Data Logging
This graph shows a volts data log from the Race of Champions 2005. Each peak represents the coasting into the chicane entry (which allows the voltage to recover slowly). The pit stop is clearly seen (no volts at all!!) and then our weak battery pack can be seen showing its colours. We actually thought it was the drivers fault and pitted him for a change. As it transpired, the pack had lost capacity over the winter (despite being kept fully charged) and we limped along to a gratefully received 4 th place. To get a reasonable resolution we use a 20 volt zener diode to ‘suppress zero', leaving the top 5 volts to give a good range of values. This method is not entirely linear but it is the trend we are interested in this case.
This graph shows current drawn around Goodwood. You can see the coasting phase as we run into the chicane and the high amps peak as we accelerate back up to cruising speed, followed by the rest of the course. The ‘hills and vales' are clearly seen on the graph and illustrate how easy it now becomes to choose a suitable ratio for a given course. That is why we always try and complete a few practice laps before every race. Power removed from race batteries is easily replaced in the first hour of the race.
This graph shows our unbelievable (but TRUE) current draw at the Surrey/Sussex/Hants heat. Note the relative absence of ‘hills and dales' peaks and troughs compared to the graph from the ROC. We were basically pushing against a ‘wall of air' and after 1 ½ hours, the motor cried ‘enough'. By adding higher CD (coefficient of drag) curves to the excellent graph on the Turbo Tortoises website – (with thanks to the team for the original Excel file) we were able to ‘plug-in' our speed and current, assuming other things were probably similar to TT. You can see that 3 extra mph does really cost another 500 watts. F35 teams take note; you won't go significantly further than a good F24 until you reduce your aerodynamic losses to their level. Good luck trying though!!
This is a graph from the modified Excell spreadsheet supplied by Team Turbo Tortoise (many thanks!!, see http://www.portaprompt.co.uk/greenpower/index_page0002.html ). We added two extra curves for higher CD values. By plotting on the graph our known speeds and current draw we were able to deduce that out CD was pretty much 0.30. This explains why it cost us so dear at the 2005 Surrey heat and why we used the whopping amps we did in pursuit of greater speed. Teams who are getting close to 30 mph average speeds need to study this graph very carefully if they are not to destroy their motor too…
 |
 |
| Hall effect current sensor and psu | Basic data logger, enough to log 32,000 readings as it stands. |
 |
 |
| Memory expansion data logger, enough to log a very large number of readings but a good way of using up smaller memory chips. | Serial LCD screen (standard Picaxe item) good for 2 rows of 16 characters, but the drivers seem unable to remember or read what it says, hence telemetry is the next step!! |
This photo shows our new ‘home made' data logger board. It has been designed to accommodate the standard Picaxe memory expansion board (available from Rapid Electronics, picture here) should you wish to log a lot of data. It includes a pair of spare outputs to use as you wish, probably most use as a green/red bi-colour LED in the drivers field of view. You chose the levels to switch it on green or red within your data logging programme, based on your practice laps, and it can tell your driver when to change gear up or down. Outputs for the Picaxe serial LCD and the four A-D inputs are on-board, one of which can be configured as a digital temperature sensor. We have ‘killed' far too many of the digital temperature sensors and will probably move to a non-linear but electrically robust thermistor. Linearity can be dealt with in software, although for motor cooling the temperature is the key question, not the actual temperature, if you see what I mean!! A low drop 5 volt regulator is included on the board, as the A-D converter divides whatever the supply voltage is by 256, so any voltages derived from accurate sources maintain their accuracy. 3.5 mm stereo jack sockets and plugs provide a quick easy way to connect and disconnect things like an LCD or sensors (+,-,signal). You can use 4 AA alkaline cells to power the unit, but we chose 6 AA 2000 ma/hr NiMH re-chargeables.
To measure current we use a ‘contact-less' Hall effect current sensor made by Honeywell and available from Farnell. It needs it's own +/- power supply and this is taken care of by a ready made module from the same source. As shown, it produces 0.1 volts per amp that flows through the wire passed through the hole in the unit. PCB designs (PCB Wizard) for both these units are available by email (tsr@trinity.croydon.sch.uk ) or ready built for a small premium. Contact us for cost and availability details.
The ‘G' sensor we use is mostly for interest although it does have a real purpose. Divers, especially on tight circuits like Bedford Autodrome, can seriously comprise the tyre life by aggressive cornering. We were able to distinguish a good driver from a ‘needs training' driver from the peak ‘G' recorded. We will share details of cost and availability should anyone be interested, just ask…. The part is normally used for active suspension control in road cars and is ONLY available in surface mount. Consequently, it's a bit of a pin to solder in!! It is made by Analog Devices and can be purchased off the web for a few euros.
Picture shows our dual axis ‘G' sensor, +/- 1.8 G – the biro tip points to the actual chip! The most we have ‘pulled' so far is 0.6 lateral as we cornered at Bedford Autodrome.
This graph shows lateral ‘G' for most of a lap around Goodwood from the National Final 2005. See if you can link it up with the real bends!!
Finally on the data logging front, Everyday Practical Electronics magazine has recently published a speed camera warning unit. Using a GPS receiver to produce location information, this data is then compared to a built in database of speed camera locations. When a match is found, a warning is issued to suggest the driver slows down. These modules are a mere £40 each and output all sorts of data, not least current speed, every second. Our intention is to log this speed, look at the data compared to different drivers, and then decide what it shows…….
We are also looking at telemetry, but realise it will be short range as the car goes past the pits (e.g.) so we are not sure what we can get across in the time, certainly not a whole laps worth of data, but quite possibly average speed for last lap, average current, voltage etc followed by a reset command from the pits. Might be a trial to get it working…we shall see.
 |
 |
 |
 |
Kent Regional Heat 2006
Amps Average:16.07
G Average:126 units
Amps Average:16.7
G Average:127 unit
Top graph is TSR-2 current from a practice lap, bottom is GPS Speed, both lined up.
This shows average of under 20 amps, so we upped the gearing and got a Silver by 7 miles!! No data for the higher gearing as the loggers both went out on TSR-2.
.jpg)
TSR-2 practice lap from National Final 2006 (true data)
