(untitled)

>dosim- no results. dosimeter thread. gc ok please help with my project : SBM21 + SPI/I2C screen + Arduino nano + ??? = profit i am fully lost and i don't even think it's doable at this point (you don't just plug a gm tube into a nano and call it geiger)(USER WAS WARNED FOR THIS POST, You should always elaborate on your opinions rather than just spilling the soykaf.)

nice, i put the thread title in the name field. oh well? gc = geiger counter. i have severe questions regarding the pulse detector circuit (?) and the voltage required as well as the calibrating for the GM tube in question. oh and how to actually code the whole thing into the arduino.

Available on Libgen: The KFM: A Homemade Yet Accurate and Dependable Fallout Meter by Kearny Radiation Detection and Measurement by Knoll You can also buy a plug-in radiation detector for a smartphone, but it's probably limited in the kinds of radiation it can detect.

>>11957 oh, I have been working on a geiger counter for the last month or so, learning electronics along the way. I can't tell you much about calibration since I'm not interested in building a dosimeter, I just want it to emit ticks every time there is an event. anyways, with regard to power supplies: the high voltage required to bias a GM tube is the majority of the challenge. this necessarily implies the usage of a switching power supply. thankfully, since the GM tube barely presents a load at all (in fact, it mostly isn't a load, until a couple of microseconds during a detection event), you are working with very very low power so its not actually all that tricky. pretty much all of my knowledge on this front comes from the voltage regulation section of the Art of Electronics and katkimshows lectures on youtube. They are both very good resources. You'll want to make sure you understand DCM boost switchers especially, I'll talk about why later. anyways, rest of this post assumes background there. the design from this article https://www.nutsvolts.com/magazine/article/pocket-geiger-unit is very common to see in various incarnations, but I am not terribly fond of it because the output is essentially unregulated. which is fine I guess, since the tube is very predictable in the sense that it mostly is not a load, but as described in the article: > This transimpedance ratio is the square root of L/C, and with practical components, can exceed 20 kV per amp. A 450 V output, for example, requires a peak inductor current around 20 mA. as far as I understand (and this may be wrong, I haven't been able to find this discussed much elsewhere, but simulation in LTSpice seems to support this) that value C is the parasitic capacitance of the inductor used. to me (with a computer science background) this is like relying on undefined behavior, which I don't like. anyways, I've misplaced the notebook I did my circuit analysis of this in and I'm not particularly keen on redoing it at the moment, but figuring this out is a good exercise in switching power supply analysis. more realistically you'll probably want a more normal switching power supply design with feedback. another thing you will often find is advice along the lines of, oh if you need such a massive voltage gain (i.e 9v battery to 400v), you'll need an excessively high duty ratio on the switch, which will strain the switching components with high currents, so instead of using a boost topology you'll need to use a flyback topology. I never had luck with simulating effective flybacks that can use practical components IRL, that might just be a skill issue (I have had a hard time understand DCM flybacks period, another skill issue, but that seems to imply huge voltage ripple and generally seems like the wrong thing to do), but more importantly: the geiger mueller tube is mostly a non-load! the resistor divider in the feedback network, if fed with the voltage you are biasing the tube at, will absolutely guzzle power relative to the tube, and so your design ends up having to consider the feedback divider the load, which is just ridiculous. finally, with regards to the DCM vs CCM part: as has been repeated multiple times now, the tube is somewhere between not and barely a load. which means CCM operation will be extremely impractical to maintain for any real implementation since CCM always has some minimum load it needs to be operating with to maintain. that leaves us with two problems a power supply design needs to solve: 1, feedback needs to be taken such that it doesn't load the power supply too much. 2, we'll need a DCM supply with a very very very short duty cycle in quiescent operation. for 1. another approach you can take, as an alternative to a flyback transformer, is use a cockcroft-walton multiplier as the rectifier of a boost supply. this way you can have the boost supply put out a reasonable voltage for a boost to handle, take feedback from one of the earlier stages so that current flowing in the feedback network is much lower, and then have a high voltage output at the end of the multiplier. this works very acceptably because of the incredibly low powers involved. for 2. hysteretic operation is usually undesirable, but plenty of experimentation will show that it is about the only control mechanism which makes any sense for this application, since otherwise the duty cycle at a fixed frequency will become impractically short. its better to fix the on period while decreasing the frequency, that is hysteretic operation. that brings us to my two favorite implementations of a geiger mueller tube supply I have seen so far, from Maxim application note 3757 https://www.bristolwatch.com/geiger/pdf/MAX4162.pdf and analog devices circuit note 0536 https://www.analog.com/media/en/reference-design-documentation/reference-designs/cn0536.pdf . Both of these implement some very straightforward power supplies, with an oscillator and a comparator that enables the oscillator when the power supply is out of regulation. simple as. I've built a prototype on a breadboard using a tlc555 and lm311 that I am very happy with and is what will end up on the final PCB for my design. with regard to sensing: the geiger mueller tube acts as a small capacitor. the SBM20, for example, is listed as having a capacitance of 4.7pF on its documentation. during a discharge event, the charge built up on this capacitor jumps to the cathode, and this charge will be what you need to detect. there is always a resistor in the megaohm range at the anode in order to limit current through the tube and cause a voltage drop so that the avalanche of the tube can quench itself as soon as it started, so that it doesn't continue conducting forever and quickly consume the tube. you can detect events at either the anode or the cathode. I haven't had much luck analyzing anode side sensing, but the only example I've seen is in the aforementioned CN-0536, where as far as I can tell there is something ... wrong about it. anyways, cathode side sensing, mightyohms geiger counter kit has served as my favorite resource. I'm pretty much copying this design, which itself is copied from the design that appeared in the nuts and volts magazine, since it doesn't have any problems and is easy to understand: before a detection event, the high voltage plate of the capacitor formed by the geiger mueller tube will charge up to 400V (or whatever voltage you need in your implementation, but here its 400V). when there is an event, this will discharge into the capacitor C2. the charge of a capacitor is defined as Q=CV. charge will not be lost, so after a discharge event, the voltage at C2 can be found via 4.7pF * 400V = 220pF * Vout, and solving for Vout yields 7.6V. this brings it down to a reasonable voltage to use for logic from this point. then, because of the diode drop from the base to the emitter of Q2, we can assume that the base of Q2 is at 0.6V when it is conducting. the current through the base of Q2 is set by the voltage at C2 - 0.6V all over the resistor R4. this helps protect whatever logic you may have after Q2: the 2N3904 is a BJT, and it can handle a fairly high current before it explodes. FETs are much more sensitive. this protection is important since you can't really rely on C2 alone to limit the peak voltage to 7.6V in a practical implementation. R7 is there so that C2 can continue to discharge towards 0V after the diode in the base-emitter junction of Q2 stops conducting. there is another way to look at this, where R2 and R4 form a resistor divider, which would limit the maximum voltage C2 can charge to. this is only a valid way to look at the circuit in DC, i.e. the GM tube is conducting continuously which should never happen and is very bad. thats pretty much it