I made this circuit to amplify the output of a load cell.


simulate this circuit – Schematic created using CircuitLab

The instrumental amplifier has a Gain = 750. I use a dedicated voltage regulator (LM350) because I want an output of max 5V.

The circuit seems to work well (there's a bit of background noise), but when I'm close to the circuit the output is greater than before.

The Output change a lot when I move my arms in the air.

The cable from Load Cell to circuit are close in an aluminum foil connected to GND.

Is this the correct way to read data form load Cell? Do you know how to remove the influence of a body?

More details: Thanks everyone for suggestion (it's only the second time I use electronics.stakexchange)

The schematic shows only a single part of the PCB circuit of my project. In the real circuit, between load cells and amplifier, there are two demultiplexers. They switch signal from 8 Load Cells to 1 amplifier. (1 demux manage the +In and the other the -In). On PCB there aren't capacitors. I tried to add capacitors in the breadboard version.

The PCB without capacitors

The circuit and the load cells are mounted into a big aluminum case, not yet completely closed.

The aluminum case

In this picture the cables are uncovered, but during tests they are covered with silver foil. Every part of the aluminum case is connected to GND.

For @ANDY AKA This is what the oscilloscope sees when I put my head near to the circuit in Breadboard version.

CH 1 is the ampli supply, CH2 is the ampli output

If I set AREF to 1V and move my arm near to the wires, you are right, the output do the opposite: it decreases.

  • \$\begingroup\$ As alluded to in rawb's answer, the details of how your ground and power are connected are very important here. Could you make your schematic more explicit about how ground and power are connected, where the long wires are, how the shield is connected (one or both ends), etc? A photo of the set up might also help if you can do that. \$\endgroup\$ – The Photon Oct 17 '13 at 3:30

It is not an schematic problem. It is implementation problem. What type of prototyping you are using? IMO, this schematic have to be prototyped carefully on a PCB, not on breadboards and any kind of rat-nest type of solder-less technologies.

Note, that such high gain, precision schematics are highly sensitive to all kind of EMI and must be constructed correspondingly.

Take special attention to the input circuits - make them with screened cables that are properly grounded and as short as possible.

Yes, the instrumental amplifiers suppress very efficiently the common mode EMIs but on gain 750 and bad construction, some differential EMIs can be inducted as well.

Edit (after PCB images has been posted): This PCB is not well designed. Routing the power and ground as such a thin tracks is wrong. You must use wide copper areas for the ground and tracks as wide as possible for the power tracks. Your PCB must be almost full of copper. Also the density of the PCB seems to be a little bit small. Higher part density will short the tracks and lower the possible influences.

  • \$\begingroup\$ OK! thanks for suggestion! I will try to make a new smaller PCB (maybe double layer) using your advice. \$\endgroup\$ – Grigno Oct 18 '13 at 12:31

It would seem that you have made a very good electrometer or E-Field detection circuit. It is likely that you are coupling in the E-field from the fluorescent lights in your lab or from the mains wiring around you. You can prove this by turning off the lights and/or isolating the circuit.

You can solve this by:

  • moving the amplifier closer to the bridge to reduce the pickup.
  • extending the shield that you have to surround the sensitive devices (a faraday cage)

You will have to give more details, drawings and pictures to allow peple to fully understand your situation.

  • \$\begingroup\$ I use a halogen lamp. But I'm really close to the power supply. Do you think that using a double layer PCB with one layer connected to GND can help? \$\endgroup\$ – Grigno Oct 17 '13 at 8:57

It's a mistake in the circuit - you have REF tied to the most negative rail. This means that any ac superimposed on your load cell lines will be half-wave rectified at the output.

REF is meant to ideally lie somewhere between V pos and neg supplies. As you are using 0V as your neg supply, REF ought to be a bit higher.

Waving your arms induces small changes in the ac signal on your load cell signal which is normally averaged out to zero but, your circuit can't do this - you can prove this by placing a weight on the load cell to offset the output in a positive direction. Try waving your arms again and see what happens. I expect this will prove it.

Or maybe your circuit diagram is drawn incorrectly?

  • \$\begingroup\$ In the PCB version there is a Voltage divider that set AREF of AD623 to 0.1V that allow me to calibrate "not perfect load cells" in the Breadboard version the AREF is set to GND. I will add an image of what happens when I put my head near to the circuit. Is fun ... I know. \$\endgroup\$ – Grigno Oct 17 '13 at 8:35
  • \$\begingroup\$ Leaving detail off the circuit diagram like this is really bad. However, I still think 0.1V is too low - try setting it to half way between 0V and power rail to see what happens. If the induced ac waveform on top of the signal is 1mVp-p it will be amplified up to 750mVp-p and still become partially half-wave rectified showing itself as a dc level change. \$\endgroup\$ – Andy aka Oct 17 '13 at 8:41
  • 1
    \$\begingroup\$ @Grigno the scope picture you had up for a few moments seem to indicate it wasn't ac power influences - the frequency was too low - it looked like small changes in dc levels and I notice that your cabling (ribbon?) feeding the load-cell is unsuitable. You need 4 cores screen/shielded cable to reduce this type of problem and the screen must be connected to signal ground at the AD623 - do not earth anywhere else. Imagine if an audio engineer used ribbon cable with home made aluminium foil around it for connecting to a microphone!! \$\endgroup\$ – Andy aka Oct 17 '13 at 9:22
  • \$\begingroup\$ ok! aluminium foil was a bad experiment, you are right. The Phidget load cell are sell with this cable. Maybe I can replace most part of the cable with a 4 cores screen/shielded cable (like an ethernet cable right?). \$\endgroup\$ – Grigno Oct 17 '13 at 10:21
  • 1
    \$\begingroup\$ @Grigno you can try 4 core + screen ethernet cable but the screen may still not be good enough - copper braid and a supplemental foil screen should work OK. \$\endgroup\$ – Andy aka Oct 17 '13 at 11:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.