load cell drift: humidity

Hi everybody.

I'm using

Battery 3.7v
Voltage booster
Arduino mini 5v
HX711
Wireless module
Single point Load Cell
BME280 sensors

The weight is drifting with humidity, at the beginning I thouth it could be something else, but now with all the graphs on the screen ! Is it possible? I never heard about humidity drifting, maybe temperature. Is it the load cell or the HX711 causing the problem?

from 30% umidity to 50% the weight change is 80 grams on a 20.000 grams weight.

thanks

Load cells are sensitive to temperature, normally wouldn't expect any effect from humidity. Is the load
cell a half-bridge? In that case I'd expect temperature effects to balance out.

If its a single element balanced by a fixed resistor, that resistor must be a precision metal film 0.1%
tolerance or better, otherwise you will see the kind of effect described (high sensitivity to temperature).

If you use carbon comp resistors you will only have enormous grief.

You didn't post details of the load cell in question, so I have to guess - always post full details of all
the hardware if you want informed responses...

I use a single point load cell from a Chinese manufacturer, the specs seems to be good: aluminium, IP54, Class 3, temp compensated, etc. all the things u would expected from a good load cell, really don't know wich kind of resistor they used to temp compensate.

There are some poor load cells and HX711 amps about - I bought a cheap Chinese job and it drifts all over the place

What do you mean with "a single point load cell"
A half bridge with two fixed resistors?
Are they 1% metalfilm, and from the same batch?
Is VCC of the HX711 >= 4.75volt, so the excitation voltage regulator of that chip can work properly.
And did you power the load cell and resistors from that E+ E- voltage?

Do you expect us to give you proper advice without seeing a connection diagram, a picture of the setup, links to the parts used, and code?
Did you read the "how to post" sticky that's on every page?
Leo..

Sorry guys, i was so concentrated on the humidity effect that i forgot to post some details. Unfortunately i don't have the specifications for the items, i can post the links:

I have connected the Load cell to Hx711 this way:
Black wire to E+, white to E-, Red to A+, green to A-
The Hx711 VCC is connected directly to the booster + output pin, and the GND to - output
Arduino is connected to the booster through the raw pin and GND pin, pin2 and pin3 to Hx711 as specified in the library.
I have used this library: GitHub - bogde/HX711: An Arduino library to interface the Avia Semiconductor HX711 24-Bit Analog-to-Digital Converter (ADC) for Weight Scales.

For the power circuit i have followed this instructable:

I have attached 2 diagrams of the data collected in 5 days. The weight is 20Kg of metal and the weight scale was inside the house. The trend of weight curve follows almost perfectly the variation of the humidity. How can it be possible? I can buy a new load cell, or a different Hx711 board, which one is to blame?

I know there is something better than Hx711, but I am not sure i am able to build this: A Reference Design for High-Performance, Low-Cost Weigh Scales | Analog Devices

If i had experienced temperature drift i would have bought a better cell, but why do i have humidity drift?

There are no standards for load cell wire colours.
Are you sure you have it right.
Did the load cell come with a datasheet.

Disconnect the load cell, and measure the resistance between the wires.
The high-resistance pairs are (E+ E-) and (A+ A-)
Could be red for E+ and black for E-
Leo..

Yes I am sure the wiring to the load cell is right, I checked it many times. At first I connected the hx711 with jumper wires to the load cell cable, but then I soldered those wires on the hx711 board to avoid interference.
I have followed the wiring information written in the link I provided earlier, where you can also find the specifications. But no internal details are provided about resistors and withstone bridge. Do resistors modify their behaviour with humidity? unfortunately I have little knowledge of electronics.

This is the load cell with specs and wiring: https://www.aliexpress.com/item/1PCS-X-100KG-150KG-electronic-platform-scale-load-cell-pressure-balanced-cantilever-load-weight-sensor/32344819630.html?spm=a2g0s.9042311.0.0.3da24c4dycyyl1

Humidity is closely related to temperature.
And a load cell bridge can be rotated and still work.
But could only be temp compenstated in one orientation.
Leo..

Thanks a lot for your replies, can you be more specific? I connected the cables as specified by the manufacturer. Maybe I did something wrong.

thanks again bye

I did some more tests. Load cell seems to be not responsible for the drift. Is the hx711 board. I used an hair drier to heat the board the weight changed enormously. 30% or even higher. I heated the board till 45 Celsius. I don't know why or wich component is causing this change in the output of the hx711 board. Maybe even humidity affect this board. And the load Cel has nothing to do with incorrect reafings

You said you use a Lipo and a voltage booster.
Did you measure that volage.
The HX711 needs 5volt to make a stable ~4.2volt excitation (E+).

Did you solder the wires to the board yourself, and did you use rosincore (elecronics) solder.
NO paste or flux of any kind.
Leo..

Voltage is very stable at 5 volts, i did not solder the load cell is connectected with terminal blocks wire to board, i even tried with no load cell connected, the value is still drifting and unstable, so i took another Arduino, connected directly to usb port without booster without battery, just Arduino hx711 load cell, same happening, if I heat the hx711 the load will heavily drift.

I changed the load cell interface now I'm using ADS1230 inside this module.

Still experiencing drift over time, here the graph

Is this normal?

Can you help?

the graph is about a fixed 20.000 grams load made of steel