The most important measurement to take from the firing of the railgun is the projectile velocity. Given the velocity of the projectile, its mass and the energy stored prior to firing, the efficiency can be calculated. In order to give myself a separate project to work on alongside of the railgun I attempted to design my own optical ballistic chronograph (despite the fact I had access to a commercial one). This was the first circuit of the scale that I had tried to design. The photodetectors work but the digital circuitry is non-functional. Despite that, this circuit served as good practice for the railgun instrumentation board.
The theory behind calculating projectile velocity is simple. To calculate velocity one need only know the distance between sensors, the frequency of the counter, and the final counter value. A small bit of algebra reveals that using easily realized counter sizes and frequencies, the primary source of inaccuracy is the precision to which the distance between sensors can be measured and to lesser extent uncertainty in delay times throughout the circuit. The resolution given by a practical clock and counter is makes an off by one counter error insignificant.
The chicken scratch below is some chronograph design work showing the whole idea better than the schematic does. It is simply a counter turned on and off by the passing shadow of the projectile. Using the distance between sensors, the frequency of the counter, and the final value of the counter one can calculate the speed of the projectile.
The schematic is fairly hard to read due the fact that most pins were connected with aliases instead of actual lines. I believe the manner in which I drew the railgun instrumentation board schematic was much better.
The far right component is a LCD, to the left of that are six LCD drivers that covert BCD values to 7-segment. To the left of the drivers are three dual BCD counters. The Bottom left circuit is a power supply meant to supply stable ± 5 VDC. The positive supply is capable of supplying 1.5 Amperes and the negative supply is capable of 0.5 Amperes. The positive supply was built to handle more current in order to supply 4 infrared emitting diodes that I was going to use for circuit testing. The infrared LED part of the circuit is pictured above the power supply. It turned out that any sort of indoor testing since the optical detectors respond to slight flashing of lights powered with alternating current. The detectors do however work fine in sunlight so I never bothered to use the infrared LEDs. The 555 chip is a low power oscillator, like all the components I chose it operates at low quiescent power. It was designed to operate at 75 Hz in order to drive the LCD. LCDs can not be driven at DC. The average voltage across any given segment of an LCD must be close to zero volts to prevent destroying the LCD. Therefore the ground plane and the input are driven with a square wave. The segment is on if the signal and ground plane are out of phase, off is the signals are in phase. The 555 was initially missing a 1 kOhm resistor which is not on the schematic. To the right of the 555 are two quad AND gate chips. These are used to chain the 6 BCD counters together to count from 0 to 999999.
The optical detectors are in the upper left hand on the schematic. They utilize OPA380 Op-Amps from the Burr-Brown division of Texas Instruments which are specifically design for this type of use. They function as transimpedance amplifiers with a current to voltage gain of 100,000. Using the data sheet for the photodiode and information on the intensity of sunlight from the National Renewable Energy Laboratory website I determined that a 10 Volt reverse bias across the diode would result in a satisfactory voltage output that did not saturate the Op-Amp. In darkness the Op-Amp output is about 0 Volts since the diode current is at most 60 nA, probably much less. In bright sunlight the Op-Amp output is around 3 Volts and the diode current is around 30 uA.
The output signal from the OPA380s is simultaneously fed directly into the input of one comparator, while the other input of the comparator is fed through a filter and reduced in DC magnitude slightly. When a shadow passes over the photodiode the comparator then momentarily sees a difference in its two inputs and the output then goes high. I used both comparators in single-ended fashion. The signals from the comparators are then fed to SR latches used to activate the counter circuit. The counter circuit is enabled by using an output from the SR latches to control the enable pin on the 20 MHz counter.
The detectors work fine, detecting very fast moving shadows. However when turned on the counter seems to start immediately counting, and the display shows garbage. Quite a bit of troubleshooting, and a new board are in order here.
The board art, apart from a few small mistakes on the schematic that I was able to correct I though I was doing great. Unfortunately I gave the quad And chips 16 pin footprints when they where actually 14 pin footprints. This was the cause of much consternation and a very ugly fix. I would rather have fixed the small schematic errors and the massive footprint error and ordered another board, but I was at the end of the semester and didn't have time.
Here is the spacious Chronograph board with components. Notice the angled 14 pin components that were painstakingly made to fit in a 16 pin footprint. the six AA battery pack is for the +5 Volt supply and the single 9 Volt battery clip is for the -5 Volt supply. The big blue components are EMI suppression polypropylene capacitors rated at 275 Volts, a little bit of overkill there.
If I had to do this again I would keep the same photodetectors, and add a lens to narrow their field of view. I would try to go to a single side supply and use a microprocessor to do the counting and display work.
visitors since June 10th, 2005.
