Those of you who have seen my hobby page on robot boats will know that our more recent robot boat prototypes are using a small, very low cost, PICAXE computer to navigate the boat, based on the GPS.
These brief notes are for those interested in the wind direction sensor. This includes the teams around the World participating in the friendly competition to be the first robot boat to sail the Atlantic. Please visit the "robot boats" page above for more detail, including details of the "Microtransat Race" and my amusing "Snoopy Sails!" video :-)
This page dates from 2011, when we experimented with a wind direction sensor, based on the 360ASMC module. We never did use wind direction sensors in our 2012, 2013 and 2014 robot boats, because we did not achieve significant improvements upwind, in exchange for the increased power consumption. However, in 2014, we have been working on electronics for the 2015 boat, using sensor power switching, to reduce consumption, and so we are using these wind direction sensors again.
when I contacted the European supplier of the old Ametes 360ASMC-01, based in Serbia, they confirmed they were out of production, but also that they had run out of stock of the 360ASM-01 that is still available. I am grateful to GMW in the USA, who were able to supply the 360ASM product, which is practically identical, when used how we use it here. The only difference is not connecting the "configuration pins", and I have updated the pin-out picture below. I have tested the 360ASM product, in a third sensor, and it works without software changes. I had purchased two of the procucts from GSM in the USA, at 24 USD each, plus 42 USD for shipping by the USPS. See details and pricing on the GMW page.
The quick answer is "maybe not" - but this has to be confirmed in our tests on Bray Lake. I've been building a prototype wind direction sensor, ready to be added, should our tests confirm one is needed. Our present solution is extremly simple: just a GPS, connected to the PICAXE computer, which drives the rudder servo.
This solution has already steered the boats around our Bray Lake test course on several occasions. However, this was with the software having knowledge of the wind direction, so the boat could "tack" to make headway against the wind, when needed. For this, we have a manually set potentiometer, plugged into the picaxe where a wind direction sensor might be put.
When we are happy that the existing solution will work reliably in different wind strengths, we intend to test again, but without the software knowing the direction of the wind. It is just possible that this works reliably enough to dispense with the wind direction - resulting in a simpler and more reliable solution to sail the Atlantic.
But if our tests indicate that a wind direction sensor IS needed, then we will add a prototype, similar to that described below, and modify the software to use it. i.e. convert the wind direction relative to the boat, to an absolute direction by adding the GPS heading - already used to steer the boat.
We've looked at several similar products on the market, and what attracted me to the Ametes 360ASMC-01 from Senis was the simplicity of the module. It only has five pins, and you can see how easily it can be connected in the picture below. We are already using the PICAXE, very popular in schools, and costing only about 10 GBP each. So with 20 GBP for a GPS, and 5 GBP for a rudder servo, maybe even a school boy could put together his own robot boat :-)
The other products considered included the AS5040, GMW 25A-10, Melexis MLX90316, and Renishaw AM4096. Some of these are already being used by other Microtransat teams, but the Ametes was the first I found which appeared simple to use based on the online documentation. Some of the other products have many more pins on the chip, and many pages of documentation to plough through.
We soon found details of the Ametes 360ASMC-01 here on the GMW Associates web site. Those in the USA may purchase from GMW who are in the USA, and there is plenty of information about GMW Associates on www.gmw.com. The technical data is readily available here.
For the rest of us, outside the USA, you are directed to purchase through Senis, listed as being in Switzerland. Their web site is www.senis.ch and this is the route by which we purchased sample product, and obtained technical support when we had problems. Both Senis (the distributor) and Ametes (the designer/manufacturer?) claim the same trading address in Zurich, Switzerland. However, other information, such as Sentronis inport/export paperwork included with the samples, indicate an address in Serbia, eastern Europe. See www.sentronis.rs
Staff at Ametes/Senis were reluctant to confirm that they were based in Serbia, rather than Switzerland, but were still helpful in overcoming problems such as postal delays, and fact that some of the samples failed to work. At present, I've put these technical problems down to my not exercising enough care with the sensors, including when soldering, and treating them as Static Sensitive Devices.
Soon after discovering the Ametes product, and details on the GMW web site, we emailed both GMW and Senis, asking for the cost of providing two samples. Senis replied, and after exchange of several emails, they quoted us what resulted in a cost to us of approximately 50 GBP, including postage and bank charges. They do not accept payment by PAYPAL or credit card, and so a transfer must be made into a Swiss bank account.
The picture below shows you how simply the sensor can be used - simply connected to a servo lead.
My mechanical construction here may appear "messy", making use of meccano gears and "fast glass" car filler resin :-) However, I'm hoping it will be robust enough for our tests on Bray Lake - if we get to that stage. I've now put the sensor and magnets into the plastic enclosure of a WH180 replacement wind direction sensor, and you can see a picture at the top of this page.
Only a few lines of code were needed in the PICAXE software, to test the sensor assembly. I've put the relevant statements below the picture.
Here are the few lines used to test the device, including use of READADC to read the analogue voltage from the sensor, and scale it to a value of 0 to 360 degrees.
windtest: 'reading wind direction pot.. readadc 4,w3 'read voltage from sensor if w3 > 24 then 'values below found by test w3 = w3 - 27 'scale for Ametes 360ASMC-01 w3 = w3 * 240 'which gives values 25 to 255 w3 = w3 / 153 'for angles 0 to 359 degrees if w3 > 359 then w3 = 359 endif sertxd ( " ", #w3, 10, 160) 'output value to PC else sertxd ( " raw ", #w3, 10, 160) endif pause forp5sec goto windtest
Accuracy seemed to be better than 3 degrees, looking at values returned for the four primary directions. Even better accuracy might be obtained by adding simple tests in the software, but this is accurate enough for our application.
The Microtransat Challenge: www.microtransat.org
Robin's hobby robot boat page: www.gpss.co.uk/autop.htm
PICAXE products, including Picaxe 08 servo driver: www.picaxe.com
Technical details of the Ametes 360AMSC-01 sensor: www.gmw.com/magnetic_sensors/ametes/360ASMC.html
USA source of Ametes 360AMSC-01 is GMW Associates: www.gmw.com
Outside USA distributor of Ametes 360AMSC-01 is Senis in Switzerland: www.senis.ch
Designer/Manufacturer of Ametes 360AMSC-01 is Ametes in Switzerland (or Serbia?): www.ametes.ch (or www.sentronis.rs ?)
Alternative direction sensor: AS5040 from Austria Microsystems,
Alternative direction sensor: Melexis MLX90316
Alternative direction sensor: Renishaw AM4096.