Faulty bme680 breakout?

Support
#1

Hi all,

I’ve recently purchased the Pimoroni bme680 breakout, and it’s behaving unexpectedly.

Running the “read-all.py” script stops returning correct values after an indeterminate period of time (between 10-20 minutes). The gas readout stops working, and the other measurements freeze at odd values:

18.36 C,993.54 hPa,52.35 %RH,70984 Ohms
18.38 C,993.54 hPa,52.27 %RH,70924 Ohms
18.39 C,993.55 hPa,52.20 %RH,71104 Ohms
18.40 C,993.54 hPa,52.16 %RH,71406 Ohms
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH

Is this an issue with the breakout, or user error?

Thanks,
Richard

#2

I should say that CTRL-C’ing the script and restarting fixes the issue, but only temporarily. The issue returns.

#3

How is it wired up? Breakout garden, header and jumper wires, soldered connection etc.

#4

Hi, thanks for the reply.

2x I2C 1x SPI breakout garden on a RPi Zero W. Raspberry Pi OS, but not the very latest.

After playing with it for a while, it seems to have settled down. It’s been running for about 2 hours without issue, which is by far the longest it has been going. I’ll report back later with an update.

#5

Ok, I have an i2c SPI Mini breakout garden and the i2c mini Breakout Garden plugged into the Mini Hat Hack3r, that connect to a Zero W. No issues I am aware of.
BME280, RV3028, LTR-559 in the i2c mini and a 0.96 LCD display in the i2c SPI mini.

#6

Happened again. :(

17.92 C,991.67 hPa,53.30 %RH,97295 Ohms
17.92 C,991.68 hPa,53.32 %RH,97126 Ohms
17.92 C,991.69 hPa,53.35 %RH,97464 Ohms
17.91 C,991.70 hPa,53.37 %RH,97014 Ohms
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH
33.38 C,659.23 hPa,100.00 %RH

#7

I was running my Pi off of a laptop USB port, which I know isn’t ideal. I’ve moved it to a good USB power supply now to see how it runs. I suspect these kinds of precision components are sensitive to power quality.

#8

OK, happened again. I’m going to assume I’ve got a faulty sensor.

#9

I’m thinking its a faulty sensor. You could try the temperature-pressure-humididty.py example. Just to see if its only the air sensor part that’s glitching.

#10

Yeah, disabling the gas sensor in the read-all.py script, it’s rock solid with just temp/press/humid. Looks like it’s the gas sensor that spoils things. Anyway, I’ve emailed Pimoroni so I’ll take it from there.

#11

Did you get any resolution to this? I get the same with two Pimoroni BME680 boards.
Using temp-press-hum.py but with a 1s delay between readings, after about 15 mins:

Timestamp            Secs       Cnt  Temp    Presr  Hum
18:20:30.007849      956.2      919  20.22   993.06   52.39
18:20:31.049289      957.3      920  20.22   993.06   52.38
18:20:32.090693      958.3      921  20.22   993.06   52.38
18:20:33.131989      959.3      922  20.22   993.06   52.38
18:20:34.140396      960.4      923  32.93   658.71  100.00
18:20:35.148797      961.4      924  32.93   658.71  100.00
18:20:36.157225      962.4      925  32.93   658.71  100.00
18:20:37.165592      963.4      926  32.93   658.71  100.00
18:20:38.173951      964.4      927  32.93   658.71  100.00

The output is slightly different but measurement code is the same and I get the issue with the vanilla temp-press-hum.py too.

If I leave it running for a while, I get an exception:

Traceback (most recent call last):
  File "temp-press-hum.py", line 36, in <module>
    if sensor.get_sensor_data():
  File "/usr/local/lib/python3.7/dist-packages/bme680/__init__.py", line 244, in get_sensor_data
    status = self._get_regs(FIELD0_ADDR, 1)
  File "/usr/local/lib/python3.7/dist-packages/bme680/__init__.py", line 296, in _get_regs
    return self._i2c.read_byte_data(self.i2c_addr, register)
OSError: [Errno 121] Remote I/O error

If I don’t have the 1s delay, I typically just get the exception, after about 10 mins.
It’s plugged in to a pi400 via the breakout garden. I only have one in at a time and default 0x76.
Thanks, Richard

#12

Hi, I got it replaced in the end. Pimoroni are awesome.

I didn’t get any Tracebacks, but I only used the “read-all.py” script.

Could you try the “read-all.py” script, but change the following line as follows:

sensor.set_gas_status(bme680.DISABLE_GAS_MEAS)

This (obviously) disables the gas element. If you get solid results without error, it does sound like the sensor issue I had.

Let me know.

#13

I get the same issue with or without the gas element on with read-all.py. I’ll contact Pimoroni.
Does your new sensor work?
Thanks for your help.

#14

It’s still in the post, so fingers crossed.

#15

Using a pi 4, read-all.py with gas sensor on has been running all night. pi400 it fails.

#16

Interesting. I’ve sent my “faulty” sensor back to Pimoroni so I can’t test it on my Pi4.

Does your Pi400 have the same Raspberry OS build as your Pi4?

#17

You may have better replies if you post your issue here.
Issues · pimoroni/bme680-python · GitHub

I have two BME680’s but I’m not using the gas sensor part. Mine are in headless setups with the WIFI turned off which makes it a bit of a PITA to mess with them.
I get the occasional glitch which I “think” is an io error. I have multiple i2c devices though, and no monitor connected to see what actually happened.
On the whole though IMHO my BME680’s are pretty reliable.

#18

Thanks. Not sure if it’s the BME680. Both my sensors reliably fail in the right hand I2C slot on my breakout 400 (in a pi 400), but so far OK in the left hand I2C slot, and on the pi4.

#19

My two BME680’s have male headers soldered on, which connect to female to female jumper wires, which then plug into a male header on a Proto hat.
I have a BME280 in an i2c Breakout Garden mini, first slot on the left. It also has an RV3028 and LT-559 in it. All working just fine, as far as I know. I’m running Enviro code on them. I have the 0.96 LCD in a i2c SPI Breakout Garden Mini showing the results. The two Breakout Garden MINI’s are plugged into a Mini Black Hat Hack3r. Pi Zero W.

#20

Sounds good, I have to go away and learn about temp calibration etc, they’re miles out, out of the box. But do vary in proportion at least.