cmps14 is a lottery ticket, one lucky, another not. There are jumps in one of them, I will take out another one and it works smoothly. On average, one in 4 works well. Perhaps cmps needs calibration in the workplace?
Similarly with bno080, these sensors are not all identical, there are some items that work badly and need to be replaced.
I have three Adafruit BNO085âs. Iâve not noticed untoward activity with any of them. The CMPS is susceptible to EMF activity which is a lottery in itself.
I keep looking at the Adafruits web site and they never seem to have any stock of the BNO08,
I got mine from Digi-Key.
Iâm not really up on the subject, but what about putting the cmps in a Faraday bag to shield it from EM interference? The little Iâve read about them says that a compass still works inside, but it should be less susceptible to noise from the environment.
You can look at this page, mine arrived in Spain in less than 10 days
Hello,
Simpliest solution: use a BNO085. Itâs run into Game Rotation Vector that use only gyroscope and accelerometer, no magnetometer. So not sensitive to EMI.
CMPS14 use a BNO085, but running into Rotation Vector mode, so with magnetometer, so sensitive to EMI.
Math
I guess you could.
But you are then using a more expensive BNO085 based board, locked in a less accurate mode, that is also more difficult to mount due to itâs EMF susceptibility.
Why anyone would bother baffles me.
It was just an idea that might help if you happen to be having CMPS problems and canât find a Bno085. They seem to be out of stock everywhere I look.
Ali express seems to have good stock. It just takes a long while for it to arrive.
Donât worry. I wasnât criticising you personally or aiming the comment at you. Just donât really understand the recommendation to use the CMPS14 over the BNO085 when it is likely to cause users more grief with installation and hard to pin down interference issues.
@Alan.Webb No worries, I didnât think you were. I understand that the Bno085 seems to be more stable, but I think lots of us already have a cmps so I just thought a Faraday container of some type might help with the jumps in heading people are experiencing. I donât even know if it would work for sure, I thought some of the science whizzes on here might know.
If it would work then it would seem like a pretty simple fix.
It will not help. I used 20 cmps 14 on various tractors, directly in the box and on the i2c extension cord in search of a good place, from this whole pool 5 or 6 pieces work very well without any problems with jumping and I have already replaced the rest with a bno 80.
why dont use 0x16 adress in place of 0x02 on cmps14 ? 0x02 IS magnÊtomèter
I donât appear to have any issues with my 3x CMPS14. Two of which are in the enclosure with the v2 board and cytron, down beside the seat, with the cab fans somewhere just under the seat. Might just be lucky or the steering could be even better than it already is.
For some reason my board was sending unchanging roll and IMU heading figures while combining the other day (probably human error when updating the AgOpen version). The effect was noticeable, but I was pleasantly surprised how well it could managing to steer the combine like that, provided it was reasonably online to start with.
However I can see the advantages of the BNO080/5. Is there anything significant difference between the 080 and 085 for use with AgOpen or any of the different boards they can come on, besides availability and price?
Using 1 cmps14 and 2 witmotion dont have any problèm with cmps14 in aluminium enclosure with esp32 in bluetooth mode. Witmotion on esp42 too ans non problèm. Just seem more rÊactive
None experienced so far, and the alixepress BNO 8? is inside an aluminum enclosure taped to the top of the opto isolators.
torriem
Just noticed that Adafruit has released a breakout board that contains the BNO085, which they say is virtually the same as the BNO080, but with some fixes for I2C. In any event I believe both the BNO080 and BNO085 have the ârobotic vacuum cleanerâ mode which lets you read the integrated roll and tilt values with no fuss and no complicated calculations. RVC mode looks very simple, and itâs supported by an Arduino library supplied by Adafruit. This looks pretty simple (way simpler than the BNO055). I see a few of you are using the BNO080. Are you using the RVC mode?
That is the board being used by many of us for probably 12 months now. The adafruit library will not fit on a nano so the pcb code uses the sparkfun library instead.
@whiterose In the thread for bno085 you refer to around december 2020 it ends up with the use of Game-rotation-vector. Which as far as I can see is the one the Ver 5 ino picks today.