One of the core elements of this project is the analogue amplifier that is the intermediate stage between the sensor and the Arduino board. Its purpose is to amplify the incoming signal from the sensor to measurable levels and sustain the voltage (that indicates the particle energy) long enough for the Arduino to sample it and convey it to the interface. The simulation of the whole circuit was done in Mindi.
Figure 1: Analog Design Block Diagram
1st: Create the Charge Sensitive Amplifier and simulate it.
2nd: Create the Shaping Amplifier Circuit and simulate it.
3rd: Create the Peak Detector Circuit and the Reset Part and simulate them.
4th: Creating the Power Supply board and test it.
5th: Creating the amplifier board and test it.
The Charge Sensitive Amplifier (from now on called CSA in the text) is the first step towards our goal of amplifying the sensor signal. It is a design specifically made for radiation detection purposes, as we expect weak charge pulses from the sensor. Those charge pulses have to be integrated and amplified, so that the result would be a voltage pulse, fitting to be shaped and measured by the next stages of this design. As it was mentioned, the resulting voltage is going to indicate the energy of the particle that was released. This information is going to be used in the later stages of the design of the interface. As for the structure of the CSA, we can say that the first stage of it is a low noise JFET which drives the second stage, a low noise operational amplifier which integrates and amplifies the signal. The integration is being accomplished by the RC feedback that is connecting the output of the circuit with the input of it.
A schematic of the circuit that was detailed is given below.
The input of the circuit is a 2.2pF capacitor, which can emulate the input of a radon detector. After that we see that we are using a BF256B JFET and after that a resistor network that is calculated below. Those resistors are connected to the AD817 that does the amplification and the pair of C2/R5 is doing the feedback integration.
The main resource for those values as well as the circuit itself was (Arnaboldi and Pessina 2007). As we read it we find out how these values were calculated.
With R1=RA, R2=RB, R3=RC, R4=RD. Through these calculations in this paper we find that R1=1.2kΩ, R2=40Ω, R3=300Ω, R4=4500Ω (here we use 4300 because there was no other value available).
The simulation with 100mV of square wave input is resulting in this waveform:
Figure 3: CSA Simulation
The shaping amplifier is the next stop in the processing of the input signal. Our need in this stage is to make the pulse long enough that the peak detector will be able to “pick it up” and make it suitable for sampling. One of the most important things to notice is that we get a negative pulse from the CSA part. So, to get a positive pulse with set characteristics that are given, we are using an inverting amplifier configuration (using the OPA604 chip). To shape the pulse to those set characteristics we are using CR-RC shaping as it can be seen below (ORTEC® 2008).
Figure 4: CR -RC Shaping
We are looking for a pulse of 15 us of width (at half amplitude level), with an amplitude of 1V and we are told to use a capacitor of almost 500pF. This means that the time constant is R*C = 15us.
In the following schematic, though, we can see what we have done instead. The first stage of the shaping amplifier is a base line restore capacitor (bringing the level of the voltage at zero) and zero pole cancellation (capacitor with resistor in parallel) which serves at creating a pulse without a negative love in our circuit. After that we have the inverting amplifier stage, as we have mentioned it and the output of that is sent to the RC stage where we use the same mathematical equation (Spieler 2003) (the result of 5n*2kΩ= 10 us, but is still acceptable as a result).
In the next figures (6 & 7) we see the simulation result with and without these signal manipulation tactics and appreciate the reason we went to those great lengths as to correct the pulse’s shape.
Figure 5: Shaping Amplifier Schematic
Figure 6: Pulse output with negative lobe
Figure 7: Corrected result
After creating the desired pulse, we now need the final stage of our circuit, the peak detection. The peak detection is a form of amplitude measurement, as we will find the maximum voltage that is corresponding to a certain input and through that, we can identify the energy of the particle. To be clear, the input we are using as a test input is a 100mV square wave. This is corresponding to a 5 MeV particle.
The functionality of the peak detector is discussed below. As you can see in figure 8, the peak detector is a circuit that is consisted by two operational amplifiers (CA3140 in our case). The shaping amplifier’s output is connected with the non-inverting input of the first amplifier of the peak detector. The first amplifier compensates the voltage drop that happens by the diodes D1 and D3, whereas D2 is in cut-off as long as we provide a positive input to our peak detector. The provided voltage then is going to be at the input of the second operational amplifier and the voltage is going to be sustained by C5. The pair C6 and R10 is designed to eliminate the leakage current in the circuit and with the pair C7 and R12 they are working as time constants that are given by the manufacturer for the optimal result (Analog Devices 2008).
There are two more elements of interest in the circuit.
The first is the resistor R2 that acts as feedback compensation. This correction was implemented because we noticed that the peak detector after getting the initial value, started to increment its voltage output. This meant that there was some feedback in the circuit that needed to be corrected.
The second is the transistor connected to the capacitor C5. This circuit acts as a switch that discharges the capacitor, thus bringing the voltage to 0 and our circuit ready for another sampling of voltage. It is, as we say, a reset circuit. The Reset signal is a high pulse (5V) given by the Arduino board though the base of the transistor (BC547B) though a resistor of 5k. The simulation of the peak detector with the reset signal can be seen in figure 9.
Figure 8: Peak Detector with reset
Figure 9: Simulation of peak detector
For our board to work we need to provide ±5V. For that reason we were given 2 5V DC adapters. To get the wanted voltage what we have to do is connect the ground pin of both of the connectors together. We shall set those as ground. Then if we measure the voltage we are getting from the 2 connectors we will find that we are getting +5 or -5V. The schematic, board and result can be shown below.
Figure 10: Schematic of Supply
Figure 11: Supply Board & Measurements
To make our lives easier, we designed the board to a freeware layout program for strip boards. The advantage of this procedure was that by doing this we were almost ready to construct the board with minimal problems. In the end we had quite a few problems as it was but if this program was not used, then the assumption would be that the problems would be much more in quality and quantity.
The schematic as well as a picture of the board is given below. A complete photo log of the procedure can also be found in the Appendix. The verification of the board in the form of oscilloscope snapshots will be given just below.
Figure 12: Analogue Board & Layout
Figure 13: Signal Generator and CSA output
Figure 14: Signal Generator input and SA output
Figure 15: Shaping Amplifier pulse width
Figure 16: Shape Amplifier and Peak Detector
Figure 17: Peak Detector and reset signal
During the construction and verification of this board we encountered several problems.
One of the most important was that we had a feedback issue in the peak detector that made the output voltage to rise for a period of time until it reached the saturation value of the amplifiers (2.7V). This was fixed by adding the feedback correction resistor as it was discussed in the peak detector circuit.
The other issue was that despite the design and the careful soldering of the board, the reset signal travelled towards the input of our board, making it not stable and forcing us to make a backup CSA circuit in another breadboard by using the 2SK209 JFET.
Appendix – Photo log of prototype construction
Analog Devices. (2008). Low Cost, 80 MHz FastFET Op Amps. s. 24.
Arnaboldi, C., & Pessina, G. (2007). A simple charge sensitive preamplifiers for experiments with a small number of detector channels. IEEE Nulear Science Symposium Conference Record , 5.
ORTEC®. (2008). Introduction to Amplifiers. s. 12
Spieler, H. (2003). Front-End Electronics and Signal Processing II. Itacuruca: ICFA Instrumentation School