Analogue input can measure voltages from zero to 5V in 1024 steps.
So each step is 5/1024 volts. This will probbly not be enough for you given you have a measurements to make over three decades.
I think it will be very difficult to measure much less than about 10^-12A (1pA) with the TS271, for two reasons:
-
The input bias current of the TS271 is about 1pA. So it will be difficult to distinguish between a reading due to the input bias current and a reading due to ionization.
-
1pA across a resistance of 1Gohm gives a voltage drop of 1mV. This is quite small, comparable to the typical input offset voltage of the TS271. However, you can add a trimpot to the TS271 to cancel out the input offset voltage, as shown in the datasheet.
I looked for op amps with less than 1pA input bias current. The best I found have input bias currents of around 300fA, for example AD8663. This has a minimum common mode input voltage of 0.2V, so you would need to connect the bottom end of the 1G resistor to the diode bias arrangement I suggested earlier. However, I also found the TI INA116 instrumentation amplifier, which has only 3fA input bias current. This would allow you to measure much lower currents, but it is designed for +/-15V dual rail supplies.
You could consider using a resistor greater than 1Gohm for increased sensitivity. Another possibility is to integrate the current over time, by using a small capacitor in place of the resistor, with a reed relay connected in parallel so that you can discharge the capacitor from time to time.
To convert 0 to 10^-12 A into 0 to 5V for the Arduino, using a 1G ohm resistor, you need a gain of 5000. So you could use a 1K resistor from the inverting input to ground (or the diode bias if using AD8663), and a 5.1M resistor from the inverting input to the op amp output. I suggest you also connect a capacitor in parallel with the 5.1M resistor to reduce the bandwidth and hence the sensitivity to noise. You will need to trim out the input offset voltage of the amplifier, either by using any means provided on that particular amplifier, or by adjusting the bias applied to the inverting input. To be really sophisticated, you could generate the bias from a D to A converter controlled by the Arduino, to allow automatic zeroing.
One way of covering a wide range of input currents would be to have several ranges. Set the amplifier gain to suit the least sensitive range, and add a second-stage amplifier built around a standard op amp. You can use a mosfet as a switch to switch in different resistors to vary the gain of this amplifier.
You might want to add a SPDT switch or similar that allows you to switch the outer electrode of the chamber between +9V and the bottom end of the 1G resistor, so that you can measure the reading when there is no ionization. You could even use an electronic switch under the control of the Arduino, so that you can re-zero the instrument programmatically.
It sounds to me that this is more of an electronics project than a mechanical engineering project.
wow 
I think I need to show this to my FYP supervisor. He probably wasn't aware of how complex the electronics would be when he suggested this topic.
My groupmates checked the math again, and we expect background radiation to give us a current reading of around 10^-16 Amps. And that the max current we could get, for pretty high radiation, would be 10^-14 amps. So a current of 10^-12 amps would be a bit optimistic.
Commercial detectors on the market all do deal with this level of current; so they must have some very intense, well designed circuitry to do it. Something that is probably out of our capabilities it would seem 
The circuit here looks pretty simple:
http://madscientisthut.com/wordpress/?s=radon&submit.x=0&submit.y=0&submit=Search
Except for that we can't get one of the components (The MPSW45). Surely this circuit does the same thing that we want of ours, but less precise? Considering he should be dealing with similar currents, and is getting voltage results in mV.
EDIT: and one more:
The sensitivity of those circuits will vary enormously with the gain of the MPSW45, and there is no indication of the leakage current of the MPSW45. But if all you want to do is show that you can detect an ionization current rather than measure it, and if the ionization current is greater than the leakage current of the MPSW45, then they may suffice.
It's very hard to say how sensitive those circuits are, because they are using the MPSW45 at much lower currents than they have specifications for. So the current gain of the MPSW45 will be much lower than the specified 25000 to 150000, which is measured at 200mA collector current. I would be surprised if they can detect currents as low as you are looking for.
Instead of the MPSW45, you could use 2 NPN transistors connected in the darlington configuration. Use transistors designed for high current gain at low current, such as BC548C. This might actually work better than the MPSW45.
These chambers should be dealing with about the same current levels and changes that we are.
The thing is, if the circuit is too complex, then it's going to be very hard for someone with my lack of skill to build, and impossible for me to debug if the schematic is incorrect.
So what I'm really looking for is to just be able to tell when the current rises and falls. Even if not very precisely. And I should be able to send that voltage into the Arduino board.
Lots of these do it yourself detectors have circuits that work and aren't so hard; it would be ideal to use one like those
Even several patents outline circuits:
But those do look complex.
I really just need a schematic that can get voltage changes with current changes in a range that Arduino Uno can process, and one that isn't too complex to build and has components that I can get from RS Components Hong Kong. I can always get the board made from outside (the professor has given us permission to do that), but in order to get the board made for me I need to provide them with a schematic and the components.
If I need to sacrifice some precision, so be it. It's an ME project; really our focus is on the design of the detector and the airflow through it with the fan. The circuit shouldn't be so hard and shouldn't take up so much of our time.
In that case, I suggest you try 2 x BC548C connected as a Darlington pair, with a series resistor and DMM to measure the collector current, as in the second link in your previous post. Don't bother with a circuit board, you are better off with air wiring for that part of the circuit in order to minimise leakage. If you find that is sensitive enough (i.e. you can get significantly different DMM readings with/without radiation present), then you can develop that circuit to feed an Arduino, for example by adding an op amp to amplify the signal.
If that isn't sensitive enough, then I suggest you use the INA116 and 1G (or greater) resistor, which you can power using a 5V -> +/- 15V DC-DC converter (or even two 9V batteries), and an adjustable bias (trimpot for simplicity), again followed by an op amp.
The circuit shouldn't be so hard and shouldn't take up so much of our time.
I would suggest that you buy an instrumental amplifier instead of making one. If you look at the price of them you will see why you will struggle to make one yourself.
I know amplifiers have moved on a bit since I tried measuring the open loop gain of a 741 way back in the 70's. I used a totally screened room and was still troubled by the interference it picked up. Very high impedance inputs are wide open to pickup the tiniest bit of radio frequency interference and that will probably swamp any signal you have to begin with. This is a hard core electronics project and I don't think your supervisor has any idea of how difficult it is.
The BC548C is not available here
Would it be asking too much to ask you to try and design a circuit that would work based on what I can find here:
? I realise this is asking for a lot; I would do it myself if I could. I understand if you can't/don't want to. I'll try and visit one of the technicians here and sit with them and see if they can help me with designing and manufacturing it, or if they know someone I can talk to. Probably meet the professor and see what he says, or look on a dedicated electronics forum.
Similarly, if I am to buy an Instrumental Amplifier, which one should I get, based on what I can find on that site?
Grumpy_Mike:
The circuit shouldn't be so hard and shouldn't take up so much of our time.
This is a hard core electronics project and I don't think your supervisor has any idea of how difficult it is.
The thing is, if this is so hardcore, then how come so many DIY sites have schematics and simple circuits that are so easy to build and work? What exactly is the difference between those circuits and what we want to achieve. Can a middle ground be reached?
I need to know this before I tell our supervisor what the problem is. Even my own groupmates think I'm doing something wrong now, as they see those websites and are wondering why I'm suddenly panicking about how hard this circuit is when it looks so easy there.
Looking at a random "build your own Radon detector" site, the detector counts pulses, doesn't try to measure the current flowing. I would expect that if you abandon the need to accurately measure and just count alpha particles you detect, you can still calculate the radon concentration in your detector.
we expect background radiation to give us a current reading of around 10^-16 Amps.
Based on that I think it is hard core. I don't see how those other designs can can cope with such low currents.
nchemlani:
The BC548C is not available here
They have BC547C, that one will do.
nchemlani:
Similarly, if I am to buy an Instrumental Amplifier, which one should I get, based on what I can find on that site?
They have the INA116.
Grumpy_Mike:
we expect background radiation to give us a current reading of around 10^-16 Amps.
Based on that I think it is hard core. I don't see how those other designs can can cope with such low currents.
I agree, I can't see how those circuits can detect current as low as that. Unless that is the average and the current actually comes in much larger pulses.
Maybe that's an error with the math then.
I've asked the group to try and get a basic chamber up and running ASAP, so we can plug it into a power source and a super sensitive ammeter to see what currents we get from background radiation and with a weak alpha source, to figure out the kind of range we are actually dealing with.
I will try and go and build the madscientisthut detector. We already tried building it, but failed, because since we couldn't get the MPSW45, and my groupmate ended up getting the TIP122 when I asked him to talk to local suppliers for a close replacement (I would have done it myself but I don't speak the local language and communicating with them in English is really hard). We did't realise the EBC setup of the two pins was different, and ended up wiring the TIP122 in the wrong way. The TIP122 also didn't see a close enough match to the MPSW45 according to the data sheets.
I'll go to the lab tomorrow and try to fix what we have now; solder the TIP122 in the right way and see if that circuit works remotely close to how it should. Will then try to use just the tin can+wire+battery and connect it to the super sensitive ammeter and see what I get.
I've also got an order of 25 TS271 Op-Amps on the way here...what should I do with those now.
nchemlani:
The thing is, if this is so hardcore, then how come so many DIY sites have schematics and simple circuits that are so easy to build and work? What exactly is the difference between those circuits and what we want to achieve. Can a middle ground be reached?
I have been looking at those simple circuits and the thing is that they detect the do not measure. You might think it is the same thing but essentially they work by calibrating out the leakage of the electronics and the effect of the background radiation to achieve a neutral balance point. Then when a radiation source comes close it upsets that balance and you get a reading of that upset. They do not measure the current and so you have no measure of how much radiation is being detected. They work on a very long time constant integrating the small upset in the balance and so detecting very small things. But you see they are not actually measuring anything independently.
Grumpy_Mike:
nchemlani:
The thing is, if this is so hardcore, then how come so many DIY sites have schematics and simple circuits that are so easy to build and work? What exactly is the difference between those circuits and what we want to achieve. Can a middle ground be reached?I have been looking at those simple circuits and the thing is that they detect the do not measure. You might think it is the same thing but essentially they work by calibrating out the leakage of the electronics and the effect of the background radiation to achieve a neutral balance point. Then when a radiation source comes close it upsets that balance and you get a reading of that upset. They do not measure the current and so you have no measure of how much radiation is being detected. They work on a very long time constant integrating the small upset in the balance and so detecting very small things. But you see they are not actually measuring anything independently.
Hmm.
Does the reading increase with higher levels of radiation? In a linear manner? What would happen if I take this reading of the upset, and send it to the arduino?
Right, just tried fixing it our prototype thing; doesn't work. Everything wired in correctly and not touching each other, and it gave pretty much no result. Erratically fluctuated with no pattern, then settled down to one value which wouldn't change for any source. We used the TIP122 instead of the MPSW45, and a 100K variable resistor instead of 100K trimmer pot. Took a video of it in testing if any of you want to see it.
I think you have been told before that a TIP122 is totally the wrong sort of transistor for this job. So it is no supprise.
Does the reading increase with higher levels of radiation? In a linear manner? What would happen if I take this reading of the upset, and send it to the arduino?
No it is the rate of change of the reading that changes with radiation and I do not think it is linear.
wildbill:
Looking at a random "build your own Radon detector" site, the detector counts pulses, doesn't try to measure the current flowing. I would expect that if you abandon the need to accurately measure and just count alpha particles you detect, you can still calculate the radon concentration in your detector.
How do we go about doing this?