I went to my local electronics supplier and they were kind enough to show me how to correct it (and correct one in the process), so my HX711 number 2 now resembles this:
If I connect the load cell, it goes back to "HX711 reading: 8388607" and doesn't budge whatever force I apply to this [insert vulgar word] sensor from hell
After a few minutes, the HX711 also gets quite hot, so it can't be good.
Everyone seems pretty OK with these, so why the heck does it behave like that with me? is there no mercy for the beginners?
I don't think the board gets hot without the loadcell connected to it. What does it mean in your opinion ?
I'm done with this for today (Europe, 17:30, 35°C, time for a fresh glass of wine to honor the concept of the "jeudredi" (jeudi = Thursday, vendredi = Friday) ;) Tomorrow i'll solder a fresh HX711 and give it another go. At some point, I'll solve this, but what a pain !
You connected it wrong or the loadcell has a short inside or the HX711 board is bad, could be a lot of things.
If you have an ohmmeter/multimeter you can check the load cell. You should see a fixed resistance between any combination of the four wires ( green-white, green-red, white-red etc.). The resistance will depend on the loadcell but is probably somewhere between 350 and 1000 ohms.
If the loadcell is good, I'd try a different HX711 board.
I measure 350 Ohms between red and green and 402 Ohms between black and white, which is weird since, the way i understand the specs, it should be the other way around... no ?
I have been soldering for 60+ years and I never use flux for this sort of thing.
Some green boards have a manufacturing fault, where the ground plane is missing.
Test for continuity with your DMM between E- and ground on opposite sides of the board.
It should be close to zero Ohms.
Faulty boards are also transparent if you hold a string light behind it.
They can be easily fixed by adding a short wire between E- and GND
@Wawa : When I put the DMM in continuity mode, it doesn't beep between E- and GND (both sides of the HX711). when tested for resistance, it reads about 1500 on the 2000k scale of the multimeter (not sure how to read that exactly, but I'd say "A lot")
Can anyone confirm that I should have continuity there and that I should solder a small wire to connect both? (would make sense... Ground is ground after all, but I'd rather make sure that I won't mess up something...)
Load cells resistances (for both load cells, one already been plugged, the other brand new):
Red-Green: 351/351
Red-White: 288/288
Red-Black: 288/289
Green-White: 288/288
Green-Black: 288/289
White-black: 401/402
--> seems at least consistent, but I'm not entirely sure which pair (red/green or white/black) is the power and which is the signal...
This what the HX711 looks like when exposed to light
Yes, you definitely have the board with manufacturing defect.
Just solder a ~4cm long wire across the board (bottom side) between E- and GND.
After that you should have 4.25volt between E+ and E-
It's not a trick. This fault is well documented all over the internet.
Can't explain the hot HX711. Shouldn't get hot even before the mod. See what it does after the fix.
It's weird how - at some point in a troubleshooting like that - you somehow kind of lose hope. It seems that whatever you do, you'll always end up in the same place
Yet, my load cell (on of the 3 "originals") is now connected to the HX711 (number 5), clamped to my desk and the example from the library reads between 67450 et 67500. If I press on it, it goes down, if I lift it, it goes up...
Try first without a load cell. Measure the voltage between E+ and E- (4.25V)
And see if the chip gets hot.
If the voltage is Ok and the chip stays cold, only then connect the load cell.