Does gyroscope requires calibration? (MPU 6050)

[ckkkkk]

Re: How to convert accelerometer data into G with the ADXL345 digital accelerometer


Maybe this is what i've been looking for but whether is it -(output (x) / 4) or is it just (output (x) / 4) without the negative sign. Also why it needs to be divided by 4?

 

[ckkkkk]

Re: How to convert accelerometer data into G with the ADXL345 digital accelerometer

Well maybe try imagine that u strap the accelerometer on your wrist, the acceleration data should always be in positive if you did not move your wrist upside down. With that, how can i perform all those calculations mentioned above?

 

[KlausST]

Re: How to convert accelerometer data into G with the ADXL345 digital accelerometer

Hi,

I´m confused.

Initially you asked how to get the G value.

Then I´ve told you how to calibrate offset and gain.
After this you get exact values for G in all three axis.
(Now you show an AN how to calibrate offset (and gain?) with a similar method but with the use of internal registers.
My method is a common method applyable for all acceleration sensors with or without internal correction features)

Since we live on the earth (and not on ISS) there is 1G of acceleration when we stand still. Not zero.

the acceleration data should always be in positive if you did not move your wrist upside down.

Please explain: why do you expect it to be always positive?

Klaus

 

[ckkkkk]

Re: How to convert accelerometer data into G with the ADXL345 digital accelerometer

Right by using the method you provided in order to get G values, first needed to get the offset value which is to use (z_before - z_now)/2.



Imagine strapping this on your wrist with the accelerometer on top, and if you do not move your hand then the accelerometer will always output positive data? As the accelerometer is not upside down therefore no negative data produced. With that, i can only acquire the "z_before" i will not be able to get the "z_now" then neither could i find the offset for this nor the final G value.

 

[KlausST]

Re: How to convert MPU6050 accelerometer's measurement range from +-16g into 2g?

Hi,

Imagine strapping this on your wrist with the accelerometer on top, and if you do not move your hand then the accelerometer will always output positive data? As the accelerometer is not upside down therefore no negative data produced. With that, i can only acquire the "z_before" i will not be able to get the "z_now" then neither could i find the offset for this nor the final G value.

I can only repeat myself:
Z_now, z_before is just to get calibration values.
It makes no sense to do the calibration run when the sensor is on the wrist.

Klaus

 

[ckkkkk]

Re: How to convert MPU6050 accelerometer's measurement range from +-16g into 2g?

Hi KlausST,

I found out the offset and sensitivity using the method you mentioned. But should I follow the sensitivity from the datasheet or to replace it with the one I found using the method you provided?

 

[KlausST]

Re: How to convert accelerometer data into G with the ADXL345 digital accelerometer

Hi,

in doubt I recommend to use the information from the manufacturer, they should know their devices and methods best.

Klaus

 

[ckkkkk]

If it does, what is the method that i need to follow?

 

[KlausST]

Hi,

you want me to read the datasheet for you and write the contents here in the forum?

Klaus

 

[ckkkkk]

Hi,

you want me to read the datasheet for you and write the contents here in the forum?

Klaus

Hi KlausST,
According to the datasheet, it says that they have reduced the need for user to do the calibration but i do not get 0 when the gyroscope is stationary. So, is there any manual way to do the calibration just like the accelerometer?

 

[FvM]

Datasheet specifies +/- 20°/s zero tolerance. Do you see larger gyro offsets?

Unless you do the initial zero adjustment in a roller coaster, there should be no actual problem, isn't it?

 

[ckkkkk]

Datasheet specifies +/- 20°/s zero tolerance. Do you see larger gyro offsets?

Unless you do the initial zero adjustment in a roller coaster, there should be no actual problem, isn't it?

No, i do not see any larger gyro offset but shouldn't all the axis get close to 0 while being stationary?

 

[FvM]

Close to zero, but expecting deviation up to the specified zero tolerances plus drift. The specification seems pretty clear to me.

 

[ckkkkk]

Close to zero, but expecting deviation up to the specified zero tolerances plus drift. The specification seems pretty clear to me.

These data are acquired when the gyro is on flat surface also are these offsets correct?

 

[FvM]

The range is?

 

[ckkkkk]

The range is?

Hi FvM,
Its on 250 dps.

 

[wwfeldman]

think of it as if you were calibrating a meter.
you connect it to the calibration machine, make the adjustments and use the meter for (some time period usually specified by contract or by local practice)

after the defined time period, bring the meter back to calibration machine and test again.
if its still calibrated, great
if not, calibrate it.

(some time period) is usually every 3 months, or every 6 months, or every year, depending.

 

[ckkkkk]

think of it as if you were calibrating a meter.
you connect it to the calibration machine, make the adjustments and use the meter for (some time period usually specified by contract or by local practice)

after the defined time period, bring the meter back to calibration machine and test again.
if its still calibrated, great
if not, calibrate it.

(some time period) is usually every 3 months, or every 6 months, or every year, depending.

Hi wwfeldman,
Alright, so i have to re-calibrate it after sometimes. But, what i am wondering now is whether i did the calibration correctly or not. I taped the gyroscope on a wooden block and i placed the block on a table then i started taking the measurements and on post #34 those are the data i get from the 3-axis.