amplifying a signal from a load cell with ad623

Hi,

I’m trying to build a force sensor using a load cell from a kitchen scale for a project (I'm a marine biologist with no experience with electronics). My ultimate goal is to make a stand-alone force sensor and data logger to measure wave forces in a coastal area.
However I'm stuck in the first steps of the project.

The output signal from the sensor is 0.44 mV (at rest) to 18,2 mV (with load), and I’m trying to amplify it with an ad623 amplifier (Gain=100).
I don’t seem to be having much success with the amplification step. As the output voltage is -0.93V, for whatever gain I use, and regardless of the load I put on the loadcell.

Can you tell me what am I doing wrong?

The circuit I made is as follows:

**broken link removed**


Thanks!

djacinto said:

Hi,

I’m trying to build a force sensor using a load cell from a kitchen scale for a project (I'm a marine biologist with no experience with electronics). My ultimate goal is to make a stand-alone force sensor and data logger to measure wave forces in a coastal area.
However I'm stuck in the first steps of the project.

The output signal from the sensor is 0.44 mV (at rest) to 18,2 mV (with load), and I’m trying to amplify it with an ad623 amplifier (Gain=100).
I don’t seem to be having much success with the amplification step. As the output voltage is -0.93V, for whatever gain I use, and regardless of the load I put on the loadcell.

Can you tell me what am I doing wrong?

Thanks!

Click to expand...

hi,
Is the load cell a bridge.? or a strain gauge element.?

I believe it's a bridge. Since there are 4 wires running into it, black, red, white and green.
Also it works pretty well when force is applied, however the output signal is within the mV range

djacinto said:

I believe it's a bridge. Since there are 4 wires running into it, black, red, white and green.
Also it works pretty well when force is applied, however the output signal is within the mV range

Click to expand...

The output levels are typically in the mV region.

Why is the REF pin #5, floating.? It should be at 0V for zero offset.. try it.

Do you mean wiring pin #5 to the "-" (black) pole of the battery. Is that it?

djacinto said:

Do you mean wiring pin #5 to the "-" (black) pole of the battery. Is that it?

Click to expand...

hi,
Link pin 5 to pin 4.. thats 0V.

Then measure the output of the AD623 from pin 6 to pin 5/4 [0v] OK.?

hi,
Did that wiring change give the output you were expecting.?

You should get about +1.8V at full load.

hi,

I followed your suggestion. But there's a strange result...
The output is now 34mV, even if I apply some force on the load cell, or if change the gain, or even if I unwire the sensor from the circuit.

Do you know why is this happening?

thanks.

djacinto said:

hi,

I followed your suggestion. But there's a strange result...
The output is now 34mV, even if I apply some force on the load cell, or if change the gain, or even if I unwire the sensor from the circuit.

Do you know why is this happening?

thanks.

Click to expand...

hi, You are working from a single battery supply its likely that the bridge signal is trying to drive the AD623 below zero.

The cure is to reverse the output from the bridge wires to the AD623 input.
This should give a +V output from the AD623..

If you dont follow, please ask.

EDIT:
with your meter what dc voltages do you measure on the AD623 input pins from the bridge.??

hi,
This is how your circuit should look and measure.

Are you wanting do this yourself to save money or learn about processing load cell signals? If you have $325 USD, look at the D15xx models here: Omega D1000 Digital Transmitters or you can use a scale indicator.

If you want to do this yourself, I would recommend searching for "instrumentation amplifiers". The datasheets often have diagrams for strain gauge applications.
Dale

ericgibbs said:

hi,
This is how your circuit should look and measure.

Click to expand...

hi eric,

thanks a lot... I followed your circuit, and voilá... it worked!

I guess that one of the things I was doing wrong was to use 2 different batteries to power the load cell and the ad623. (I tried using both batteries with the circuit you showed me and it didn't work)

This way it seems to work just fine... thanks again.

By the way, is the capacitor really needed? What is the purpose of using it?
(without the capacitor, I get to the same results)

I really enjoyed this experience with electro-tech, I suppose I'll came back soon, with new questions.

thanks,
david

DSGarcia said:

Are you wanting do this yourself to save money or learn about processing load cell signals? If you have $325 USD, look at the D15xx models here: Omega D1000 Digital Transmitters or you can use a scale indicator.

If you want to do this yourself, I would recommend searching for "instrumentation amplifiers". The datasheets often have diagrams for strain gauge applications.
Dale

Click to expand...

hi dale,

Thanks for the advice. I'm really trying to do this myself, and to learn a little about this "strange" world of electronics.

david

djacinto said:

Thanks for the advice. I'm really trying to do this myself, and to learn a little about this "strange" world of electronics.

david

Click to expand...

I'm glad you are learning. I didn't know if you wanted to learn or solve a problem quickly.
Dale

djacinto said:

hi eric,

thanks a lot... I followed your circuit, and voilá... it worked!

I guess that one of the things I was doing wrong was to use 2 different batteries to power the load cell and the ad623. (I tried using both batteries with the circuit you showed me and it didn't work)

This way it seems to work just fine... thanks again.

By the way, is the capacitor really needed? What is the purpose of using it?
(without the capacitor, I get to the same results)

I really enjoyed this experience with electro-tech, I suppose I'll came back soon, with new questions.

thanks,
david

Click to expand...

hi,
The capacitor on the power line is to decouple the supply, it should be fitted.
It may work without it, but in many cases not fitting a cap could give unexpected circuit performance.
Its good engineering practice to fit decoupling to all power lines.

It is possible to use two batteries, one for the bridge and one fror the AD623,
but its important that both battery -V [0v ref] are linked together.
That is, a commoned 0V [-Vbty] lines.

Well done.

Hi Eric and others,
I am also wanting to amplify a load cell voltage using an AD263 but are having issues getting it to amplify the load cell voltage. The circuit is set up as in your example, message response #5 from Eric Gibbs.
The following is what i have done /observed;
Supply voltage for ad623 is 5v (regulated).
Supply voltage for load cell is 12v (regulated).
Common mode voltage is 5.76V (should be 6v ???).
Load cell gound and ad623 gound are connected and a decoupling capacitor is used.
Load cell works and produces output in the range 0 - 36mV (3mV/V), i require amplified voltage in the range 0 - 5v (gain = 138 so have have set Rg to 748 ohms via a multiturn potentiometer).
Output from the ad623 is ~35mV no matter which way i connect the output wires from the load cell bridge (green/white).
Circuit is relatively simple so cannot see where the eror is that is preventing it from producing an amplified votlage. The only thing that is a bit odd is the common mode voltage which should be 6v for a 12 regulated supply not 5.76V.
Have also tried different ad623 units with same result. The only thing i havent tried is fresh batteries (Currently 2 x 9v batteries in series but are getting a bit dead - producing 16v but are running other suff OK - PIC 18F micro for example and still maintains load cell at 12V via regulator).

Any help is much appreciated.

Cheers.

katesfb said:

Hi Eric and others,
I am also wanting to amplify a load cell voltage using an AD263 but are having issues getting it to amplify the load cell voltage. The circuit is set up as in your example, message response #5 from Eric Gibbs.
The following is what i have done /observed;
Supply voltage for ad623 is 5v (regulated).
Supply voltage for load cell is 12v (regulated) .
Common mode voltage is 5.76V (should be 6v ???).
Load cell gound and ad623 gound are connected and a decoupling capacitor is used.
Load cell works and produces output in the range 0 - 36mV (3mV/V), i require amplified voltage in the range 0 - 5v (gain = 138 so have have set Rg to 748 ohms via a multiturn potentiometer).
Output from the ad623 is ~35mV no matter which way i connect the output wires from the load cell bridge (green/white).
Circuit is relatively simple so cannot see where the eror is that is preventing it from producing an amplified votlage. The only thing that is a bit odd is the common mode voltage which should be 6v for a 12 regulated supply not 5.76V.
Have also tried different ad623 units with same result. The only thing i havent tried is fresh batteries (Currently 2 x 9v batteries in series but are getting a bit dead - producing 16v but are running other suff OK - PIC 18F micro for example and still maintains load cell at 12V via regulator).

Any help is much appreciated.

Cheers.

Click to expand...

hi,
Your main problem is the AD623 powered by only 5V and the Bridge by 12V, that will give a common mode voltage of 6V at the input of the AD623.

I would use a 12V to the AD623.

Reducing the Bridge to 5Vx will reduce the bridge sensitivity.!

hi,
These two images show the result of a AD623 at 5V and 12V supply.

Adjusted the gain a little, 5V out for 36mV input

NOTE: If you use a MCP6002 it has a maximum supply voltage of 5.5V [ I have used 12V for the demo], a CA3240 would be OK at 12V

The circuit I use for loadcell amplification is 5v excitation on a 350ohm bridge with a INA126P with the reference at 1.25V the gain set to x330. It's more than good enough for my purposes. My cell is usually around .75mV/V.

Hi Eric etc
And thanks for the replies. Did not even consider effect of higher input voltage than supply rail for ad623. So thanks a lot for spotting that rather amateur error. I will make the necessary modifications and let you know how it goes.

I have one other question - what is the advantage of the input circuitry you have used for the input from the load cell to the ad623 as opposed to just directly connecting the sensor wires.

Thanks again,

Cheers.