Sunday, 1 December 2019

PIN diode radiation detectors?

The Geiger-Müller gas discharge tube is of course only one of several ways to detect ionizing radiation. It is perhaps generally the most convenient - its sensitive area is greater than optical systems; its much lower cost than photomultiplier tubes and scintillation crystals; and its a heck of a lot more portable than a cloud chamber!

But I'd quite like to try all the methods that are within my reach! One of these, the cloud chamber, I am working on. I have suitable high current power supplies and heatsinks for a small chamber cooled using Peltier effect devices - I'm just awaiting delivery of those devices! The cloud chamber is of course the best for visually demonstrating radiation, as the paths of the ejected particles are visible to the eye.

Another method is by detecting the impact of a particle or energy quanta on a semiconductor junction. Most junctions, for instance a TO-92 transistor, are tiny, but there are a number of PIN photodiodes that have rather large junction areas, in the region of 5-7mm² which, while pretty small still, represent a much greater target area. One such, the BPW34, has been used in several simple detector circuits, and can be obtained for very low cost. As the likelihood of a particle event is still quite low, wiring them in parallel to increase the effective surface area is a trivial matter. Ive ordered five for less than £1.50. My intention is to build all five into a detector, but perhaps allow switching to select the actual detection area.

Of course the big problem with using PIN photodiodes for radiation detection is keeping them from detecting the radiation they were designed for - light! And extremely light-tight enclosure is needed, plus any indicator LED has to be very well isolated optically!

Friday, 29 November 2019

Further testing of the 3V HV generator

Well, this has proved fun! For some unknown reason, the entire circuit decided to stop working regardless of which transistors I installed! I ended up having to completely start again.

Eventually, I managed to get it working with the BC327/BC337 transistors at 3V supply, only to find that the maximum unregulated voltage was barely 300V. Even allowing for a little loading by my 10:1 HV probe, this is too low.

So I played about with the transformers, with interesting results! It would seem that the audio transformers I have are not 1:1 isolating transformers, but output impedance matching units - put them in the wrong way and the current goes up dramatically!

In the end, I found that the unidentified transformers taken from the old emergency lighting switch-mode inverter worked best. With one of those fitted, i'm measuring just over 400V unregulated. How much higher the true figure is, i'll have to wait for the 1GΩ resistor to find out! I also found I had to lower the oscillator capacitor value as the 100uF was causing visible pulsing! 4u7 seems a good value at present.

Generating High Voltage from Very Low Voltage

One of the things I want to do with these Geiger tubes, is to make an ultra-portable 'pocket' device. This is planned for the SI-19BG miniature α tube, which is only about 20mm long!

At present, I'm working on this circuit -

Where possible I've kept the values as stated, but as I don't have any of the specified 2N series transistors, I'm using whatever I have in stock, namely a BC212L and a 2N3904. The 1N914 is replaced by a series string of 47V Zeners and a couple of miniature neon bulbs! The transformer is a miniature audio transformer. The transistors have lower specs than those stated, and this might well affect the results. The closest I have in stock to the specs of the 2N4401/2N4403 are a couple of BC327/BC337 pairs. These have total power dissipation of 625mW same as the specified devices, but slightly lower collector-emitter voltages, however the collector current is higher at 800mA against 600mA. I might try these instead of the quickly-grabbed BC212L and 2N3904.

It is working on the bread-board, but due to loading effects of my HV resistor chain (remember this is only 100MΩ) the voltage reading is poor with the 3V supply (2xAA). With a 9V supply (PP3), the loading is much less of a problem, and I can get the circuit to produce a reading of about 380V with two neons and four Zeners, oh and the neons glow quite nicely! At 3V the neons glow is quite dim, and extinguishes when the voltage is measured.

I quite like the idea of having at least one neon in the feedback circuit - its glow is a good safety check!

I'm not entirely sure what controls the available power of this circuit yet, which may be critical to getting it to work at 1.2V or lower, which is my ultimate goal, so the pocket unit can run on a single AA NiMH cell. This might prove too difficult a voltage to start from, so I may end up using a 3V supply and finding a way to miniaturize the battery! Using higher voltage Zeners will also drastically lower the component count and physical size of the built circuit - as would using a single inductor in place of the transformer.

The 6.3mm fuse clips for the tubes have been delivered. I expect the 1GΩ 2kV resistor to arrive tomorrow. That will massively assist in getting accurate voltage readings!

There is a variation of this circuit that uses the feedback transistor to control the base bias of the oscillator transistor, which is said to give lower current drain, so I might try this out. I'll try the circuit at just 1.5V as well from a single AA cell, and see if it runs!

Delaying testing the G-M tubes - with good reason

As any electronics enthusiast could probably appreciate - I'm itching to test these Geiger-Müller tubes! But, I've decided to force myself to wait! Why? Well, although I have my scratch built High Voltage divider chain, it is only 100:1, and built from standard 300V resistors. So I've decided to wait until I can get a very accurate voltage reading - which means waiting for the Next Day delivery of a 1GΩ 2kV resistor, coming from RS Components, for the sake of another couple of quid.


I've also had to enter into a dispute with, yet again, a Chinese ebay seller. The 3W IR LED modules I bought to create extra illumination for the trail camera, turn out to be just 1W. Of course, the seller will now try and give me the run around, but I don't play games with these people!

I've also turned down an appalling counter-offer made by seller "dosimeters_radiometers_counters", who believes the 'Best Offer' option is for wholesale, and sent a counter-offer of exactly the asking price! No, its isn't. I thought only the Far East sellers used that dirty trick!

Amazingly, it's actually stopped raining, and there is sunshine! The ground will still be sodden, but I might make a little foray out later, to play with the Radiofix receiver up on the top of a local hill - away from all the electrical crud!

Thursday, 28 November 2019

Another 1090MHz Spider Antenna

Its been some time since I last worked on the 360Radar receiver external mount project, due to giving the PVC radome build plenty of time for the cement to cure. So today I finally got around to fabricating the antenna.

The first job here was to drill a bit of PCB stock, to mount the BNC socket on. This was deliberately drilled a little too small, allowing for creating a keying flat during filing it out to size.


With the PCB drilled and roughly cut to size, the BNC panel socket was fitted. The corners were then cut off and the PCB filed until circular.


The circular PCB is single sided, and so is fitted with the copper facing the socket. This is the 'bottom' of the antenna.


Using my 150W iron, the PCB was tinned, and the antenna elements soldered on. Each was cut a little long, to ensure that there was some play in the dimensions.


With all the elements soldered, a marker for 68mm was made on the jig block, and each element measured and trimmed. The BNC to SMA patch-lead was connected, and the ground-plane elements bent to shape. That done, the coax patch-lead was threaded through the antenna mount on the radomes internal equipment board, and the antenna secured in place with hot-melt glue.


A test fit was made to ensure that the driven element of the antenna would fit cleanly into the spire of the radome.The next stage of this project is to size up the equipment board and drill it for mounting pillars to attach the electronics.


I now have the BOI-33 G-M tubes, and most of the necessary parts for the Geiger counter. Ive ordered the correct sized fuse clips to attach to the tubes (6.3mm rather than the common 5mm), these should be with me by the weekend.


Tomorrow, I am going to bread-board a 3V to 400V zener regulated flyback HV generator, for powering these tubes. I've some 100V zeners on order, which will make regulating these circuits a bit easier, as only four would be needed! The prototype tomorrow will use the only zeners I have at the moment in 'high' voltages, so will be a string of eight 47V devices, plus one 24V unit! I could probably get away with fewer if I took the feedback from after the first multiplier, but the regulation will be that bit poorer. I might try both and see how they do.

Wednesday, 27 November 2019

A more effective 'clicker'

Almost all simple Geiger Counter designs I've seen, use a piezo element or a small loudspeaker, to give the audible 'click'. Ive tested some of these, and to be frank I think they are rubbish! Yes they 'click' but its just the instant of the loudspeaker coil being pulled in. Its not very satisfying.

Since for some of my Geiger builds I'm wanting a simple count indication of this 'click' type, I decided to work out a design that would give a much nicer, cleaner, and louder indication.

The result is a single transistor design, which when triggered by a pulse from the G-M tube, switches a parallel arrangement of an LED, fed by a 1k series resistor, and a low cost miniature 5V buzzer. As this initial circuit is designed for 9V, there is a potential divider made of two 100Ω resistors which provides a roughly 4.5V supply for the buzzer. In order not to draw any current when not needed, the 'ground' connection of the potential divider connects to the buzzer negative, and so to the transistors collector.

To prototype this, I had to rob the buzzer from a cheap Chinese 'pixie' transceiver, that was lying about in the workshop. I also, belatedly, added a series resistor to the transistor base on the breadboard prototype - after destroying one transistor by repeatedly applying raw supply to the base! I also destroyed a few LEDs trying to find a way of using a capacitor to 'extend' the pulse.

The nightmare of finding parts at low cost

Its really no wonder that so many companies are losing out to internet trade, especially when it comes to the hobbyist. Many small electronics suppliers, who im sure most hobbyists and amateurs would prefer to patronise, either dont have an internet presence, or dont stock enough and varied parts. Major suppliers either charge extortionate small quantity postage rates, or rediculous 'handling fees', or both, and/or have stupidly time consuming and confusing parts selection filters, that make it all but impossible to quickly and easily find and order a simple component. And hence, the likes of ebay thrive, as a small supplier, or private seller, can quickly and easily list their goods, and a buyer can equaly easily find what they want.

Of course its not always as simple as that. The plethora of Far East suppliers using off-hand tricks in listing titles, which make it look like a part is available cheaply, but then you find it isnt, and the number of sellers listing parts they dont actually have, makes it a tedious task. Then of course there are those with stupid prices, and barmy postage.

Take this morning. I wanted to buy some 5MΩ high voltage resistors. Why? Ive just invested quite a substantial sum in G-M tubes, and Im not going to risk a standard 1/4w resistor breaking down short-circuit and destroying my tubes!

Finding these on the big suppliers websites gave me a headache. I found some on ebay that were not extortionately priced, and were in the UK, but in smaller quantities than I wanted, or in the wrong values. I even enlisted the help of forum members to try and locate some, which is what worked in the end!

I now have 50 4M7 3kV rated resistors coming. Its cost me more than I wanted to pay, but it gives me a handy surplus. Now, many of the Russian tubes advise an ideal anode resistance of 5M1, but that can be made up with a lower voltage rated series resistor before the 3kV unit, if desired. It probably wont make a lot of difference. But, the presence of the 3kV rated resistor gives the protection to the tube that is wanted. The drive voltage of 400-500V probably means normal resistors will survive, but the trade of is cost of resistor against cost of tube. These have cost me less than 8p each, against a tube cost of £10-20 or more. I can offer some of the surplus to other builders - who would shun 99p inc. postage for the protection and piece of mind? We'll see!


Monday, 25 November 2019

Tubes and Parts, and waiting

OK, so now I have on order the Russian Geiger-Müller tubes - two BOI-33 beta/gamma types (SBM-20 equivalents), the dinky SI-19BG alpha tube, and the whopping STS-6 beta/gamma tube. I believe some less sensitive SI-3BG glass beta/gamma tubes are also coming my way. I also have ordered high voltage transistors, a load of 2N3906 PNP transistors, plus various other parts. These will be used in the counters, but will mostly just bulk my stock of general purpose components.

Ive also ordered some sub-miniature 1:1 audio transformers. These will be used in the HV generator circuits, but I also use them in such things as digital mode transceiver interfaces, and other audio projects, so they are always worth having!

Although the current HV generator works, further reading suggests that using the transformer on it as a transformer rather than a dual inductor package, might well be more efficient. This hopefully will also get me to the 'holy grail' of this project - A 400V G-M tube supply from a 1.2V AA NiMH cell!

I now have to await delivery. The 3W IR illuminators to help light dark areas of the trailcams coverage have arrived today - no instructions and no hint as to which terminal is which (other than making the assumption that the one with the most copper attached will be negative!), as also have the BN73-302 pig-nose ferrites to wind the transformers for the Wellbrook Loop clone amplifier.

Of course, once I build a Geiger counter that works, I'll need to prove that its working! The beta-lights in the Clansman equipment might work for this, as perhaps might the dial illumination on my Silva Mk4M compass, but to be on the safe side, I've also ordered a 2% Thoriated Tungsten TIG welding rod!

Thorium Series (courtesy Wikipedia)
Thorium is the most abundant naturally radioactive element on earth. Its long half-life is a touch longer than the age of the known universe, and its slow alpha decay leads to the above decay chain, providing alpha and beta particles to test Geiger counters with! And a Thoriated welding rod is a damn sight safer to handle than old Radium watch dials!

Sunday, 24 November 2019

Russian G-M tubes

Any brief search on ebay for Geiger Tubes, will bring up a whole raft of different Russian devices. If you forget to put the word 'tube' you'll get a whole load of shoes...

It seems the Soviets were a bit obsessed with G-M tubes! They can be had in all manner of forms, sizes, and sensitivities! As the photo below (courtesy pocketmagic.net) shows

Soviet G-M tubes
It would seem that the SMB-20 tube (4th and 5th from top) is considered the standard for low rate homebrew counters, and so this will be one of the tubes I will try. I have on order though a huge STS-6 (2nd from top) and a tiny little Alpha capable device, the SI19BG.

SI19BG Alpha window G-M tube
Ive now made the HV generator permanent. Its not a winner in any design elegance awards! I've literally just pulled the components off of the breadboard, and cobbled them together on a bit of Perfboard! Ive made one change - the 10mH choke has been replaced with a 25mH/20mH transformer taken from an old switching supply. Ive wired the windings in parallel, and I'm able to just about get 530V from it now. I don't know how, or even if, using both windings in any way makes much difference. That's something I can investigate later.

HV Generator Board, and HV probe board
 With the 500Ω single turn preset, setting an exact voltage is very tricky. Future builds will have proper 100Ω multiturn presets, but for now I've managed to set this one to near enough 430V, which is the ZP1481s voltage. The SBM-20s I believe run at 400V.

One of my main goals with these circuits, is that it should be reproducible using 'junk-box' parts. Now, I have spent quite a bit on parts for this - but generally this is in bulk, and so building up my 'junk-box' further! The electronics for a simple Geiger counter should, ideally, cost nothing to any constructor with a reasonable stock of common or garden parts. The expense should all go on A) the G-M tube, and B) the case to put it all in!

HV Go, G-M No Go, And Inside the Euromarine Radiofix Mk5

Late last night, I worked up the Geiger HV generator on breadboard. The first attempt didnt give me more than the supply voltage! Clearly, I'd got something wrong in building it, so I pulled everything out and did it again - this time it worked!

Parts ready to prototype, and the HV divider board
Using my newly built HV resistor chain, I was able to set the HV generator up to show 43V on my DMM, which equates to 430V. This is in the middle of the Mullard ZP1481 tubes 'Geiger Plateau'. I didnt have a 100Ω preset for the voltage setting, so was using a 500Ω with a 100Ω fixed resistor in parallel. This was very temperamental, and hard to adjust! Ive some 100Ω 10-turn presets on order which will make that much easier!

HV generator, and Mullard ZP1481 G-M tube
The 4M7 anode resistor was connected right on the tubes anode pin, and a 2x100k potential divider in the cathode to ground path was provided to allow me to connect the oscilloscope to look for pulses.

Sadly, not one single pulse was detected! After a lot of checking and testing, including with various 'beta-light' sources, I've sadly had to come to the conclusion that the tube is dead. I will of course retain it and test it again once I have another working tube and a working counter circuit. I have Russian tubes on order, but they will take some time to arrive.


With the intention of maybe retuning it for better coverage of the Aero-NDB band, I've also had the Euromarine Radiofix Mk5 open. Internally, the most prominent feature is the ferrite rod antenna. From the wiring I can see that there are two coils wound together on this. The circuit itself looks to be a very simple superhet. There are three IF transformers, which are all Toko 468kHz units, plus an oscillator coil (Red). Five BC171B transistors and a germanium diode detector.

Ferrite rod directional antenna

Simple Superhet
 Being a 468kHz IF, this rules out retuning to cover up to 550kHz, as this would require tuning through the IF! But, it seems there are few beacons in the UK above 440kHz anyway (not sure of the rest of Europe), so if I decide to retune, it would only be shifting around 40kHz, and still clear of the IF.