As we've discussed elsewhere here, there is a need for some kind of standard that lets us compare results between labs that might be across the earth from one another. From that thinking came the "standard counter" project, documented on some other threads on the board, for the various parts (hard and software, PC and embedded). Joe Jarski and I have formed a collaboration to make this and other things that sometimes seem to be holding others here back, under the assumption that those people just want to get on with their main research, and not have to dive so deep into the details of what it takes to make a quality, reliable product. Our plan has been to do this completely "open source", and we feel that if you want to build your own of any of these things (we have a long list of things to come), that's fine with us. The point is to serve the community and collect good karma, not to get rich off this (as if anyone here was going to make us rich!).
So, this is our first product outing, though there are a few more in the wings coming soon.
This is a pancake geiger, shown here with the protective cover (and beta stopper) on the tube. It is USB powered, so it requires 5 volts on a USB jack to run -- and it's a lot nicer when you use it with a computer and plotting/logging software which we also provide. The particular pancake tube we've managed to get 10 of is optimized highly for beta radiation over all other types, making this perfect to count silver or indium after activating it with your neutron-producing device, and that's its main intended purpose. It's sensitive enough to count background, which it counts in the low 20's of CPM in a few locations I've taken it to, well off the ground (the rocks around here are a little hot). However, the specialty of this is superior beta sensitivity. While it counts less background than the larger Geo pancake, it counts a beta source a lot higher. The gamma to beta ratio is greatly enhanced in this, as I will demonstrate on this thread somewhere. I measured this using an "official" plastic encased Cs-137 .25uC source -- something that emits one beta and one gamma per decay (no alphas, and both the betas and gammas are monoenergetic). With the beta shield in place, this counts around 1k CPM. Without the beta shield, it counts closer to 30k CPM -- so it's about 30x as sensitive to betas as to gammas, making it ideal to count sample activations against a mostly-gamma background (the usual situation).
Here's what it looks like on my dual monitor setup.
You can see when I did what easily on this plot, which is log Y so you can even see the background in the presence of the 30k cpm points. Gnuplot lets you get numerical data back by placing your cursor over a spot, and the cusor position, normalized to plot scale is displayed at the bottom left (I didn't do that this time). If that's not a good enough record, then it will also log the raw data from the standard counter.
Later versions of the open sourced perl PC code will run on windows (and get the line terminations right for that) as well as allow inserting into a MySQL database on any host in your network.
The idea is to build up an overall physics lab data aquisition and data mining facility at pennies on the dollar compared to what's out there now, and tailored for fusor (or other nuclear) type situations. The sharp eyed might have noticed an extra set of colums in the log that are all zeros, or that BNC jack on the box. We provide another counter input channel, ttl type levels but AC coupled, meant for negative going pulses. This is so you (or I) can attach a neutron counter and get both on plots and logs, all nicely in time-sync.
The provided host perl software uses very little of a PC - it was carefully done to never spin in a polling loop waiting for a second to expire, for example, and it uses little memory for reasonable length runs. Just about any laptop is overkill for this.
We are not using much of the PIC microprocessor here at all. It's loafing at about 1% cpu load most of the time, and most of the memory and I/O are "left over". That's on purpose. The main board we designed (I did the schiz, Joe did the PCB layout) has quite a lot of capability, and will be used in other things we make available here. Here's what it looks like inside.
Note the dual row of "spare pads" on the edge of the board. Thats' the leftover IO from the PIC 18F4550. It includes regular digital IO, PWM outputs, A/D inputs, interrupts, I2C bus, and power.
This board can be plugged onto application specific other boards, and either power them or be powered by them, providing a uP and an PC interface that can really move some data.
You'll notice that we designed this in modular fashion -- all the geiger-tube-specific stuff is on its own board, so if that should change for some reason, you only change that part.
As we have only a limited supply of these excellent Russian tubes, and since this is really specialized for counting activations that make betas, we'll price it like this:
$200 + shipping for all comers
$180 for board members or fusor.net "neutron club" people.
$150 for board members that have an actual neutron producing thing to make the proper use of this with.
These will come with a calibration source (and we will log the response here by serial number) and a sample of silver foil to test-activate and count. We are using a thoriated lantern mantle for the calibration source, since one nicely covers the active area of the pancake, as does the silver foil sample. You'll have to make your own neutron oven, a sample is described here.
If anyone just wants to make their own, feel free. All the information to do so is posted on various forums here already, including complete source code for both ends of the communications link. I'd like this to be GPL-v2, meaning mainly that if you make some improvement, I'd like you to share it back to us all.
(note to self, edit and put links to the other stuff about this here)