Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Load Cell junction box and offset trimming

Status
Not open for further replies.

htekin42

Member level 2
Joined
Dec 10, 2012
Messages
45
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,699
Dear All,

I am having problem little bit about platform load weighing system.

I have 2 design. One of them will have 2 load cell and other one will have 4 load cell. I am using INA129P as amplifier.

I know that I need junction box to get accurate weighing for load cells. but when I make search on google, I did not find any helpful information about that. some of them says, connect them in series, but most of the products has multiturn potentiometer on connections. Is there any reference circuit or information to make a junction box reference circuit?

Also, load cell has drifts and they can be adjust with offset trimming on op-amp. Which is the better, offset adjust on Load cell signal side with pot, or offset adjust on op-amp reference input?

Thanks
 

Hi,

It´s not clear to me what you want to do and how your (desired) connections are.

--> a picture says more than thousand words.

Klaus
 

Hi,

Here is the pictures. first one explains the junction box (summing box). Other one explains the offset adjustment.

For the junction box, I need a reference circuit or how to make design information.
juction_box.jpg
Second one, I need suggestion for offset adjustment.
offset_adjust.jpg

Thanks
 

Attachments

  • offset_adjust.jpg
    offset_adjust.jpg
    192.9 KB · Views: 483

Hi,

There are different solutions. Ech has it´s benefits and drawbacks.

Junction box:
You didn´t draw the sensor excitation signals. Is it DC or AC?
DC is more simple, but with AC you automatically cancel out all thermocouple voltage errors.

Use low pass filters to reduce noise and increase ESD immunity.

For high signal quality (or long cabling) I recommend to put the amplifier as cloase as possible to the sensor. Amplify the signal. Then either do the ADConversion at the sensor or use differential analog signal wiring to the junction box.
It seems you want to add all four signals. Instead of one suming amplifier consider to use an ADC with 4 differential inputs. This makes signal processing more flexible. For example: if the sensitivity of all four sensors is not exactely the same then you can compensate for this.

Transmitting signals digitally makes it more immune against errors than transmitting analog signals.

Summing amplifier: (If you don´t convert 4 x digital signals)
I´d use (for high signal quality) 4 x the INA, and then use a differential OPAMP circuit with individual input resistors for each channel.

Signal processing with ADC and microcontroller:
Select a fixed, stable, continous sampling frequency that is an integer mutiple of lines frequency. Do digital averaging over one line full wave period (or multiple).
This cancels out almost all mains introduced errors.

If you expect high thermal drift and you want to compensate for it, then consider to install a temperature sensor.

Offset correction:
* the best quality is that you buy a sensor with already corrected offsets. --> if it is done by laseer trimming directely at the load cell it will result in very low future temperature drifts.
* I personally don´t like pots, because they will drift with time, temperature and mechanical vibration
* But if you use a pot, then place the pot where you expect abput constant temperature and/or no mechanical vibrtion. I see no big difference in your two modes.

My question is: Do you need analog offset correction at all? If the analog value is converted to a digital one and this data processed in a microcontroller, then do the offset cancellation in software.

For sure there are alot of factors that need to be considered.
* cost
* expected resolution, prcision and accuracy
* is there a way for automatic calibration
* cable length and cable type
* software effort
* environmental parameters: temperature, humidity, vibration...
* and so on
We can´t know about this, therefore you need to decide it on your own.

Klaus
 
Dear KlausST,

Thanks for your advices.

I will going to supply load cells with DC voltage.

I am using low pass filter on output side of INA129P.

I need the total weighing of load cells. So I need a junction box to get total weighing on platform.

I will going to use minimum cable length from load cell to junction box.

I am trying to do something like this.

Connect 4 load cell to junction box, and get only one output. and amplify it then read it with 16 bit or 24bit ADC module.
Here is an example of junction box which I am trying to do.
junction_box_example.JPG

We have a supplier which manufacture C3 class load cell and they do not make laser trimming. All the load cell we bought from them has offset.

we also make software offset trimming, but some load cell has much offset and we want them to trim and then we can re-trim on every usage with software. Is it better or making trim of all on software?

Thanks for your replies.
 

Hi,

All your decisions are OK.

I need the total weighing of load cells. So I need a junction box to get total weighing on platform.
As said before: If there is the chance that there is a sensitivity mismatch between sensors: How do you decide to comensate it?

My recommendation:
4 x load cell --> 4 x INA --> 1 x summing amplifier --> ?? (I recommend either analog differential, or digital) ADC on this bord or remote?

Offset trimming:
OK. If the offset is small, then do it with software only. But with some sensors the offset is more than the estimated masurement range, then you need to do it with the INA offset input. Maybe you can use an EEPOT (quad? with pusbutton interface or I2C/SPI? I recommend to use a comparator with "zero" threshold and a LED to ease offset trimming. Does the end-user need to perform the offset trimming? )

Klaus
 
Hi,

As your recommendation summing amplifier has inverted output. How can I normalize it? What is your suggestion.

Also when I sum all the outputs of sensors. Is it true weight. I am not expert on load cell.

Thanks a lot for your helps.
 

As shown in post #5, the usual way to sum/average the signals of multiple load cells is to connect it in parallel. No summing amplifier used.

Equal sensitivity and output impedance of load cells is required.
 

Hi,

Averaging by paralleling the sensors will work...if the sensors match.

Suming with amplifiers ... is no problem, either. Just interchange the INA inputs to correct for inverting suming amplifier.

Klaus
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top