Speargun Speed Sensor

I have managed to measure the speed of speargun spears along their whole trajectory through the water. The plots show the energy and the speed, respectively, of two different spears, fired from the same gun. Energy and speed are plotted against the distance the spear has travelled. Energy is important, because that tells you how hard the spear is going to hit the fish. Speed is important, because that tells you how quickly the spear is going to get to the fish.

The first part of each curve is the rapid acceleration of the spear in the gun. The gun has a 90cm barrel. The acceleration happens only along about half the barrel, then the spear starts decelerating because of the water drag.

The magenta curves are for a 7mm x 1.3m spear of about 450 gram. The yellow curves are for an 8mm x 1.5m spear of about 600 gram. Since the 8mm spear is heavier, it is slower initially, but far along the range, it is faster - it keeps its speed better. The energy difference is even more significant - after the first 1.5m, the 8mm spear has more energy than the 7mm. The 8mm gets a smaller initial energy than the 7mm, because it has higher drag, but it retains this energy longer, because of its higher mass.

This example illustrates how measurements like these can help a speargun builder decide on parameters like spear dimensions. To find out how I made the measurements, read on:

Sensor Basics

I used a white line marked with a series of black marks, that runs through an optical sensor. The line is attached to the spear. The signal from the sensor is recorded via the input to a computer sound card.

The recorded signal for a shot sounds like this. The first 50 milliseconds is plotted below:

The valleys represent the black marks on the line. Every black mark is unambiguously identifiable. Note that the spacing between pulses becomes less as the spear gains speed. The signal was recorded at the maximum sample rate of the sound card, namely 44100 samples per second. I calculated the speed by counting samples between consecutive pulses. The period that elapsed between consecutive pulses is the sample count divided by 44100. The distance between marks, namely 3cm, divided by the period, gives a speed reading for that instant.

The measured graphs are not perfect. The small fluctuations are errors. The spear speed cannot fluctuate like that, but the line can perhaps. There are other sources of error as well, including imperfect spacing of my marks. Nevertheless, the curves are good enough to give us a very good idea of the spear behaviour in the water.

Sensor Detail

The sensor consists of a phototransistor and an infra red light emitting diode (LED) that are mounted in a solid polymer housing, which forces the marked line to pass close to the components. The housing is made from 30mm delrin (or acetal) which has a low friction coeficient. The acetal was drilled on a small lathe. The hole is 2.5mm at its smallest, but flared to 13mm at both ends. The phototransistor and LED have 3mm diameter.

The marked line is 1.5mm solid-weave dyneema. The dyneema also has low friction and has very low stretch. (It's also very strong.) I marked the line using a balck felt-tipped laundry marker pen. I wound the whole line around a spear, and stuck some masking tape along the spear, exposing only a portion of the line along the side of the spear. The exposed part was marked.

The LED is powered by a 9V battery. A series resistor limits the current through the LED. I used a SFH409 high-output, infra-red LED. (Its output light is invisible). It can draw up to 100mA. The LED has a voltage drop of 1.5V, so the resistor can be as small as (9-1.5)/0.1 = 75 ohm. The LED anode is connected to postive.

I used an SFH309 phototransistor. It is fast enough, has high enough current output and is well matched with the LED. The phototransistor is powered by the same 9V battery, also with a resistor in series. The collector of the phototransistor is connected to postive. In this case the resistor is not to limit the current, but to convert the current through the transistor to a voltage that can be input to the sound card. The phototransistor has a very low current flowing through it when not illuminated. Since I used white acetal, some light still reflects off the inside surfaces, so the transistor does not completely swith off when it sees a black mark. It has a current in the order of 1mA flowing through it when light from the LED reflects off an unmarked white part of the line. I used a 1 kilo Ohm resistor. That gave me a voltage swing between white and black of about 1V.

Datasheets for both components can be found at Infineon. I got them for less than 1 US$ each.

The two sides of the 1kOhm resistor are fed via an audio jack into the computer sound card line-in. A good place to put the 1k resistor is inside the jack. I used a 10 meter twisted-pair cable between my sensor and the computer. (Most sound cards have both a line-in and a microphone jack. This design is suitable for the line-in. The microphone input expects a smaller signal on most sound cards. Most microphone jacks also have a 5V supply on them which complicates matters further. First test the voltages, before you connect to your sound card.) Line-in is stereo: use just the one side of the jack, or use a mono jack.

Note that the sound card input has a frequency range of about 20Hz-20kHz. (This may vary between makes.) The 20kHz is fast enough for a fast spear shot. The bottom cutoff of 20Hz means that you don't get DC into the sound card. When the line is not moving, you cannot tell from the recording whether the line is at a black mark or not. This may obscure the first pulse of the shot.
Back to top.