HX711 and Strain Gauge HW debugging

Waste of time.
If you want to tinker, then modify the board to Sparkfun's standard.
It's not as hard as you think, assuming you have basic soldering skills.
Leo..

I don't have solid soldering skills but I don't fear it :grinning:

Following your advices I've already bought sparkfun load cell amplifier and until it will not arrive I'm without work :weary:

@Wawa Could you help me sharing something to adapt my hx711, please? I've googled and I found this article but you are not agree with that..

Not a good way.
Lowering excitation voltage is degrading the performance of the setup.

If you don't have any soldering skills, then wait for the Sparkfun board.
Because modifying that green board includes de-soldering and lifting pin1 of the HX711.
Leo..

Also, if the photo in the OP's Amazon link is accurate (and the OP's voltage measurements are accurate), the OP's green board may be another one of those with floating E- (as is the board in the honeybee hive monitoring project).

OP's board from Amazon:

Honeybee board:

Board known to have floating E-:

This is what I would do to modify that board.

  1. de-solder and lift pin1 (near the dimple) of the chip from the board.
  2. scrape/expose the copper of the long trace (VCC) near the dimple, and connect to pin1.
  3. cut away the short trace between C8 and the long VCC trace.
    Now the VCC pin of the board is the 5volt (AVDD) connection.
    You should add a 100n decoupling cap from VCC (pin1) to ground.
  4. connect a wire to pin16 of the chip (C8). This is the 3.3volt (DVDD) connection.
  5. solder a wire between E- and GND. This fixes the design fault of that board.
    Leo..

I play with GIMP in order to be sure about the changes to do. :grinning:

It should be the pin in red circle below:
image

I cut the link toward Q2 and I connect VCC to pin1.
image

I cut the link between C6 (you said C8 but I think it's C6) and I add decoupling cap 100n.
image

The wire will be connected to the micro.
image

At the end, E- will be connected to GND (probably by the back).
image

Could you confirm that's all right?

Not what I said.
The link to Q2 must remain.

Correct, it looked like an 8.
Leo..

OK!

I post again the result expected.
image

So, this image comes from @DaveEvans.
My HX711 is a little bit different from it.


Anyway, I will try to change it as described.
I will give feedback as soon as done.

Your board is indeed different.

Basically, pin1 (dimple) of the chip and the top pin of Q1 must be connected to 5volt
(with that 100n decoupling cap to ground).

And pin16 must be connected to 3.3volt (with C5 as decoupling cap).

The problem is that there is a track under the chip that connects pin1 to pin16, which has to be cut.
Leo..

Today it's arrived sparkfun load cell amp.

I've done more tests again but the result is not changed: the values were nan or not stable (eg: remove weight from the balance there was the same value).

I consider untrustable sensors (and load cell amp) and I've already started the return procedure.

Do you suggest continuing with the four strain gauge load cell or bar type load cell?

The field application is a beehive scale.

Did you consider user error? (most likely).

Did you also use the combinator, or did you join the load cells yourself.
Did you measure the resistance of each load cell, and are they the same.
What values are they (include wire colours).
Did you measure voltage between E+ and E- (should be 4.3volt).
Did you measure voltage between A+ and A- (should be 0.0volt).
Post clear pictures of the setup.
Leo..

I suggest you continue with a simpler project. Placement and connection of such sensitive sensors and amplifiers requires some experience which you may collect during dedicated experiments with e.g. a single load cell.

Not to hijack the thread, but I'm trying to wire this same circuit for a school project. I've purchased the Sparkfun load cells, load cell combinator, and HX711 amplifier board. After wiring the load cells to the combinator board, the A+ to A- voltage difference is non-zero (1.6volt). I've looked at all of Sparkfun's guide's and unless I'm making a stupid error, I don't know what the problem could be.

My wiring diagram:

Physical hookups:

I'm powering it from my microcontroller's 5V/GND outputs. I also measured the resistances between each load cell wire and listed them in the diagram below (red is the center wire for all).

Is there anything you see that I'm missing? Thanks in advance, I appreciate the help!

What happens if you swap cells?

Swapping any two cells results in the same 1.6V output. Reversing the pin header orientation (white on right, black on left) also doesn't affect the output. Removing any load cell from the combinator results in near zero output (0 to 5mV). Perhaps I could try bypassing the combinator and wiring the wheatstone bridge myself? Is there an easy way to test each individual load cell?

I also just tried connecting each load cell to +5V and GND on +/- leads and measured the center wire. All 4 readings were around 2.5V as I expected.

EDIT: after bypassing the combinator and self-wiring the wheatstone bridge (following this datasheet), I'm still getting 1.6V on the output.

I guess that you mixed up the combinator pins. Check the voltages on the boards with regards to GND.

Spot on actually, the combinator board lists the pins as "C+-" and I had them wired as "+C-". I fixed this and the voltage reads near zero on the output, thanks for your help!

Fine :slight_smile:

You can mark the topic resolved.