I’ve been thinking about what other parts I need to buy for the high-school Robotics Club, so that they can finish their underwater vehicle.
Some I’ve known about for a while, but just haven’t gotten it together to buy the (expensive parts). For example, they will need several IP-68 connectors for the tether power, for the tether CAT-5 cable (see Long USB cable problem solved), for the motors, and for whatever suite of tools they build for this year’s challenge. I bought a 4-pin Buccaneer standard connector (made by Bulgin) for the power tether, and am leaning toward 8-pin mini Buccaneer connectors for the motor and tools. I’ve still not decided whether to get another 8-pin mini Buccaneer for the Ethernet connection, or to try using their IP-68 ethernet connector.
One that I only just started thinking about is a pressure sensor for determining depth. The challenges for the past 2 years have involved going to a particular depth, so having a depth gauge is a good idea.
Pressure sensors are classified into 4 types: differential, absolute, gauge, and sealed gauge. In truth they are all differential, measuring the difference in pressure. The absolute gauges measure the difference to a vacuum chamber, the gauge-pressure sensors measure the difference to atmospheric pressure, and the sealed gauge sensors measure the difference to a sealed chamber that has a known pressure in it (probably around one atmosphere).
If the club members put the pressure sensor sticking through the wall of the dry box, we could use any of the types of pressure sensor, but absolute or sealed gauge would be best, since we don’t know how much the pressure inside the drybox changes as we submerge it. The sidewalls will flex, so the pressure will go up, but we don’t know how much.
I spent some time today looking for pressure sensors at DigiKey. It looks like absolute pressure sensors are cheaper and easier to find than sealed-gauge, so we’ll go with them. There are also price variations depending on whether the sensor has temperature compensation, and on whether the output is a low-voltage differential signal or an amplified one ready for input to a microprocessor A/D input. I’m pretty sure that we need the simplest interface we can afford, so the ones that don’t need external amplifiers are the best bet.
I also spent some time reviewing how much pressure to expect at different depths, so that I could help the club select the appropriate part. I found an on-line pressure calculator that can convert depth of water to pressure (in PSI, kPa, or atmospheres). Each meter of depth adds 9.8 kPa (or 1.42 psi). In a 20′-deep pool, they could have up to 59.7 kPa, and in a 12m deep pool (like at NASA) they could have up to 117.6 kPa. Of course, this is just the water pressure that is added to the atmospheric pressure, so the real range they need to cover is 100 kPa to 220kPa.
It looks like Freescale Semiconductor makes some absolute pressure sensors with a 20–250 kPa range (for example, the MPXH6250AC6T1), which is just about right for us. They run on 5 v and provide a ratiometric output, that we should be able to read to 10 bits of precision, or about 0.25 kPa, which would be a depth resolution of 1 inch. The accuracy of the pressure sensors is only ±1.5% of 250kPA, which is ±3.75kPa, ±15in, or ±38cm. They get full points if they are within 50cm, so this is accurate enough for us. The could calibrate to higher accuracy, by lowering the machine at the side of a pool while looking at a ruler or tape measure along the side of the pool and taking measurements. So, now that I know what is needed, I can have the robotics club try to solve the design problems.
I think I will have to do one more thing for them, though—the pressure sensor is a surface-mount component, so I think that they will need a breakout board for it. I’ll design a breakout board that has room for some bypass capacitors and an output filter capacitor. Although surface-mount, the part has 0.05″ pitch leads sticking out, and so should be hand solderable, though it might be worthwhile to get a heat gun and solder paste to try reflow soldering.
Come to think of it, since the pressure sensor will be glued to the side of the drybox, it might be a good idea to mount the 3-axis accelerometer on the same board. That would reduce the amount of cabling they need to do. But the ADXL335 chip has the SMD pads underneath the chip, which can’t be easily hand-soldered. That brings up the question of whether we try soldering with a heat gun or toaster oven, or pay $12 more for the chip already on a breakout board. (The DigiKey breakout board is the same price and a similar design to the LadyAda breakout board—both are copying the ridiculously overpriced breakout board from Analog Devices.)
I could play around with a few different designs and stick them on my next PC board run—I want to do a revised version of the HexMotor board anyway, and it won’t cost any more to toss in some sensor boards on the same fab run.