming for being as fast as possible -- 10MHz, 100MHz, maybe I'll get up to a GHz in the coming years.

> Regarding the transistors. These switching types easily handle currents of hundreds of milliamps. They do this by having physically large substrates/junctions that aid heat diffusion. The downside is that it introduces more capacitance. Look at small signal RF types and usually they handle maybe 5ma.

    That might be why the page 78 design did not need a capacitor where I suggested would help.  However, my suggestion would still help with the loop area situation because it's possible to have a miniscule loop between a photodiode and a tiny capacitor point-to-point soldered right across it.  So it might be better to use a small-signal integrated Darlington Pair (if such things exist, which I suspect not, seeing as such integrations are probably only ever sold for driving loads that draw higher currents), or even the 2 discretes as before, and then ensure that the majority of the capacitance is at the designated locations, rather than relying on parasitic properties.

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/20250314/21f5b88b/attachment.sig>