[Swlug] DIY Geiger counter

Tue Mar 11 00:28:53 UTC 2025

At Z+0000=2025-03-10Mon22:32:01, Rhys Sage via Swlug sent:
> Brief answers right now. I didn't sleep well last night, did a 12 hour day and I've just come back and will be down for the night in a couple of hours/
> 
> I've only tried the BC574 circuit given albeit with tweaks as of Sunday. I probably won't get to my electronics lab until Saturday. I have a somewhat exhausting job driving extra-long special needs bus routes.

    Oh right.  Sounds like a good service.  I try to use buses but I miss them most of the time and get stranded because there are only 4 per day on my local route.  It's about a mile walk/cycle at each end if I catch the bus.  I would cycle all the way every time, but the roads are so dangerous.

    I'm special needs -- aspie & OCD, probably evident in the way I write my emails! :-P

> I'm what they call a floater - I have a route that I normally drive with a partner (whose route it actually is) but being a floater I can be thrown on any route with any group of special needs riders at the drop of a hat. Over the last couple of weeks I've had to learn and drive from memory 3 extra routes in addition to my normal route and deal with some passengers with challenging needs. I used to have my own route but when they closed the base I operated from, I chose to operate from a base local to my house rather than the main base (which is where my old base moved to). It means I am way out in the sticks, away from any shopping centres in a town where the whole town is run by one family. The Mayor, Police Chief, town council members, big town business owners are reportedly related to each other.
> 
> Sunday is usually my productive day. Saturday I'm still largely recovering from the week.

    Sounds very tiring.

> I did look at the Tip120 instead of the BC547 pair but it's a slow-switching thing. I suspect  a pair of 2N2222 might be better than the BC547 pair. The switching is a lot faster. Bonus - I have a massive pile of 2N2222. I have about 30 different kinds of transistor having bought a bulk pack of different types a few years ago.

    Faster transistors will help later, but I see bigger problems in the circuit you have so-far.  Particularly: your loop area is probably more limiting to your bandwidth than your BC547 transistors with their typical 300MHz current-gain bandwidth product; the use of a decoupling cap in this circuit is probably preventing it from working; and your transistors might not be sufficiently current-limited.  Hopefully my suggestions will fix some or all of those problems: reduce loop area (especially the smallest loop that the photodiode is on); replace the decoupling cap with a 1kΩ resistor; add another 1kΩ resistor to limit the current between collector of the 1st transistor and the positive supply rail.

> The ultimate goal - after being able to read (preferably background radiation) will be to slow the output down sufficiently that I can feed it to an ATTiny13 with its 1mhz clock speed. The radiation reading will be displayed on a TM1637 4 digit LED display and might be accompanied by sound from a piezo speaker. I want to be able to have a slow click of background radiation and a high click of high radiation. 
> 
> Should I look for smaller non-electrolytic capacitors? My smallest is .1uf metal film.

    In a previous email you reckoned that you had some .01µF metal film caps, in other words 10nF.  I reckon that trying a tiny capacitor in-parallel and directly adjacent to your photodiode will go a long way to mitigating the delays due to your loop area inductances.  It basically means that as soon as your photodiode starts conducting, current starts moving quickly between the capacitor and the photodiode, in that very tight loop.  Then, the charge that was taken from the capacitor through the photodiode will recharge by the same amount as the quick discharge, but this time through the more distant and slower amplifier, but this time it can take as long as it wants to do this, because the capacitor does not have a time-limit.

    The photodiode has a stricter time-limit, and by placing a capacitor right there across it, you convert this time-limit into more of a charge-limit, delivering some amount of charge with each pulse -- that's the idea, anyway.  You may later want to tighten the loop area of the rest of the circuit if you want the amplifier to be able to detect individual pulses, but for now, just having a capacitor at the crystal with a really tight loop there should be enough to get something working, even if the amplifier distorts the signal too much for now because it's too slow.

> I believe they make lower values. I normally order my electronic components from AliExpress, eBay and Temu since there are no nearby suppliers and those further away cost a lot more and save me only a few days off ordering direct from China. Radio Shack went bust or they'd be perfect.

    Aye!  Also, Maplin was no Radio Shack or Tandy, but I do miss being able to walk into a real electronics shop and see what I'm buying, buy it there and then walk away with it go home and work with it that same day, and even occasionally have to return it but be able to return it that same day and swap it for the right thing, go back home and carry on all on the same day.  It's only when it's gone that you realise what you had, even if it did seem a bit overpriced compared to online, there was so much less faff and waiting back then.

> I'll have to try to unscramble the text diagrams. That will have to wait. Right now I'm feeling very tired.

    Following my email to the list at 20:55:01, I did send you & the list another at 20:59:58 with screenshots attached showing how the ASCII-art diagrams look for me.  Unfortunately it bounced from the list complaining of an archaic 38KB limit, which is the lowest limit I have ever seen on any mailing list -- wow!!  “Message body is too big: 144200 bytes with a limit of 38 KB”  The smallest I've seen prior to this record was 200KB at SB LUG, but even that has changed to 1MB as of last year.  You did say that this list doesn't support images, now I see what you mean.  (Then again, I did send a tiny image just the other day (the envelope detector schematic from Wikipedia), so it is possible if they are really tiny enough.)

    But despite this, the CC'd copy direct to you did not bounce, so you should have a copy of it.

    But I'm wondering whether even without the scramble problems, there's something about my ASCII-art that you don't get on with.  I noticed that the Instructables article that you started with for the amplifier did not have any schematic, just photos of the point-to-point wiring.  So maybe if I describe my suggestions in terms of where you are already at, it might be more useful to you.

    I think I have 4 main points (starting from your photo that you sent us of the breadboard):-
(1) swap the electrolytic for a 1kΩ resistor (otherwise the DC-bias is just going to flood the capacitor and nothing much will happen once it is charged; this was surely intended for an AC signal, such as audio -- but in any case, I don't think it should have been an electrolytic because reverse polarity damages them);
(2) swap the tiny yellow flat jumper wire for a 1kΩ resistor (may not be necessary in this circuit with such small signals on the input, but this will ensure that a larger input signal will not cause the 1st transistor to pass too much current, and also protect the base of the 2nd transistor);
(3) add a tiny-valued capacitor in-parallel with and directly adjacent to the crystal (this allows more charge to draw through the photodiode while it is "on", given this charge much more time to accelerate through the rest of the circuit, which should result in a higher "dose" of charge per pulse getting into your amplifier);
(4) once your circuit is working, albeit with a lot of distortion and maybe noise as well, try reducing the loop areas of the other current loops, placing one or more capacitors across the voltage supply to truncate the effective loop area of the supply, using twisted wire pairs, swapping the transistors for higher-bandwidth ones, trying different values for the capacitor and resistors, and various tweaks like this.

> For those wondering how a video comes out every day - I shoot several at a time then schedule them.

    Nice.

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/20250311/2b01ce1a/attachment-0001.sig>