Load Cell Output

[willeng]

I have aqcuired a load cell with a digital scale setup at the right price.
I want to use the load cell for another purpose so I need to build a circuit for it & then check it from the scale readout to verify things.

I have the datasheet Attached for the load cell, & have a few questions to get a better understanding of things before attempting the circuit.

The Rated Output on the Datasheet is 3.05mV/V:

Could someone explain this for me, I know what 3.05mV is, to me it reads 3.05mV/Per Volt?

I'm not sure if I have made the correct decision regarding the instrument amplifier chip but I purchased an INA125P.

Is this suitable?

On the datasheet - Attached:

I see that the more gain you have the greater the gain error, so is it best to have low gain from the chip for accuracy & then amplify it again with a precision opamp or does that defeat the purpose of the chip?

Or is that the correct way to do this?

Thanks

 

[JimB]

Regarding the units mV/V,

The load cell has a bridge circuit with an input and an output, so for every volt of excitation on the input you will get 3.05mV on the output when the cell is loaded to its rated capacity.

An example;

The BSS-10K cell, when loaded to 10,000lb and with an excitation voltage of 5 volts, will give an output of 3.05 x 5 = 15.25mV.

Easy?


As fas as the "gain error" goes, I think it is a case of swings and roundabouts.
The instrument amplifier is a precision device, if you use it at low gain and then use another op-amp to give any further gain required, the errors in the other amp could be worse than those in your instrument amp.
If anyone else on the board has more experience in this area, I stand to be corrected.

JimB

 

[willeng]

Thanks JimB,

I was unsure if the rated output was at maximum capacity, although it makes perfect sense now that it is in black & white.

Don't I require some sort of buffer amplifier on the output though?

Thanks

 

[ericgibbs]

hi willeng,

A couple of questions.

What supply voltages do you plan to use with the 1NA125.??

What are you connecting the output of the1N125 into.??

E.
Sorry for butting in Jim, its my area of interest, hope you dont mind.

 

[JimB]

Sorry for butting in Jim, its my area of interest, hope you dont mind.

Not "butting in" Eric, I guess you are just adding to the thread, I don't regard it as a private conversation.
If you have related information, by all means lets hear it.


Willeng wrote:

Don't I require some sort of buffer amplifier on the output though?

I guess that all depends what you what to do with the output.

A simple op-amp on the output of the in-amp would offer a good opportunity to perform some scaling and offset on the signal if required.
Or, just a cheap op-amp set as a unity gain buffer would offer some protection from an "Ooh, I never thought of that!" situation smoking the more expensive in-amp when you do something silly with the output.

JimB

 

[willeng]

Thanks JimB,

I'll have to look into what you mentioned as my electronics skills are very limited at this stage, I do know about the lovely Smoke though!

Gday Eric,

After reading your post I have attempted to draw the schematic of what I was thinking so far, obviously I need some help as I would like accuracy, stabillity & reliabillity with this circuit.

ASC Attached, I have no models for the power supply voltage regulators or the INA125Chip etc.

The Load cell is the BSS 2K 2000lb unit, for the purpose I want to try it out for it is a lot bigger than needed & I will only ever use up to 300-400lb for this application

Am I on the right track with the Gain figures I have attempted below, I have chosen the 5V ref, pin 15.

Load Cell output 3.05mV/V * 5V = 15.25mV maximum output.
Load cell rating 2000lb ( .01525V / 2000lb ) = .000007625V/lb

Output Required = 0.01V/lb

Gain Required = 0.01V / 0.000007625 = 1,311.5

RG = 60004 / 1311.5 = 45.752R

On the Datasheet for the INA125 it has Vo = (V+in - V-in) * G, how to find V+in & V-in without building the circuit?

Could you suggest some improvements for the circuit & protection etc.

The circuit will feed into an existing circuit board with a digital readout & it has a maximum input of 5V & .01V = 1lb on the readout.

Thanks

 

[ericgibbs]

hi willeng,

Assume the maximum load will be 400lb, so for a 2000lb LC , thats 400/2000 ie: 0.2 of it max rating.

Given that its 3.05mV/V, so with 5V excitation = 15.25mV * 0.2 = 3.05mV [ at 400lb load]

You want a 5V output from the INA125 for a 3.05mV input, so 5/0.00305 = Gain of 1639 overall.

Using your calculations for a normalised output of 10mV/lb to suit the existing digital readout.

The formula in the d/s is Gain = 4 + [60k/Rg], which transposes to Rg = 60k/[Gain-4]

So Rg = 60000/ [1311.5 - 4] = 60000/1307.5 = 45.89R [ close enough]
I would use a 39R and a 10Rpot, this will enable you to fine tune the Gain.

As you are using +/-9V supplies, the worst case limits for the Vout of the INA125 are +9v - 1.7v = +7.3v
and the -9v + 1v = -8v.

So a maximum available output swing of +7.3v thru -8v.

A point you should consider is the weight of the Base that is used with the LC. [ the structure which rests on the LC , on which the load is placed]
Its weight/mass will give a positive output even when there is no load on the Base.

You could offset the Base reading by connecting a -V variable offset voltage to the INA125 Vref pin #5.
If you do use an offset voltage, use a buffer OPA to drive the Vref pin.

E.

 

[willeng]

Thanks Eric,

I see I had erros in my calculations--damn!

So the -V offset voltage you mention is to Zero things, do I have that correct?

Just so I have it right:
You mentioned to use a -V variable offset voltage to the INA125 Vref pin #5.

(Pin 5 is IAref) & is connected to ground in my schematic, (Pin 15 is Vref5) which is the 5 volt reference to the load cell, I'm not sure which one to use & don't want to make any mistakes?

Do I require a buffer on the output between the INA125 output & the Circuit board it feeds?

One other query, the existing circuit board with the display requires the output signal from the INA125 Circuit & a Ground so, do I just run a ground wire from the INA125 Load Cell circuit board that I have to make & connect it to the existing display board ground connection.

I don't want to create a ground loop or any noise etc & am a bit unsure of this?

Thanks very much for your time.

 

[ericgibbs]

Thanks Eric,

I see I had erros in my calculations--damn!

So the -V offset voltage you mention is to Zero things, do I have that correct?

Just so I have it right:
You mentioned to use a -V variable offset voltage to the INA125 Vref pin #5.
If you do not need to set the INA125 output to zero output when the load cell is used with a Base load keep pin 5 connect to 0V as shown on your circuit.
Its possible your existing display board has a Zero set pot.

(Pin 5 is IAref) & is connected to ground in my schematic, (Pin 15 is Vref5) which is the 5 volt reference to the load cell, I'm not sure which one to use & don't want to make any mistakes?

Do I require a buffer on the output between the INA125 output & the Circuit board it feeds?
It really depends upon the input impedance of the existing display PCB. I would expect any decent display would have an input impedance exceeding 10K, so a buffer will not be required if that is the case.


One other query, the existing circuit board with the display requires the output signal from the INA125 Circuit & a Ground so, do I just run a ground wire from the INA125 Load Cell circuit board that I have to make & connect it to the existing display board ground connection.
Is there a grounding point connection on the LC .??

I don't want to create a ground loop or any noise etc & am a bit unsure of this?
Use a Star grounding point for all your supplies and LC cable screen

Thanks very much for your time.

hi,
I dont have a LTSpice model for the INA125, but I could use a 'similar' IA if required.

E.

 

[willeng]

Thanks again Eric,
I'll make up the circuit & see how I go & what if anything it may need.

I just noticed the note at the bottom, is it ok to use a similar IA model, not that I have one.

I'll be back!

Thanks.

 

[ericgibbs]

hi,
I have made a 'model' for the INA125, shown in this asc file.

You may find the circuit useful for testing.

In the plot I have made V5 a PWL ramp from 0mV thru 3.05mV... change it as you require,

E.

 

[ericgibbs]

hi,
I have made a INA125m hierarchical model using the LT1128 OPA.
Its not identical to the INA125, but it should suffice for testing

The files are in this zip.
Move the INA125m.asy to the LTS /Lib/Sym folder/OPA
Keep the INA125m.asc in the same folder as the INA125m_test1 & test2 asc files.

Use the INA125m symbol as normal F2/OPA and select INA125m.


E.

 

[willeng]

Gday Eric,

Thanks for the LTSpice Models Etc, very helpful, I'll try to load them in, see how I go.

Although probably not a good Idea, I made up the circuit first on a breadboard to get it all working.

I have a couple of issues, firstly is noise at the output of the INA125, it's about 4mV peak, I'm not really sure how bad that actually is but it can't be good I guess.
As I understand, the breadboard won't help the issue & hopefully a circuit board will help.

When I switch the circuit on I get a reading of -172mV at the INA125 Output, with a 10k load resistor to ground at the output & without it actually.

I was expecting to be a little closer to Zero than this but that is an uneducated guess only.

I weighed a lb out & suspended it from the load cell & it seems to give the 10mV change as required although it is not Zeroed as mentioned.

Even though this is on the breadboard at this stage I would like to get it Zeroed to understand it more, not much good making something when you don't have some idea how it works.

To get it Zeroed I need the Offset Trimming you mentioned, I think?.
I can see figure 2 on the datasheet is for Offset Trimming.

If this is what I require I don't understand the schematic, in particular the V+ & V- inputs on the drawing - Attached.

They show what I believe are current sources on both these inputs & I don't understand?
Also they have
100μA
1/2 REF200.
Not sure about that either.

I have some OP37E Opamps, can I use these instead of the OPA237 shown, or is their a more common opamp I can use that I may have, I don't have a big range of things but may have something suitable.

I may not require this Offset Voltage when it's all connected to the existing digital readout but I would like to give it a go just now.

Thanks

 

[ericgibbs]

hi,
The circuit regarding the offset adjust is only for a +/-10mV range, it would not cancel out -172mV at the output. The +/-V on that drawing are your +/-9V's

The d/s for the LC states a Zero balance of +/-1% of the rated output of the LC.
The rated output for the 2000lb is ~15mV so that means, +/-150uVolts, which after amplification by ~1300 gives a possible maximum offset of +/- 195mV.

So the -175mV you see on the output is within specification, it indicates a -175mV/1300 = -134uVolts from the LC cell. At 10mV/lb a Base weighing say 17.5lb would move the INA125 output upto close to zero.???
To cancel LC offset I use the method of 'injecting' a very low level DC voltage into the bridge outputs.

Are the values you have posted with the LC in isolation, ie: no Base structure loading the LC.??

The noise you see is most likely due to using a BB for the construction,

Scoping the output, is it mains related frequency noise.???

I would check the existing display PCB for a Zero Display adjustment, to null out that -175mV.
E

 

[willeng]

Thanks Eric,

I am learning as I go so I'm a bit slow on posting answers.

I see now that the -175mV is within specifications after your explanation.

The figures quoted were with just the load cell suspended by the screw in attaching ring at the top & nothing in the bottom thread where another attaching ring screws in.
I made a typing error when posting the figures it was supposed to be -165mV, my appologise for that.

Being new to all this I stumbled at the next question for a while, regarding is the noise at mains frequency.
Yes, on the scope you can see it is mains related, 20mS cycle, 50Hz.

Yes, I will try to zero the existing display.

I would still like to learn how to zero this seperately though.
You mention:
"To cancel LC offset I use the method of 'injecting' a very low level DC voltage into the bridge outputs."

I can't seem to get that to work in LTSpice, I can get the output to zero by doing it but when you increase the output from the LC the output from the INA125 is wrong, I am not doing it correctly obviously?

Thanks

 

[willeng]

Sorry for the double post but I forgot to ask about the Grounding scheme for this circuit.

I have all the grounds going to a single point that the 0V connection from the power supply goes to.
I have pin 12, pin 5 of the INA125, the (LC Ground, Black wire) & 0V connection from the power supply connected together at one point.

Is this ok, thought I may have it wrong & it is creating some noise?

Thanks

 

[ericgibbs]

hi,
This is an example of offsetting, note I have deliberately unbalanced the bridge in order to introduce the offset error. [ the R1,R2 ,R3 are for offset nulling]

The plot shows the AC response, added a LPF after the INA125 as a trial only.

Note the Res/Caps near the INA125 inputs, these could be required if you are subject to RF interference.

 

[willeng]

Gday Eric,

Thanks once again, I have been studying LPF'S etc.

If I have it correct on the last asc posted, INA125-test3:

Before the non inverting input +in of the INA125, the 4.7k resistor & the 1n capacitor form a low pass filter & is one of the 400Hz -3dB filters, when I select Node N002 on the sim I can see this.

What I don't understand is that If I use the Formula for the cut off frequency:
Fc = 1 / 2pi*R*C I don't get the 400Hz cut off frequency from the formula?
I'm not sure if I am doing this right or there are other things to consider in the formula for this circuit.

What is the C1 47n cap for, does this equalize both inputs for the INA125?

I see the LPF on the output as mentioned & the Unity Gain Buffer.

I didn't have the Potentiometer model for the sim but found one on the LTSpice users group pages.

I changed the sim from the AC analysis to the transient mode to see if I could change the wiper value of the potentiometer model to zero the output.

I don't get any change on the output by changing the wiper value between 0 to 1, basically nothing happens?
Not sure what I have to do to get it to work, all I get is approx -99.8mV output?

I have started to draw the circuit board required for this so i'll see how I go, is their a circuit board design program that does this?

Thanks for all the help with this Eric, it is much appreciated!

 

[ericgibbs]

hi,
Look at this sim for the Pot.

Remove V5 voltage source when doing this test.

What is the C1 47n cap for, does this equalize both inputs for the INA125?

This is effectively a low impedance at high frequencies for input signals.

 

[willeng]

Gday Eric,

Thanks for the information, all this has been very helpful.

I will make up the circuit board & test it, i'll post the results when done.

It's been a very interesting project so far!

Thank You!