# understanding load cell, instrumentation amplifier

#### varuntejavarma

Joined Apr 16, 2014
36
Hello,
I want to build a weight scale with 0.1 gram resolution. But i have some problems in my initial stage.
I am using a load cell with sensitivity of 1mv/v and excitation voltage of 3 v.
1. With no weight, the load cell output shows -0,089 mv and with full load(500gms) it shows 2,895 mv. Is -0.089 mv the noise of my load cell. If so, how could i solve this problem. why is my load cell mot showing an output of 3mv for 3 v of supply.
2. I have connected the load cell to IN amp ( PMI AMP04FP) with a gain of 100, with no load, i get an output of 3,203 mv and with full load i am getting an output of 278 mv.
my q: why is the inamp showing an output of 3,203 mv. is it the offset voltage. if so how could i solve the prob. why am i getting 278mv instead of 300mv
I have gone through the basics.But couldnt get my mind clear because of lot of information in small time. I know that my questions are basic and need some reading. I would be glad if someone could explain me what is happening and how to proceed. I am willing to learn . So, any kind of help( links, books etc) is appreciated. kindly go easy on me

#### Attachments

• 1.3 MB Views: 41

#### MrChips

Joined Oct 2, 2009
28,473
Every load cell or bridge arrangement needs to be balanced to be able to remove the zero offset.

#### ericgibbs

Joined Jan 29, 2010
17,416
hi,
If you add the -0.089mV Offset to the 2.895mV at full load thats a Change of 3mV.

The 0.089mV Offset is removed by 'injecting' a +0.089mV signal into the IA input.
Give me some details of the Load Cell ie: the resistance values.
E

E

#### varuntejavarma

Joined Apr 16, 2014
36
Rated Load - 0,5Kg, Rated Output- 1.0±0.1mV/V, Zero Balance ±0.1mV/V, Input Resistance 1090±10Ω(ohms), Ouput Resistance1000±10Ω(ohms)Excitation Voltage5～10VDC, Nonlinearity0,05%F.S, Hysteresis0,05%F.S, Repeatability0,05%F.S, Creep(3min)0,05%F.S, Use Temperature–10～+40℃, Temperature Effect On Zero 0,05%F.S/10℃, Temperature Effect On Span0,05%F.S/10℃, Insulation Resistance3000MΩ(100V), Overload Capacity150%%F.S

#### varuntejavarma

Joined Apr 16, 2014
36
I want to measure the voltage changes for weight change of 0.1 gram. i.e., a change of 0.06 mv. but couldnt see it.. Is there anything i am missing

#### ericgibbs

Joined Jan 29, 2010
17,416
Hi,
Please post your full circuit, showing the component values and the IA pins.
How do you plan to convert the analog output voltage to a display.?

Rated Load - 0,5Kg, Rated Output- 1.0±0.1mV/V, Zero Balance ±0.1mV/V,

So with a IA gain of 100, a 500gm load will give a 3mV Signal * 100 = 300mV

A 1gm load will give a Signal of 0.003/500 = 6uV * 100 = 600uV

A 0.1gm load 0.003/5000 = 0.6uV *100 = 60uV

Your options are to use 10Vex and to increase the IA Gain

That would give 10mV at 500gm and say a Gain of [330 for a 3.3V ADC or 500 for a 5V ADC]

Post more details of the project.

E
EDIT:
Note: the stated values: Nonlinearity0,05%F.S, Hysteresis0,05%F.S in gms FS

Last edited:

#### varuntejavarma

Joined Apr 16, 2014
36
I am planning to connect the o/p of IA to 24 bit ADC( hycon 3106). I control the ADC with PIC18f4550 mcu through SPI. I am not done with the ADC and uc communication yet. I am still working on it.

The ADC has max voltage upto 3.6v. So, I thought of using 3v for supply and also for reference voltage. As, i am using 3v for ref voltage , I thought it would be better to use the same for load cell excitation voltage. am i right?? Plz correct me if not.

I am using a PIC demo board with 5v. so, had to use LDO(3.3v) for vdd of ADC.

The ADC also has a PGA (upto 128). So, with this option available, how much gain would be better on IA??

http://www.hycontek.com/attachments/DataConverters/DS-HY3106_EN.pdf

I have already burned two ADC's by giving more than 3.6v at ref volt( without knowing).

how much voltage source should i choose for individual components. or should i choose a single source for evrything??

how much gain should i choose in IA and in ADC??

#### Attachments

• 1.2 MB Views: 31

#### ericgibbs

Joined Jan 29, 2010
17,416
hi,
Its advisable to use the same voltage source for the bridge and the ADC's Vref, as most ADC's are ratiometric.
I do not think that its possible to measure to +/-0.1gm using your existing 500gm load cell.
As I have posted, 0.1gm is equal to a bridge signal of only 0.6microVolt.

The spec for the load cell gives an Hysteresis/Linearity figure of 0.05% of Full Scale which is in the order of 0.25gm.??

Is it essential that the Full scale load has to be 500gm.?

Eric

#### varuntejavarma

Joined Apr 16, 2014
36
hi,

I have tested the loadcell with spider signal conditioner. It works well for small weights. I would like to know if I have to change anything in my INamp circuit to detect small changes.
do you think the INamp works fine with 3v or should i increase??

what about the gain settings. how much gain do you think is better considering the availability of PGA of ADC.

#### ericgibbs

Joined Jan 29, 2010
17,416
I have tested the loadcell with spider signal conditioner. It works well for small weights. I would like to know if I have to change anything in my INamp circuit to detect small changes.
What is the smallest change in gms, that the load cell will detect.? using your test rig.

#### ErnieM

Joined Apr 24, 2011
8,355
how much gain should i choose in IA and in ADC??
That depends on what the maximum weigh you wish to measure. You want to aim for the max to be below the ref voltage, as that is the max reading you can make.

I oft put this somewhat below, some 5-10%. That way I have room to make an over range measurement, and can always tell a max reading from an over range or even failure reading.

The SPI interface from the A2d and the PIC needs work: you are sending digital signals between 3.3V and 5V domains. I have seen this done with resistors, with voltage translating ICs, and also by running all the parts off the same (3.3V) power.

#### ericgibbs

Joined Jan 29, 2010
17,416
hi,
Set up your project in this way for a test.

Use 5Vx for the Load cell, AMP04 and Vref for the ADC, PIC18F

Note: the AMP04 d/s states, 5V min for for a single supply, which will give a Vmax out of ~4V and a Low of ~ 2mV, with a 2K load.

The L/C will should give 5mV out with a 500gm load.

To increase this to say 3V out the AMP04 gain should be set for 600.

Check:
Let the project warm up for about 5 minutes.
Place a 250gm load on the L/C, the AMP04 should output ~ 1.5V, don't worry about any offset of a few mV at this time.
Note precisely the Vout of the AMP04 ,, if you have the ADC and PIC 18F connected and programmed to display the weight, note the displayed weight.

Add a 1gm weight to the 250gm on the L/C note the change in Vout and the displayed weight, then remove the 1gm and compare the weight with the original 250gm weight...

Repeat the above using 0.1mg weight.

Post the readings that you have measured.

We are checking the Hysteresis and Sensitivity of your set up
Eric

#### varuntejavarma

Joined Apr 16, 2014
36
The SPI interface from the A2d and the PIC needs work: you are sending digital signals between 3.3V and 5V domains. I have seen this done with resistors, with voltage translating ICs, and also by running all the parts off the same (3.3V) power.

Ok. According to the diagram in my adc datasheet i had to connect both analog and digital voltage sources to the adc. I have a demo board with 5v and my adc can sustain only upto 3.6 v. So, i have used a ldo (3.3 v) for digital voltage source from mcu to adc. the analog voltage given to adc is 3v. the rest of my circuit is working at 3v( load cell, INamp). So, am i doing something wrong here. plz tell me what to do

#### varuntejavarma

Joined Apr 16, 2014
36
Hi ericgibbs,

I had an input offset of -0.089mv at loadcell. so, i placed some weights untill i read 0 at output wires.

At 5v source, with 500gms load, L/c outputs 4.886 mv
with 602 gain, 500gms load, AMP04 gives 2.85v

with 250 gms load, Amp04 gives 1.3741 - 1.3750 v
with 1 grm added, 1.3795 - 1.3810 v
with 0.1 grm added, 1.3750 - 1.3760 v

But the amp04 output is not constant, it keeps on varying. so it is kind of difficult to find out the exact value when using less weights.

I am not successfull yet with adc and pic interfacing(no correct reference available for spi interfacing for my ADC). still working on it.

Last edited:

#### ericgibbs

Joined Jan 29, 2010
17,416
Hi ericgibbs,

At 5v source, with 500gms load, L/c outputs 4.886 mv
with 602 gain, 500gms load, AMP04 gives 2.85v

with 250 gms load, Amp04 gives 1.3741 - 1.3750 mv [V?]
These values sound incorrect.?
Example:
You say, 500gm = 4.886mV and AMP04 Vout= 2.85V

I would expect for a 250gm weight half those values.
ie: 2.443mV and AMP04 Vout = 1.425V

with 1 grm added, 1.3795 - 1.3810 mv
with 0.1 grm added, 1.3750 - 1.3760 mv

But the amp04 output is not constant, it keeps on varying. so it is kind of difficult to find out the exact value when using less weights.

I am not successfull yet with adc and pic interfacing(no correct reference available for spi interfacing for my ADC). still working on it.
Hi,
Have you incorrectly named the values mV and not V.?
with 1 grm added, 1.3795 - 1.3810 mv
with 0.1 grm added, 1.3750 - 1.3760 mv

What electrical noise filtering do you have on the AMP04 input pins.??

E

#### varuntejavarma

Joined Apr 16, 2014
36
sorry, it was v not mv. I have changed it. I have no noise filtering at the inputs. whta kind of noise filtering do you think i should use??

#### ericgibbs

Joined Jan 29, 2010
17,416
sorry, it was v not mv. I have changed it. I have no noise filtering at the inputs. whta kind of noise filtering do you think i should use??
hi.
Look at page #8 of the datasheet for the AMP04 for noise limiting.

For reference only I have modeled the AMP04 in LTSpice, ref the attached imagines. It will give you a good idea of what to expect from your circuit
One is with a 0mV to 5mV signal from the L/C and the other with a 1uV Step superimposed on a 2.5mV level.
Gain = 600.
Ignore my input R/C filter.

Did you redo those checks and get new values.?

E

#### Attachments

• 40.5 KB Views: 34
• 40.3 KB Views: 37
Last edited:

#### varuntejavarma

Joined Apr 16, 2014
36
I read the datsheet. so, my input noise depends on the gain. Do you think its better to lower the gain in INamp and set more gain in ADC. But i think it would be a problem for ADC to detect small signals

I have checked with the values. when i change 1 uv at Inamp input i can see some change in mv in INamp output but not sure how much. because it is never stable. it changes rapidly

#### ericgibbs

Joined Jan 29, 2010
17,416
hi,
If there is non common mode 'noise' entering the L/C and its wiring it will be amplified and cause the amplified voltage to vary.

Could you post a full circuit diagram showing your power supplies etc and if possible a photo of the project.?
When measuring such low level signals, good grounding, power decoupling and layout are critical.

E

#### varuntejavarma

Joined Apr 16, 2014
36
hi,

I have attached the circuit diagram and photos of the project.

the first picture shows the connection of loadcell to the amplifier. The length of load cell cable is 1 metre long.

The small smd chip is the ADC.The yellow wires from ADC are the inputs (VINN1, VINP1). I am connecting them to loadcell output as for now for testing ADC. Later VINP1 is going to pin6 of INamp as per circuit diagram.

#### Attachments

• 1.2 MB Views: 50
• 1.5 MB Views: 35
• 1.7 MB Views: 32
• 1.3 MB Views: 29