search

mstrens / oXs_on_RP2040

version of openXsensor to be used on raspberry pi pico RP2040 (more protocols, more functionalities)
93 stars 23 forks source link

Adoption of XGZP6897D Differential Pressure Sensor #128

Open jasc76 opened 9 months ago

jasc76 commented 9 months ago

Datasheet

XGZP6897D-Pressure-Sensor-V2.7.pdf

In Test is the 3.3V 500Pa Variant. Currently the sensor is recognized, but no measurements are displayed when using FV. In parallel a GY-63 is connected so I²c should be terminated without using pullups

Baro sensor is detected using MS5611 Sensitivity min = 100 (at 100) , max = 300 (at 1000) Hysteresis = 5 Airspeed sensor is detected using XGZP....

jasc76 commented 9 months ago

Just downloaded the 2.12.4 Version and will try to get some debug data

If you like I can send you one of these sensors (if you are in the EU... US takes ages) . I got 2 :-)

mstrens commented 9 months ago

Perhaps we can find the issue with the debug messages. In a second stage, if I do not find the issue, then getting a sensor could help. I am in EU (Belgium).

jasc76 commented 9 months ago

looks good, values are definitely changing when blowing over it

rawPres=66695 rawPres=172799 rawPres=33783 rawPres=3823 rawPres=16777142 rawPres=115830 rawPres=205088 rawPres=320911 rawPres=486299 rawPres=16737849 rawPres=16747550 rawPres=16740163 rawPres=16736876 rawPres=783477 rawPres=16728544 rawPres=531232 rawPres=1093841 rawPres=1124545 rawPres=1056067 rawPres=649749 rawPres=16735820 rawPres=16743526 rawPres=16740288

mstrens commented 8 months ago

In the version you tested I made 2 changes:

mstrens commented 8 months ago

I made a new version 2.12.5 in order to manage negative pressure (the sensor is differential and so some values can be negative. Without this correction, negative values were processed as very big values and could generate wrong results. I hope it help to get more realistic airspeed.

jasc76 commented 8 months ago

you're a wizard :-) depending, which port gets the air pressure the value is pos or neg now. I guess ignoring pos/neg makes the most sense since its irrelevant for the application

rawPres=-30229 rawPres=-28695 rawPres=-1945446 rawPres=-40881 rawPres=-51391 rawPres=-27898 rawPres=-88071 rawPres=-73877 rawPres=-324429 rawPres=-889948 rawPres=898187 rawPres=773595 rawPres=-70024 rawPres=79785 rawPres=1257551 rawPres=700077 rawPres=-64377 rawPres=-61067 rawPres=1076767 rawPres=-61751 rawPres=-35367 rawPres=-30116 rawPres=-29381

jasc76 commented 8 months ago

hmm. at the same time there seems to be a lot of noise

rawPres=1733 rawPres=3086 rawPres=-218 rawPres=1984 rawPres=4579 rawPres=1337 rawPres=175 rawPres=-1214 rawPres=-2869 rawPres=-1598 rawPres=-1104 rawPres=1779 rawPres=1744 rawPres=862 rawPres=335 rawPres=-1252 rawPres=-547 rawPres=103 rawPres=246 rawPres=1398 rawPres=1641 rawPres=-8 rawPres=-348

mstrens commented 8 months ago

Yes but this is perhaps not so abnormal. A raw value like 2000 means in fact a pressure of 0.25 Pa which is equivalent to a speed of 2km/h

Best is to check the airspeed calculated by oXs and see if it is realistic.

Do not expect an realistic value when speed is less than about 15 km/h because this result of very low difference of pressure. When speed get higher, then accuracy/noise of airspeed becomes better.

mstrens commented 8 months ago

FYI tTaking into account the negative values is not bad:

jasc76 commented 8 months ago

ok cool

Currently this is all a crude test setup botched together with some wires. I now ordered stuff to actually mount it to my glider. I would assume I put it in the plane, hook up SBUS and tweak the compensation adjustment channel until I get a consistent result. Is this correct !?

Which oXs version should I use now? The master or still test?

mstrens commented 8 months ago

Use always test branch. I put all new functions there. Still take care that there can be some not yet known bugs.

Some users tested the airspeed while running with a car. One man drives the car, another hold the pitot tube outside. The value is read on the handset.

jasc76 commented 8 months ago

Do I have to change the k value in the code or is the calibration done solely done with the compensation channel ?

mstrens commented 8 months ago

You have to change manually in config.h the value for XGZP_K_FACTOR It depends on the max pressure of the model you selected. Here the text in config.h

define XGZP_K_FACTOR 8192.0 ; // See datasheet : K depends on the sensor model (related to the max pos pressure being measured)

                          // 8192.0 is the value when max sensor pressure is 1000Pa (1kPa) like XGZP6897D001KPDPN
mstrens commented 8 months ago

The compensation is a parameter used only for the compensated aispeed. it does not matter to calculate the airspeed.

jasc76 commented 8 months ago

I dont really care for airspeed, I'm just interested in the vario compensation. So my thought was not to bother with the k value. Also in the datasheet, all models below 1kPa have the same k value.

jasc76 commented 8 months ago

hmm I wanted to do the calibration, I used the current test branch 2.13.1 and I get changing raw values. However with FV airspeed is always 0. I did not hook up a receiver and just used usb

rawPres=-683676 rawPres=-734410 rawPres=-800049 rawPres=-25184 rawPres=-217883 rawPres=-1596020 rawPres=-1711642 rawPres=-1651593 rawPres=-28809 rawPres=-33805 rawPres=-1309954 rawPres=-635123 rawPres=-1202301 rawPres=-690018 rawPres=-769427 rawPres=-567340 rawPres=-958961 rawPres=-1552390 rawPres=-1172935 rawPres=-1505920 rawPres=-1097496 rawPres=-78826 rawPres=-33733 rawPres=-2576 rawPres=-6549 rawPres=-4949 rawPres=-8274 rawPres=-7904 rawPres=-4487 rawPres=-22924 rawPres=-32619 rawPres=-3220531 rawPres=-2690748 rawPres=-3060192 rawPres=-3926 rawPres=-4940 rawPres=-1513 rawPres=-2688 rawPres=-6399 rawPres=-7482 rawPres=-5073 rawPres=-6046 rawPres=-8062 rawPres=-5920

`Version = 2.13.1 Function GPIO Change entering XXX=yyy (yyy=255 to disable) Primary channels input = 5 (PRI = 5, 9, 21, 25) Secondary channels input = 255 (SEC = 1, 13, 17, 29) Telemetry . . . . . . . . = 255 (TLM = 0, 1, 2, ..., 29) GPS Rx . . . . . . . . . = 255 (GPS_RX = 0, 1, 2, ..., 29) GPS Tx . . . . . . . . . = 255 (GPS_TX = 0, 1, 2, ..., 29) Sbus OUT . . . . . . . . = 255 (SBUS_OUT= 0, 1, 2, ..., 29) RPM . . . . . . . . . . = 255 (RPM = 0, 1, 2, ..., 29) SDA (I2C sensors) . . . . = 10 (SDA = 2, 6, 10, 14, 18, 22, 26) SCL (I2C sensors) . . . . = 11 (SCL = 3, 7, 11, 15, 19, 23, 27) PWM Channels 1, 2, 3 ,4 = 255 255 255 255 (C1 / C16= 0, 1, 2, ..., 15) PWM Channels 5, 6, 7 ,8 = 255 255 255 255 PWM Channels 9,10,11,12 = 255 255 255 255 PWM Channels 13,14,15,16 = 255 255 255 255 Voltage 1, 2, 3, 4 = 255 255 255 255 (V1 / V4 = 26, 27, 28, 29) RGB led . . . . . . . . . = 16 (RGB = 0, 1, 2, ..., 29) Logger . . . . . . . . . = 255 (LOG = 0, 1, 2, ..., 29) ESC . . . . . . . . . . . = 255 (ESC_PIN= 0, 1, 2, ..., 29) Locator CS . . . . . . . = 255 (SPI_CS = 0, 1, 2, ..., 29) SCK . . . . . . . = 255 (SPI_SCK= 10, 14, 26) MOSI . . . . . . = 255 (SPI_MOSI=11, 15, 27) MISO . . . . . . = 255 (SPI_MISO=8, 12, 24, 28) Output level High . . . . = 255 (HIGH = 0, 1, 2, ..., 29) Output level Low . . . . = 255 (LOW = 0, 1, 2, ..., 29)

Esc type is not defined

Protocol is Sport (Frsky) CRSF baudrate = 420000 Logger baudrate = 115200 PWM is generated at = 50 Hz Voltage parameters: Scales : 1.000000 , 1.000000 , 1.000000 , 1.000000 Offsets: 0.000000 , 0.000000 , 0.000000 , 0.000000 No temperature sensors are connected on V3 and V4 RPM multiplier = 1.000000 Baro sensor is detected using MS5611 at I2c adress 77 Sensitivity min = 100 (at 100) , max = 300 (at 1000) Hysteresis = 5 Airspeed sensor is detected using XGZP.... No Vspeed compensation channel defined; oXs uses default settings First analog to digital sensor is not detected Second analog to digital sensor is not detected Foreseen GPS type is Ublox (configured by oXs) :GPS is not (yet) detected Led color is normal (not inverted)`

mstrens commented 7 months ago

What do you mean with "calibration"?

At power on oXs makes several (32) measurements and uses the average as offset. So this is an automatic calibration.

jasc76 commented 7 months ago

ok I get now also Airspeed. But I guess you need to show the absolute value, because I get neg values :-)

processing cmd

Cmd to execute: FV

Vspeed = 3 cm/s Baro Rel altitude = 24 cm Airspeed = -1235 cm/s Compensated Vspeed = -1529 cm/s pwmTop= 20000 rawPres=-733749 sumPa=-86.113831 count=1 rawPres=-1275134 sumPa=-152.200867 count=1 frawPres=-22982 sumPa=0.649719 count=1 v processing cmd

Cmd to execute: FV

Vspeed = 1 cm/s Baro Rel altitude = 30 cm Airspeed = 28 cm/s Compensated Vspeed = 187 cm/s pwmTop= 20000 rawPres=-34739 sumPa=-0.785461 count=1 rawPres=-304483 sumPa=-33.713196 count=1 rawPres=-2581819 sumPa=-311.708313 count=1 fvrawPres=-3587696 sumPa=-434.496033 count=1

processing cmd

Cmd to execute: FV

Vspeed = 1 cm/s Baro Rel altitude = 32 cm Airspeed = -2541 cm/s Compensated Vspeed = -1600 cm/s pwmTop= 20000 rawPres=-3573326 sumPa=-432.741882 count=1 rawPres=-27387 sumPa=0.112000 count=1 rawPres=-27745 sumPa=0.068298 count=1 rawPres=-18979 sumPa=1.138367 count=1 rawPres=-11486 sumPa=2.053040 count=1 rawPres=-12270 sumPa=1.957336 count=1 rawPres=-11181 sumPa=2.090271 count=1

mstrens commented 7 months ago

I use negative values because they can help during the calibration process (to calculate offset) Do you test with a pitot tube or just applying a pressure on one of the entry of the sensor? With a pitot tube, you should try to invert the 2 tubes (this should invert the sign.