Hello everyone, as the title suggests, I am trying to replicate an alt-zimuth system (an inclinometer and a compass) that I would like to apply to my telescope.
I did several searches and I fell back on the Adafruit BNO055 sensor, I purchased it and now I'm testing it but I'm encountering several problems regarding small and slow movements because they are not detected.
I'm rewriting the code to verify that I did everything correctly.
The result I would like to obtain is the display of the compass degrees and inclination on 2 4x7-segment displays (TM1637).
The reasoning I follow is that 0 degrees north and 0 degrees inclination are calculated at the beginning with an offset (an average of 100 readings).
after that calculation I calculate the values: that offset is subtracted from the value read by the sensor, 5 readings are taken and the average is value that I see on the displays.
In your opinion, is this reasoning correct or not? Are there sensors that allow you to detect even those small movements that I am unable to detect or more sensitive?
my tests consist of comparing the value read by the sensor and the compass application of the mobile phone
I had also encountered a problem with the Arduino (nano) because it turned off when powered by the power bank so I inserted some resistors on some pins set to high to increase absorption and not make it turn off.
Thanks for your opinion and sorry if the category isn't correct
Er, umm, I think the term is Goto
Did you follow the tutorial provided by Adafruit Arduino Code | Adafruit BNO055 Absolute Orientation Sensor | Adafruit Learning System
Does it show small movements?
The Goto is something that moves automatically the telescpe, a push-to it's manually and I "only" want to see compass and inclination degree on 2 display, i don't need motors or encoder to move telescope
I followed it and try to modify to see how accurate it can be but it can't catch very slow and continuos movement (for example I moved very slow and with compass of the iPhone i saw 10 degree and the BNO055 didn't moved at all, but only with that slow) and sometimes the measure isn't right at all
I was thinking to try the BNO085 because it's more sensible and precise, but I think for my project just a gyroscope sensor will be ok but I'm not sure and i read that IMU's can perform calculation in hardware and I think are more accurate... Am I right?
Are you talking about the Adafruit example code, or your code?
There are youtube videos showing the raw output when the device is stationary and the numbers are constantly changing due to noise.
Are you using a low pass filter to remove the noise, or a high pass filter to remove drift?
Yes, about the adafruit examples, I modified them by changing the delay so i can see lots of measure and if I move very very slow it didn't change.
I also tried on a new code made by me where I read 10 measure and perform an average and it didn't change.
I think noise of the sensor is removed by the mod Operation_Mode_Ndof, but I'd like to have a precision like an encoder, but i don't want to modify my telescope by adding screws or changing the rockerbox so I was thinking to add this sensor on the tube with a nylon strip and a case 3d-printed and place the arduino and display on the base without screws.
is it possible or i'wasting time to find an impossible solution?
Look at the VECTOR_ACCELEROMETER values.
If they do not change then something is wrong.
If I move very slow they don’t change, do you think it’s because it’s not too sensitive like the bno085? Because I was looking and I read that the bno055 works with samples rate at 100Hz and bno085 at 1Khz, and it have more accurate sensor fusion algorithms.
But for my application do you think I need to improve something with the accelerometer? Because I only read gyroscope and magnetometer to calculate compass and inclination
Aha, I defer to your expertise!
Or what about encoders? (for the compass)
I mean, if I put an encoder on the edge of the base, with a plastic wheel that rotate when i rotate the scope, in this way 1 turn of the telescope it means more turns of the encoder so the precision will encrease.
I don't know if you know how the rockerbox of dobson are made, but there is an recess where I can build a small base in wood so i don't have to screws anything on the scope and it will always be original...
Is that a bad idea instead of any sensors?
In this way the only problem will be the inclination, and or with a sensor or with another encoder i have to build something...
Or maybe an AS5600 with a magnet on the scrues in the middle (for azimut)
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.