Wiring connections

[drclong]

Hi GeekyDeaks, Thanks heaps for your project, I am in the process of construction but am unsure of the wiring connections. I am a complete novice with electronics so please bear with me. I am using single load cell setup. Could you please clarify the following?

1. Is the cap still required with 1 cell?
2. I am confused as to exactly which wire goes where as I thought the load cell would have a ground connection.

Looking at the instructions, it seems that the cell connections are as follows:
Red -> 3.3v
Black -> V+ (V- on PCB?)
White -> V- (V+ on PCB?)
No GND?
Is this correct?

Also "R_REF should be replaced with a link from the middle pin to the right pin" is that looking from the top or bottom of the board?
Thanks so much for your help.

[rjm83]

I used the Leo Bodnar board and not a custom PCB, but I wired it as 5v to white, red to V-, and gnd to black. I split from ground/black with a 1k resistor and from white/5v with another 1k resistor then brought together the other side of the two resistors to single wire into v+. This completes the wheatstone bridge and the difference in voltage from v- and v+ is the load registered. Reference the wheatstone bridge image shown in the original readme or in other issues/comments.

[drclong]

I used the Leo Bodnar board and not a custom PCB, but I wired it as 5v to white, red to V-, and gnd to black. I split from ground/black with a 1k resistor and from white/5v with another 1k resistor then brought together the other side of the two resistors to single wire into v+. This completes the wheatstone bridge and the difference in voltage from v- and v+ is the load registered. Reference the wheatstone bridge image shown in the original readme or in other issues/comments.

Thanks for the reply but I am trying to copy this circuit exactly and there are no images of the finished board and wires that I can find.
Cheers.

[GeekyDeaks]

Hi @drclong - you are correct. The PCB design shared on the project page does not connect GND directly to the Loadcell when there is only one. I'll go into specifics in a bit, but here are the main points you raised:

• R_REF is directly connected to GND (this was used for the dual load cell as we ran the amp in a negative gain configuration)
• Install the cap C1 if you can - it's just for decoupling the amp from the power to help smooth the supply to it, but I found the circuit wasn't particularly noisy so you may get away with it

That's a good call on the constructed board. I'll try and make another one up this weekend and take a photo. If you want a PCB, I have some spare and I'm happy to post one to you.

To go into a little more detail around why there is no GND connection to the load cell. It helps to remember that the load cell is simply just two resistors in series and not an active device. In the model used in this project, the red wire goes to the mid point between the two resistors and the black and white are connected to the ends of the resistors. Since there are only two resistors in a single load cell, this means you need two more to construct the entire wheatstone bridge. The post by @rjm83 highlights the fact that there is more than one way to do this. I chose to have the load cell take the place of the upper two resistors (with respect to V+ and GND), but @rjm83 chose to have the load cell take the place of two resistors to a side (this might have been dictated by the circuit on the bodnar board). If you take a closer look at the circuit diagram, you will see that the load cell is connected to GND via R2 + R+BAL1 and R1. These two resistors and trimpot are the lower part of the wheatstone bridge.

[tazea]

Hi, sorry to revive this old but gold thread. Im wiring a strain gauge as rjm83 did, however ive got a very small change in the output of the INA, resulting in a very poor resolution. I believe using two strain gauges as half bridge would increase the sensitivity of the bride, but, is there any way to increase the output with only one cell?

Ive tested a 4 wire loadcell and the output is perfect, 0 to 5v but im trying to achieve a considerable same result with only one strain gauge, is that possible?

Thank you..

[GeekyDeaks]

hi @tazea ! Whilst using one load cell will reduce the resolution, at worst it should be half of that with two and in practice I found this was something I could not perceive through the pedal. You should still be able to double the amplification used for the dual load cell to get the full range with a single, so it sounds like something isn't quite right if the output from the amp cannot get close to 5v. Are you able to provide some pictures of your setup? I'd probably start with checking what the mV input to the amp is like when idle and under load with both the dual and single load cells to see if there is a big disparity there. It might be that the single load cell is -ve when idle due to tolerances on the other side of the bridge, which would reduce the range.

[neilser]

Hi, sorry to revive this old but gold thread. Im wiring a strain gauge as rjm83 did, however ive got a very small change in the output of the INA, resulting in a very poor resolution. I believe using two strain gauges as half bridge would increase the sensitivity of the bride, but, is there any way to increase the output with only one cell?

Ive tested a 4 wire loadcell and the output is perfect, 0 to 5v but im trying to achieve a considerable same result with only one strain gauge, is that possible?

Just in case there's a misunderstanding here: each of the bathroom-scales type of load cell is in fact a pair of strain gauges (in a half-bridge arrangement). You mention "only one strain gauge" and "only one cell" in a way that makes me wonder if you are considering them as being equivalent. Perhaps though you are using a different type of load cell with only a single strain gauge?

[tazea]

@GeekyDeaks Thank you for your fast reply! so far its working better now, i measured the output of the strain gauge without no load and it was around 24.5 mV, so i missed the "unbalance" the strain gauge thing to balance its output to 0 when no load. Ill keep working on it tomorrow to polish some things. so far so good.

@neilser sorry for the missunderstaing, i meant 2 strain gauges as a full bridge, i dont know why i said half.