oXs glider performance measurement

Development & General Chat for the superb openxvario project.

Moderator: rainer

Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

oXs glider performance measurement

Post by Carbo »

Here is a german thread about oXs for measuring glider performance.

With friendly help from Kevin Caldwell we translated the initial parts, to bring it back to the home of oXs.
Attachments
oXs gliderperformance.pdf
(991.82 KiB) Downloaded 509 times

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Indicated airspeed test, April1_16
Indicated airspeed test, April1_16
I have run some calibration tests on the oXs airspeed. I originally used a set-up with a water manometer on a 1:3 slope as Richard Johnson did when he calibrated full-scale airspeed instruments for his famous performance measurement tests. This applies a known pressure to the oXs sensor that should correspond to an airspeed that is fairly simple to calculate. This is basically how all airspeed instruments are calibrated (mercury manometer for higher speeds), since the dynamic pressure at a pitot tube for a given speed can be easily calculated. Most airspeed error in an airplane is due to static pressure errors, so those have to be eliminated for accurate calibration.

Initially I had some issue with the aispeed units. OpenTx pre 2.1 expects knots to be sent from the airspeed device, 2.1.1 through 2.16 expect oXs to send kph, 2.1.7 and 2.1.8 expects oXs to send knots again. You can have the display and log convert the units to whatever you want with your telemetry settings, but if oXs doesn't send the units that your version of OpenTx is expecting, your airspeed readings will be out, usually by a factor of 1.8. oXs can be set to send either kph or knots in the config file, as your version of OpenTx requires.

I soon found the water manometer method was not sufficiently accurate at low airspeeds because the pressures are very low. I tried adding a Magnehelic gauge I happen to have. It's full deflection is 0.5" of water, which was reasonably good for speed below 30 knots. Speeds between about 12 and 30 knots will be of most interest for performance measurement in RC gliders. This worked well enough to see the small difference caused by air density, even though I am testing at only 120' above sea level, and at the time only a few degrees away from the standard condition temperature of 15C. We also uncovered one small error in the oXs airspeed math that was quickly corrected by Michael.

oXs was initially set-up to measure only "true" airspeed. Using true airspeed, stall speed of an airplane will vary with altitude and temperature. All full-scale airspeed instruments measure "indicated" airspeed, where stall speed and best L/D speed will read the same on the instrument regardless of altitude or temperature. Reading indicated airspeed also makes performance measurement data reduction simpler. Michael (mstrens) kindly added the option for indicated airspeed to oXs. Thanks!

https://en.wikipedia.org/wiki/Airspeed

From discussions with carbo, we came up with a hopefully more accurate way to test the oXs airspeed. I happen to have a home-built CNC foam cutter that has a vertical positioner driven by stepper motors. I can vary the height of the oXs sensor with quite good accuracy, and therefore control the water head that the sensor is seeing down to tenths of a millimeter. Minimizing the air in the system to a small bubble at the sensor eliminated any air compression effects, and using a flexible pneumatic high pressure tube reduced any dimensional changes in the connection from the water column to the sensor.

The data I have taken with this system shows the oXs airspeed sensor is extremely accurate right down to 10 knots. The noise in the electronics does increase as the reading goes below 12 knots. I believe most of the error is still in my measurement system. There is an oscillation in the correction factor that looks a lot like positioning error from the lead screws on my CNC set-up.

I believe the error in the oXs airspeed is probably less than 0.1 knots from 12 knots up. This is extremely good! The best full-scale airspeed instruments have errors of +/- several knots that vary across the speed range, and several other RC airspeed sensor I have tested have errors of 6 to 13%. The oXs airspeed should be excellent for RC sailplane performance measurements.

Kevin
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

I've got my swapable pod with an oXs ready to go. The oXs has airspeed (calibrated per above), vario, altimeter and 5Hz GPS.

I laid up the saddle on my Xplorer 2 fuse, just in front of the wing. It is mainly Kevlar, and flexible enough to adapt to my e-Supra fuse as well. the connection will run under the back of the hatches to the FrSky Rx.

The whole thing weighs 39g, and shouldn't add much drag. It should be clear of the fuselage wake, and far enough ahead and above of the wing for pretty clean air.

I hope to get some late evening flights out over the ocean with both my e-Xplorer and the e-Supra. Probably mornings would be better, but! I'm hoping the air will be stable out over the 10C water of the Salish Sea (Straight of Georgia) between Vancouver Island and the mainland near where I live.

Kevin
Attachments
Static pressure from the pitot run to both the airspeed and altimeter/vario sensor.
Static pressure from the pitot run to both the airspeed and altimeter/vario sensor.
Pod front.JPG
Pod side2.JPG
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

kcaldwel wrote:I've got my swapable pod with an oXs ready to go....
Cool, i can only show my latest low-end solution. I reactivated my good old XL3200 to test mstrens proposition of an additional mix on the elevator with airspeed as source, adjustable targetspeed as offset and adjustable weight. It should help to stabilize the flightspeed in shorter time.

Bernd
Pitot.jpg
oXs.jpg
XL3200.jpg
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Nice! All that space in the fuselage...

It will be interesting to see how the Taranis works as a speed stabilizer. I can get my Eagle Tree Guardian stabilizer into my e-Supra I think, but it will be trickier on the e-Xplorer. There isn't much space, and the roll control is on SBus.

I particularly like the flying photo!

Kevin

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

I'm having trouble getting the TEST_1, TEST_2, TEST_3 values to be sent as telemetry values. I have the #define T1_SOURCE TEST_1 and #define T2_SOURCE TEST_2 lines uncommented, and #define ACCX_SOURCE TEST_3 set, but those are not showing up on my Taranis telemetry.

I set #define ACCY_SOURCE GLIDER_RATIO, and ACCY shows up when I discover new sensors. I must have something wrong in my config file, but I'm not sure what?

What are you using for values in the glide ratio calculation? Are these default values OK?

#define GLIDER_RATIO_CALCULATED_AFTER_X_SEC 1 // value must be higher or equal to 1, set line as comment if no calculation must be performed
#define SPEED_TOLERANCE 5 // in % of speed
#define VSPEED_MIN_TOLERANCE -200 // out of tolerance when Vspeed is lower than this value (cm/sec)
#define VSPEED_MAX_TOLERANCE -10 // out of tolerance when Vspeed is upper than this value (cm/sec)

I hope to get some data one day soon! I hope the feedback of airspeed to elevator works, because it will be very difficult to fit the autopilot to my Xplorer.

Kevin
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

kcaldwel wrote:I'm having trouble getting the TEST_1, TEST_2, TEST_3 values to be sent as telemetry values. I have the #define T1_SOURCE TEST_1 and #define T2_SOURCE TEST_2 lines uncommented, and #define ACCX_SOURCE TEST_3 set, but those are not showing up on my Taranis telemetry.
The TEST_1,_2_3 values have to be defined in the code somewhere, i think it is for the people who do the programming themselve. I do not know how to use them, unless mstrens tells me ;)
They are normally not necessary.
kcaldwel wrote:I set #define ACCY_SOURCE GLIDER_RATIO, and ACCY shows up when I discover new sensors. I must have something wrong in my config file, but I'm not sure what?
So far this is OK, you can use for example ACCX for transmitting the average sinkrate since T0 and ACCZ for the time since T0 (it is transmitted in 1/10 sec). T0 is the time, when the measurement starts. You can use a special funtion to announce the time since T0, so you know, when you have a valid measurement.
kcaldwel wrote:Are these default values OK?

#define GLIDER_RATIO_CALCULATED_AFTER_X_SEC 1 // value must be higher or equal to 1, set line as comment if no calculation must be performed
#define SPEED_TOLERANCE 5 // in % of speed
#define VSPEED_MIN_TOLERANCE -200 // out of tolerance when Vspeed is lower than this value (cm/sec)
#define VSPEED_MAX_TOLERANCE -10 // out of tolerance when Vspeed is upper than this value (cm/sec)
#define GLIDER_RATIO_CALCULATED_AFTER_X_SEC 1 // defines, how long the flight has to be within speed tolerance to start calculating the performance. You can try with 5 sec, but 10 sec are better in my pov. 1s is only Ok for the first test.

#define SPEED_TOLERANCE 5 // defines a speed sector, where the measurement is valid, if speed leaves the limits (up or down) the measurement is cancelled. Lower values give better accuracy, but it is harder to keep speed inside the limits. 5% is good, you can try 10% at the beginning.

#define VSPEED_MIN_TOLERANCE -200 //
#define VSPEED_MAX_TOLERANCE -10 //
are both additional limits for cancelling the measurement. If VSpeed is not between the limits, it is probably better, not to calculate any glider performance ;)

You will have to wait for calm conditions for your testflights. I have damaged my testglider. A mobile home stood on my landing field and the required modified landing approach ended in the wineyard. Two wings need to be fixed. But my depressive aftercrashphase is almost over ;)
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Thanks!

I couldn't see a definition of the SECONDS_SINCE_T0 and AVERAGE_VSPEED_SINCE_TO. I thought these were seconds since take-off and average vertical speed since take-off, which didn't sound that useful? From your description, I think these are the values I need to log. They seem to actually be the average vertical speed and the lapsed time in the test interval when the air speed and vertical speed were within the bounds, which is what I thought the TEST values were.

I'll give up on the TEST values, even though from the config_description they seem to be what I wanted:

"If oXs is within tolerance since more than a user defined enlapsed time (defined by GLIDER_RATIO_CALCULATED_AFTER_X_SEC), it performs the calculations
* Note: in this version of oXs, if you want to sent the calculated field you have to fill the data to transmit section using following code
* - TEST1 for enlapsed time (in 1/10 of sec)
* - TEST2 for averageVspeed (in cm/sec like Vspeed)
* - TEST3 for gliderRatio (in 1/10 of units)"

Thanks for the config file, I'll log what you are logging!

Sorry to hear about your glider. I hit my foot with my Xplorer wing attempting a stupid spot landing, and damaged both my foot and the wing.The foot has healed, the wing not so much. Not bad, just annoying.

Kevin

Edit: I see, SECONDS_SINCE_T0 is seconds since time zero, not take-off!
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

kcaldwel wrote:I'll give up on the TEST values, even though from the config_description they seem to be what I wanted:

"If oXs is within tolerance since more than a user defined enlapsed time (defined by GLIDER_RATIO_CALCULATED_AFTER_X_SEC), it performs the calculations
* Note: in this version of oXs, if you want to sent the calculated field you have to fill the data to transmit section using following code
* - TEST1 for enlapsed time (in 1/10 of sec)
* - TEST2 for averageVspeed (in cm/sec like Vspeed)
* - TEST3 for gliderRatio (in 1/10 of units)"
You are right Kevin, but this is only a relic of a early testversion, it should be changed in the decription. 9.1 in config_description is up to date, here you can see, in which telemetry fields the measurements can be sent.
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Section 9.1 in the files I downloaded yesterday also still lists the T1 and T2 telemetry fields, and seems to show that TEST_1, TEST_2 and TEST_3 are valid sources. I guess that no longer works though?

"#define T1_SOURCE GLIDER_RATIO // select between TEST_1, TEST_2, TEST_3, GLIDER_RATIO , SECONDS_SINCE_T0 ,AVERAGE_VSPEED_SINCE_TO , SENSITIVITY , PPM, VOLT_1, VOLT_2, VOLT_3, VOLT_4, VOLT_5, VOLT_6
* #define T2_SOURCE SENSITIVITY // select between TEST_1, TEST_2, TEST_3, GLIDER_RATIO , SECONDS_SINCE_T0 ,AVERAGE_VSPEED_SINCE_TO , SENSITIVITY , PPM, VOLT_1, VOLT_2, VOLT_3, VOLT_4, VOLT_5, VOLT_6 "

I'm still amazed when I get a config file to load and work, so I don't need too much more confusion! I've only killed one GPS forgetting to unplug it before trying to load a new config file...

Kevin
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

kcaldwel wrote:Section 9.1 in the files I downloaded yesterday also still lists the T1 and T2 telemetry fields, and seems to show that TEST_1, TEST_2 and TEST_3 are valid sources. I guess that no longer works though?
T1 and T2 are possible FrSky telemetry fields for transmiting data. You can use TEST_1, _2, _3 also, but we do not know, which data is in TEST_1, _2, _3. This is for developers, when they assign data to the TEST_1, _2, _3 oXs-variables in the oXs-code. In earlier versions our performance data was in TEST_1, _2, _3, but mstrens added "GLIDER_RATIO , SECONDS_SINCE_T0 ,AVERAGE_VSPEED_SINCE_TO , SENSITIVITY" as oXs-variables, to make things clearer. You can have infinite oXs-variables, but you have limited FrSky telemetry fields and in chapter 9.1 you define the assignment.
kcaldwel wrote: I've only killed one GPS forgetting to unplug it before trying to load a new config file...
Mstrens suspected so, but there was no volunteer to test it. This obscurity is diminished now ;). Because i knew, this will happen, i have always a spare GPS. Fortunally the N32 is one of the cheapest, but works perfect.

You will stumble over "SECONDS_SINCE_T0", oXs transmits it as integer, but OpenTX/Frsky expects 1/10 sec. This means, that 5 sec show up as 0.5.

Bernd
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

I managed to get a data file where everything worked, and no SD card error, with my old foam highly modified Calypso. It was a thermally day so the data is not of much use, but I did get some intervals that it calculated the glide ratio. I was using the Eagle Tree Guardian during the glides. The speed is oscillating. I had the sensitivity turned up all the way. I'm not sure if that is too much, or it can't damp out the oscillations. I need to test it in smoother air.

I had the oXs glide ratio interval set to 5 seconds, and the speed tolerance set to 5%. The L/D numbers seem odd (0.5 to 2.6), even for being in sink between thermals.

My Taranis B model still often, but not always, has "SD Card Error" come up when writing log files. I have tried multiple SD cards, but it persists. I'll have to try it with my Plus.

The GPS I thought I had killed by leaving it plugged in while loading a new config file actually does still work. I too have a spare GPS now, but it seems I didn't kill the first one after all. My house has a metal roof, and I'm surrounded by 40+m tall trees, so sometimes getting a a GPS fix takes longer than my patience allows.

Kevin
Attachments
Calypso-2016-06-07.csv.zip
(125.24 KiB) Downloaded 329 times
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Everything seems to work as intended. But you should verify glideratio-calculation once in the xls-file, to see, if it is working with speed in knots (i never tried, because i use km/h). It seems, you have to multiply your LD with 10? Michel is currently offline, flying his new plane. Changes in the code have to wait therefore.

The most important thing is flying in calm air, when you look at VSpeed, you can see the influence.
Calypso.png
In calm air you will see much better results.
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Stripping down an old telescopic antenna i made a new mount for the pitot-tube. The used (and minimum) outer diameter is 3.5mm to have sufficient room for the tubing. It could also be mounted in the canopy hatch.
Pitot1.jpg
Pitot2.jpg
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Looks nice!

I've been flying rather than testing, so I don't have any new data. I did check the csv file I posted above, and it appears the L/D calculation is out by a factor of 10.

When I calculate L/D using the instantaneous airspeed reported in knots converted to m/sec, and the average vertical speed, I usually get 10x the L/D reported. It is sometimes out in the decimal point (I.e. my calculated value of L/D = 14.8 may be shown as 1.46), but sometimes they are the same.

I guess you must be using an old version of OpenTx before 2.1.6 (I think), where they changed it back to expecting knots from the airspeed? If I set my oXs to kph, then my Taranis telemetry is wrong because oTx 2.1.8 that I am using is expecting knots from the airspeed sensor.

Kevin
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Here is a test in the early morning with a Stratos F3B veteran, some wind might have caused waves in the air, that might have caused waves in the log ;)
Messflug.png
Stratos_E-2016-07-31.zip
(118.33 KiB) Downloaded 226 times
Here are 3 exmples for the flightmodes thermal, cruise and speed, where i calculated the averaged parameters with the log data:
HC_2016_07_31_4.png
32:20 to 35:30, 35 km/h, average sink 37,2 cm/s, glideratio 26
HC_2016_07_31_5.png
25:50 to 26:50, 44 km/h, average sink 49,3 cm/s, glideratio 25
HC_2016_07_31_6.png
20:30 to 22:00, 53 km/h, average sink 98,4 cm/s, glideratio 15

It shows, that the oXs performance arameters are not averaged enough, they are too sensitive for phugoide, local and pilots influences. Further calculations inside oXs seem complicated, maybe we should let oXs supply the raw data only. To have averaged values i will try a LUA-script, triggered by the pilot, when constant flight is achieved, to announce the averaged parmeters until the pilot stops measurement again. Same procedure as calculating with the log, but having the result in flight already. BTW LUA-volunteers are very welcome to help ;)

Kevin, you are right, OpenTX expects knots in 2.1.8 and does the conversion itself.

Bernd
mstrens
Posts: 1435
Joined: Fri Dec 27, 2013 7:49 pm
Country: -

Re: oXs glider performance measurement

Post by mstrens »

Should I delete the code about glider ratio ( including secFromT0 and averageVspeedSinceT0 ) from openXsensor.
If it is not used, best is to keep the code as small as possible.

Michel
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Michel,

as long as there is no real need to delete it, i propose to keep the option. The parameters are good for a quick overview, but as you see in my examples, the values have too much noise. Also only FrSky-users have the chance of using LUA at the moment.

The main advantage of deleting the option would be a more cleaned up user-interface and less code to maintain, what do you think?

Bernd
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Bernd,

Is that flight data using manual speed control, or do you have the feedback loop of speed to elevator working? That is pretty smooth flying either way!

I've got to get out with my Xplorer and get some data. I was off to a competition for the first time in 44 years, now I'm tired!

Kevin
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Kevin,

i changed my mind about forcing the plane in a constant flight. I will try to change the measurement in a way, to be able to measure the plane in its normal behaviour. If a plane wants to fly phugoids, OK, then i measure it flying them. So the results will be more real world and not artificial improved.

Maybe mstrens is so kind to change it in oXs, then it would work with all systems and the data can be logged easily or i have to use a LUA script to do the calculations. It requires the pilot to signal with a switch, when the measurement has to run. In oXs PPM could be used for triggering (e.g. -100). What do you think, Kevin?

If there is no other Kevin Caldwell flying competitions, it seems you made second place? You left not much room for improvement, but 44 years of practicing have to pay out ;)

Congratulations!

Bernd
kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel »

Your speed control on that data is very good! I think taking the average over some long periods of that data should give a pretty representative number for L/D and sink rate in your different flight modes. Yes, I think "as it really flies" numbers including the phugoids may be of more use. It would be at least theoretically interesting to know what the damped, maximum performance numbers are though, as a reference to the calculated performance numbers.

I'm not sure if it is easier to do long period calculations from the data afterwards, or on the oXs? Afterwards you can decide when the average speed is constant, that might be hard to see when flying.

I'm impressed you are getting such nice data. I'm far from sure I can keep an average speed that constant.

I would like to get at least a piece of a polar curve over a range of speeds in each camber setting, to see what the best L/D speed in each camber setting is, what the high speed L/D is, and what the minimum sink speed and camber setting is. Much testing to do, and none accomplished so far.

I don't know anything about LUA scripts at the moment, but perhaps this might be of some use as a starting point to get some telemetry feedback on the L/D?:

http://www.rcgroups.com/forums/showthre ... +ratio+lua

He is calculating L/D over the ground using the GPS speed, but perhaps it isn't too hard to change that to using the airspeed?

Yes, I finished second, but I'm on the downhill slide! I won the 1972 FAI Sailplane at the Canadian Nats (only 7 pilots I think) at 17 with a glider I designed myself, my only other RC sailplane competition. I'm not a big fan of the US ALES format, with 200m cut-offs. It is really just a "precision landing" contest, where a few seconds and a meter here or there decides the whole thing. In 1972, 10 minutes was very challenging with a balsa 2m glider, cautious 6V winch launches to keep the wings on, and not much of a clue about air reading. Flying 4m composite gliders from 200m doesn't have much to do with soaring skills today.

Have you tried the triangle flying LUA for the Taranis? I've made up my own competition: 150m cut-off height (maybe F5J launch rules?), you can't enter the course until after motor cut-off, then 30 minutes to fly as many laps of a 2km perimeter triangle as you can. That is a lot more fun!

http://www.rcgroups.com/forums/showpost ... tcount=279

Kevin
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

kcaldwel wrote:I'm impressed you are getting such nice data. I'm far from sure I can keep an average speed that constant.
i did a lot of measurement flights meanwhile, and it seems, that calming down the plan is somehow hardcoded in my brain. Yesterday i made a sundown measurement flight, when a family came watching. They wanted to know, what i am doing there. When you look at the log, you can see exactly, where the ommunication started. I was surprised, how much concentration you need to fly calm, i did not realize it until then. BTW there is still good and bad air, even at sundown.
Test_abends.png
An intended mesurement after breakfast this morning ended with some thermalling. Flipping the flightmodeswitch and start circling seems also hardcoded in my brain. The best time for measurement is really unhappily before and at sunrise.
900Test.png
kcaldwel wrote:Have you tried the triangle flying LUA for the Taranis? I've made up my own competition: 150m cut-off height (maybe F5J launch rules?), you can't enter the course until after motor cut-off, then 30 minutes to fly as many laps of a 2km perimeter triangle as you can. That is a lot more fun!
Looks challenging, i will try this. I presume my first try will result in 0.4 2km perimeter triangles ;)
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Here are some examples, where i compared the oXs results for average sink and glideratio with the averaged results from log. The boxes mark the time intervall for the calculation with the log data.
Auswertung1.png
Auswertung2.png
Auswertung3.png
Auswertung4.png
Auswertung5.png
The last one is perfect, #1 and #4 are OK, #2 and #3 are too low. I have at the moment no explanation for the deviation in 2 and 3. On the other side a glideratio of 28-29 seems unlikely, but i can not see, what went wrong.
Stratos_E-2016-08-07_1.zip
(113.81 KiB) Downloaded 211 times
User avatar
jhsa
Posts: 19480
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: oXs glider performance measurement

Post by jhsa »

Hi guys, sorry if this is a little of topic, but where can I get the info about building the oXs vario with Accelerometer compensation?

Thanks

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW
User avatar
kalle123
Posts: 905
Joined: Sat Mar 29, 2014 10:59 am
Country: -
Location: Moenchengladbach

Re: oXs glider performance measurement

Post by kalle123 »

Hi João.

The documentation is one problem, but remember, mstrens is alone ....

I would advice to download oXs v7.0 from here https://github.com/openXsensor/openXsensor

Just push the green "clone or download" button.

Then open oXs_config_description.h. That is the documentation. Looks strange at first, but you get to like it ;)

Look for chapter 4.1 and for chapter 12

There you'll find the information, how to connect hardware.

br KH
User avatar
jhsa
Posts: 19480
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: oXs glider performance measurement

Post by jhsa »

Thanks a lot for the info.. Is this working well now? Did it really reduce the response time from the vario?
Only using the pressure sensor, the delay is about a second.. Enough to miss the thermic :) At least for a beginner like me :D

Thanks

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW
User avatar
kalle123
Posts: 905
Joined: Sat Mar 29, 2014 10:59 am
Country: -
Location: Moenchengladbach

Re: oXs glider performance measurement

Post by kalle123 »

João, I am also not a specialist for thermic like carbo here.

But it could also be your plane .... Too fast and/or too heavy.

I am flying a Chinook II, and that one is really taking thermic very well. Slow and light. ;)

I am not using the IMU 6050.

KH
User avatar
jhsa
Posts: 19480
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: oXs glider performance measurement

Post by jhsa »

No, I have a good thermic glider now. It's a Climaxx Elektro ( 4 servo wing) from Höllein. It flies like a dream, that's why I would like to build the vario with accelerometer.. :)

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW
Carbo
Posts: 467
Joined: Fri Aug 02, 2013 6:55 pm
Country: Germany
Location: Freinsheim RP

Re: oXs glider performance measurement

Post by Carbo »

Here you can see a flight with IMU accelerated vario from yesterday, look at the speed! Completely accidentally you can see the topspeed in the data rows below.
30min.png


OK - enough revenge for the mod´s off topic: Seriously you can expect 0,4 s faster response from the IMU-vario. You can configure oXs in a way, that you can switch with a PPM connection between IMU- and standard-vario. Then use the one, that gives you a better feeling for the air. Maybe IMU for light planes, because you will detect smaller thermals, that you can not use with a bigger glider. Price is nearly the same, so test it, and you will hear the difference, both varios work perfect in my pov.

Bernd
User avatar
jhsa
Posts: 19480
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: oXs glider performance measurement

Post by jhsa »

Thanks Bernd, will give it a shot and report in the right thread wherever that is ;) :)

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW

Post Reply

Return to “OpenXVario - an open source vario supported by the open source firmwares!!”