A long, long time ago the folks at bildr.org got their hands on some nixie tubes and some components from ogi lumen, and built a twitter follower counter. You can see the post here: http://bildr.org/2010/11/twixie/ At the end of the post, they outline a plan to give away a set of nixie tubes and all the fixin’s required to work them, as well as a voucher for laser cutting from ponoko.com. Well, it turns out I won that contest!
After a fair amount of designing and redesigning, I decided to make my geiger counter look like an old-timey cathedral radio with a detachable wand to check various sundries for radiation. It was a bit more ambitious of a project than I realized at first but I think the results were well worth the effort.
Step 1: Materials
Well, there's a lot of things you'll need and I probably forgot some, but here's the main parts:
Nixie tubes, drivers, and power supply
-electronic components to build a power supply for said Geiger tube
-various electronics tools, such as a multimeter, alligator clips, soldering iron, etc.
-an oscilloscope is helpful
12 V wall wart/power supply
multiposition selection switch + knob
curly phone cable + 2 jacks
clear tube to use as a wand for the geiger tube
-router and router bit
Money for laser cutting
Step 2: Get the Nixie Tubes Going
The Nixie Tubes I used were generously provided by ogi lumen and bildr. This project uses 2 Nixie Duos, 2 Nixie Drivers and 1 Nixie Power Supply. These come as kits and require a bit of soldering, but it's all through-hole parts. There are instructables already written that detail the assembly of these kits, so I’ll just link you to them instead of rehashing it:
Ogi Lumen has an arduino library written for controlling the tubes which you can download from their website. This makes testing your assembly easy enough and you can have some fun with the nixie tubes before you get everything else working too.
Step 3: Get the Geiger Tube Working
I bought an SBM-20 geiger tube from ebay for about 25 dollars. It's one of the cheapest you can find, but it works well enough for this project. It needs about 500 volts to work but the amount of current it draws is miniscule. You will, of course, also need a way to register counts. I adapted the second circuit on this page: http://www.techlib.com/science/geiger.html and also the schematics from Sparkfun’s Gieger Counter: https://www.sparkfun.com/products/10742? These will provide your 500 volts and also the readout. Be particularly careful with this circuit, 500 volts can give you a pretty good zap, and the capacitors used in the voltage doubler section will hold charge for a few minutes after you unplug it too.
Quick tip, I found that fuse clips, if bent out a little, were just right to hold the geiger tube and make electrical connection to its ends.
Testing the voltage can be a little tricky. Most multimeters don't have enough output impedance to be able to read the voltage from this circuit without dropping the voltage substantially. I put ten 10 megaohm resistors in series with the multimeter to act as a voltage divider, and then multiplied the voltage I read by the appropriate amount. It's not the most precise way of doing things, but it gets the job done.
Once the high voltage is going, you'll want to test that it's actually counting as well. An oscilloscope comes in pretty handy here, but you may need a radiation source of some kind so that you get a high enough count rate for the trigger.
I've drawn up a schematic in gEDA and added it. A couple of notes: the three 0.01 uF capacitors in the voltage doubler section and the 47 pF capacitor should have a voltage rating greater than 500 volts and probably about 1000 volts. Also, the MPSA42 transistor was chosen for its high voltage rating; a regular old 2N3904 definitely won't cut it there.
Step 4: Put All the Electronics Together
At this point, all the main electronic components are assembled and tested. Now is the time to put them all together and test everything! The pinouts for the nixie tubes are all included in the .pdf's from the Nixie page. Pins 3, 4, and 5 on the Arduino are the clock, latch, and data pins respectively. Those are specified in the first few lines of code in the Arduino sketch, so they are easy enough to change. My arduino code is included as a zip file; it's adapted from the example code that came with the nixie libraries from ogi lumen.
The output from the geiger tube goes to pin 2 on the Arduino. There is an interrupt attached to this pin, which picks up a falling edge and triggers that as a count. I added a little piezo so that I would get a nice click for every count.
It's nice to have everything connected and working together, but it's no time to rest on your laurels, there's plenty more to do.
Step 5: Move the Geiger Circuit to Its Wand
Now that everything works, take it all apart. A breadboard just won't cut it for this project, so I moved to perfboard. I put the geiger circuit in it's entirety into a wand so that I wouldn't have to worry about 500 volts on an exposed line anywhere and so that the power would be close to the tube. It was a little bit of a tight squeeze though, and really ought to have been done with PCB instead. That's a project for another time though.
The wand itself was a plastic tube that came full of bay leaves or oregano or some such. I liked the clear tube so I made a few modifications to suit my purposes. There's some slots cut in the end, over the geiger tube itself, in order to increase the acceptance. You'd be surprised how much radiation a few millimeters of plastic will absorb. There's also a slot cut into the lid so that the perfboard can poke through and the jack can mount onto that.
I used one of those curly phone cords to carry the signal in to the Arduino. Pay attention to the wiring of this cable; when I put everything back together it took me quite a while to figure out that I had the lines from the phone cord mixed up. Also, those cords have tiny plastic clips, make sure to get the right size jacks for them. 4P4C RJ-11 is the appropriate designation I think, but it's been a while since I bought those parts.
Step 6: The Enclosure
There are three sections to the enclosure: the base, the laser cut parts, and the veneer. They were all finished with a few coats of tung oil. It gives a really nice matte finish that brings out the natural color of the wood. It's easy enough to apply: just wipe on enough to soak in, let it dry overnight, give it a quick sanding and dusting, then apply another coat.
Step 7: The Laser Cut Parts
Now that most of the electronics work is done, it's time to work on the enclosure. I had the front, back, and supporting struts laser cut from ponoko.com. The total cost for laser cutting, including materials, was about 80 dollars. It's a fairly big and intricate piece I guess. While I was ordering from ponoko, I also got that nifty red rocket-launch-safety switch to use as the on/off switch. It seems they sell hardware as well as laser cutting services.
Ponoko takes vector graphics files; I made mine in Inkscape. If you're not familiar with Inkscape, I highly recommend it. There's a handful of good tutorials out there that will teach you to use it. There are two files for this, one is for the light colored wood (bamboo) and one is the dark wood (walnut veneered mdf). The light files aren't entirely up to date; a couple holes need to be added for the switch, the knob, and the jack. Also, the dark inset pieces on the design are a little too big to fit without some adjustment. I may get around to amending it one of these days.
I added the design files I used as a zip file. They are a little hard to read since the laser cutter needs a very thin line of a specific color. You can "select all" and increase the line size to get a better look at it, but be sure to check ponoko.com for the latest specifications before submitting them for cutting.
It takes a few weeks to have everything laser cut and shipped to you, so you could probably send off for these first and complete all the electronics while you're waiting.
Step 8: Assembling the Structure
The structure isn't too complicated. In the walnut material, there's the back piece, a frame for the front, and the inserts for the central design. The bamboo has the actual front piece, a supporting middle piece, and the struts.
First, I glued the walnut inserts to the bamboo front. The I glued the walnut frame to the front piece. I glued the struts to the middle support next, then the struts to the front assembly, and then I glued the back on. As a backing for the central design, I found an interesting piece of perforated copper sheet. I cut it to size with tin snips and glued a few tabs in to hold it in place.
Step 9: Gluing Down the Veneer
Because of the shape of the enclosure, gluing down the veneer turned out the be the hardest part. I got a piece of really nice wood-backed walnut veneer, so it didn't need any additional support. Unfortunately, it had quite a bit of springiness to it, so it was pretty difficult to get it to conform to the contour of the structure.
What I ended up doing was clamping everything down and gluing just a little each day. I worked around the edge, moving the clamps each time and gluing a few more inches. It took about a week total.
After it was glued down, I just trimmed the excess with an exacto knife.
Step 10: The Base
The base is a solid walnut plank and the edge was routed with a double roman ogee bit. Before the rest of the structure is glued down, It's necessary to make a small support for the nixie tubes inside of the structure. Since they aren't set all the way at the bottom of the enclosure, I cut a couple small pieces from the laser-cutting scrap and glued them in to hold the nixie duo/driver assembly up. Unfortunately, the pictures of the supports aren't very clear, so you'll have to use your imagination.
Step 11: Put It All Together
Now that the enclosure is finished, all the electronics can move inside.
The knob on the front is to select modes. It just passes a high signal to a pin on the arduino depending on its position. There's a mode that just increments for each count, and a mode that estimates the counts per minute. I intend to add another mode or two eventually, since there's some space left on the selector.
The switch routes power from the input to all the parts. Input power is from a 12 VDC wall wart. The nixie supply takes 12 V and the rest of the electronics, which runs on 5 V, draws little enough current that the Arduino's linear regulator can handle it.
It's somewhat of a rat's nest inside after everything is connected together, but it works!
Step 12: Detect Radiation
Here's a video showing it just counting background radiation:
and here's a short demonstration: