Something got to be wrong with your setup. As far as I know the MS5611 has similar specs than the BMP085. I'm using a BMP085 based DroidSoar V2 and a couple of hand wired prototypes since about 2 years and the error in IAS is in the order which is to be expected from the specs. The error at low speeds is much higher than at normal flight speeds. That comes from that fact that IAS is proportional to SQRT(delta P). That comes from the Bernoulli equation. In absolute numbers, you can expect an error <= 10 km/h at speed around 0 km/h. At 70 km/h the error is ~ 0.5 km/h and gets smaller as you increase the IAS. So something got to be wrong with your device or the S/W. A couple of things to check: 1. make sure to use the correction coefficients from each individual device. They are uniquely calibrated. 2. make sure the conversions have completed when you read the data. The BMP085 has a EOC signal which can be used for that. 3. double check the calculation. The raw data from the pressure sensors (after compensation calculation) might not be in hPa.
I have a couple of Java programs which you can use to produce Apps running on Android. That is to display the pressure, and to display IAS. I can send it to you if you want. BTW: I've moved to to xcsoar-user. This is probably more appropriate for this. Ronald PS: don't even think of building a pressure sensor from discrete analog components. The result is orders of magnitude worse than a BMP085 for less than 10 EUR. Am 26.05.2014 15:52, schrieb Andy Smith: > I would like to endorse Advi's comments regarding XCSoar. Thank you all for > your hard work which has produced an excellent application > . > I too have an IOIO sensor box, currently with 3 MS5611 pressure transducers > and RS232 FLARM interface. > I chose this approach with the hope that accurate airspeed could be derived > from separate absolute sensors. I have had some difficulty getting the system > to perform a pitot zero calibration with the current s/w in 6.8 preview_2. > Eventually I got some results which showed errors of 10kts at 50kts. This is > far worse than I'd hoped considering the accuracy and stability of the MS5611 > altimeter readings (+/- 2 ft). I'm very encouraged that Advi has worked on > the code and will try to compile it on Eclipse under Windows. > > A long time ago I designed a microprocessor based system (Varcom). It used an > MPX10DP pressure transducer for airspeed with external temperature > compensation by instrumentation opamps (OP27). I had to heat cycle each unit > to calibrate the compensation. Even with this it was necessary to have a > routine that looked for steady asi readings on the ground (i.e. no wind) and > adjust the zero as it drifted over the years. This worked and it was possible > to get very accurate airspeed from the system. The MPX5010DP appears to be > the same transducer I used with added onboard temperature compensation. I > would be very interested to know how accurate your ASI measurements are Advi. > If I can compile your code I will try it on my sensorbox with two MS5611. > > Unfortunately I have a very steep learning curve before I can help with the > coding but I've made a start. However, I have access to gliders in the UK and > Australia and could help with flight testing. If we can develop an > intelligent XCSoar vario there will be a lot of work involved in this area. I > have some ideas for data logging and replay that could shorten this process. > > Thanks again for your hard work > > Andy > > > > ------------------------------------------------------------------------------ > The best possible search technologies are now affordable for all companies. > Download your FREE open source Enterprise Search Engine today! > Our experts will assist you in its installation for $59/mo, no commitment. > Test it for FREE on our Cloud platform anytime! > http://pubads.g.doubleclick.net/gampad/clk?id=145328191&iu=/4140/ostg.clktrk > > > > _______________________________________________ > Xcsoar-devel mailing list > xcsoar-de...@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/xcsoar-devel > ------------------------------------------------------------------------------ The best possible search technologies are now affordable for all companies. Download your FREE open source Enterprise Search Engine today! Our experts will assist you in its installation for $59/mo, no commitment. Test it for FREE on our Cloud platform anytime! http://pubads.g.doubleclick.net/gampad/clk?id=145328191&iu=/4140/ostg.clktrk _______________________________________________ Xcsoar-user mailing list Xcsoar-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcsoar-user
