gradient issues again after 31.31

I agree with different numbers and that is normal since no cycling comp is 100% accurate specially with elevation measurements but the difference should not be that big if your laps/rounds was tested on the same invornment.

It was not a coincidence because the difference is really big and was not measuring correctly in most areas after it updated the second time.

My garmin's gradient was working before(fairly accurate).

I don't agree with the idea that this is an unavoidable limitation of cycling computers.

A properly designed barometric altimeter is capable of extremely high vertical resolution—on the order of centimetres under stable conditions. The limitation is not the pressure sensor itself, but how the firmware filters, calibrates and interprets the raw pressure data while the bike is moving.

Gradient is calculated from changes in elevation over distance. If the barometric data is processed correctly, with appropriate filtering and compensation, the displayed gradient can be remarkably stable and accurate. This is why many cycling computers, including previous Garmin firmware versions, have produced reliable gradient readings for years.

What users are seeing now is not the normal level of variation expected from atmospheric pressure changes. It is inconsistent behaviour that appeared after recent firmware releases and is affecting multiple Edge models. That strongly suggests a change in the software processing pipeline rather than a fundamental limitation of barometric technology.

In other words, this is not a sensor problem—it is a data processing problem. The quality of the pressure sensor is only part of the equation; how the firmware handles its output is what ultimately determines the accuracy of the displayed gradient.

i know it very well, it is not working reliable in x40 devices since releasing, doesnt matter which fw is installed...

Maybe that's the reason why you said it never worked but in our experience, it was working fine before and we also know it very well too. I even compare it to other cyclo computer just to verify.

i had a few Garmins in the past 15years, and i ride 20k per year with 2-3 devices simoultanously, so i have a small knowledge what is reliable and accurate, and what is not. the x40 was never reliable, sometimes it shows near good value (within 1-2%), sometimes way off with more %, within same day on same loops or day by day or firmware by firmware. it is a known fact and the whole forum is full with it since 3 ys. grade% "worked" means in my term : always showing the accurate value like the altitude value, but when i climb on a 10% and it shows 4% then that is a bit far from "worked". and i had more x40 devices from early production and brand new, no difference, probably this new bosch bmp380 pressure sensor is worse than its predecessor in x30 family

I work in the aviation and free-flight industry, where we rely extensively on the Bosch BMP380 for flight instruments. From practical experience—not just the datasheet—I can say that this sensor is capable of producing extremely precise altitude data, with effective resolution on the order of 1 cm under stable conditions when properly implemented.

That is why I don't believe this is a hardware limitation of the BMP380. The sensor itself is more than capable of delivering the performance required for accurate gradient calculations.

The real challenge lies in how the firmware acquires, filters, compensates and processes the pressure data before it is used to calculate gradient. Two devices using exactly the same BMP380 can produce very different results depending on their algorithms.

So I would be very cautious about blaming the BMP380. It has been successfully used in demanding aviation applications where altitude precision is far more critical than displaying road gradient on a cycling computer. Everything points towards a software or signal-processing issue rather than a limitation of the sensor itself.

maybe "worse" because it is more sensitive than the 280 and much better code would needed than it was in the x30 family.

btw the altitude data is pinpoint accurate, just i dont know how much autocalibrations ares made by DEM file and GNSS position during the ride. but if you watch the raw barometer data (there is a hidden menu for it), it is very fluctuating even on the desk too.

EDGE 540 after last update doesn't calculte accumulative meter gain. A year or so ago after an update the % of cliembs were totally off... You guys keep on making GARMIN more expensive but less reliable!! FIX