Fake BME680 (Bosh temp/VOC sensor) ? [SOLVED]

[Gabri74]

Hi, I was just wondering if anyone has ever been supplied with fake BME680 sensors.
I've just bought two new modules to integrate into an existing project, but the readings are all wrong, to the point they seems fabricated.
(link to the module https://it.aliexpress.com/item/1005007312026696.html )
(I know that the price was a bit to low for this kind of sensor, but being quite old and superseded by the 688 I risked it )

I.e. on the other genuine sensor, those are the readings I get:
IAQ: 83
static IAQ: 52
CO2: 620 (ppm)
VOC: 0.8

On the fake(?) sensor I get completely different readings:
IAQ: 243
static IAQ: 814
CO2: 8140 (ppm)
VOC: 1000

Note that on the fake sensor:
- IAQ and static IAQ are too far apart
- VOC is constant at 1000 (!)
- static IAQ is always equal to CO2 / 10 --> this in particular make me think the values are fabricated

I'm using the official Bosch library.

Those sensors have a settle/burn in period and admittedly I've just stared using them for a few hours... but the values are two wrong to be a genuine part...

Does anyone know if it is possible to detect using some I2C commands if the chip is fake or genuine ?

[wraper]

I would not be surprised if sensor is genuine but screwed up during PCB production.

[Gabri74]

Thanks for the reply.

I turned out it was my fault all along... I forgot that I did implement calibration saving when full accuracy was reached.
When I swapped the old sensor with the new on my test unit, I forgot to clear the calibration.

It appears that calibration parameters for those sensor could be quite different from one another, and the old
calibration loaded at power up was screwing up the algorithm and giving me saturated values for some parameters.