Load cell Data(Force sensor) to 2 systems

Thread Starter

or789

Joined Oct 15, 2019
19
Hi Everyone,

Curently i have a Tension Compression Load cells(attached datasheet) connected to a Electronic control system which Supply's voltage and calculate the force. Right now i need to send the raw data of the load cell to another complete diferent application.

Is there any interface to send the analog data(raw values) of load cell data to current PCB control system and to another application - Time synchronized by interface plug of high quality like stereo jack plug or DB9.

Thanks in advance
 

Attachments

Thread Starter

or789

Joined Oct 15, 2019
19
hi,
Do you mean the actual bridge differential output signal.?
Over what distance.?
E
Hi Eric,

yes, I meant Differential output signal (+ and - output voltage)

This signal is send to Control system(1 meter distance) and another Device which is 3 meters far.
 

ericgibbs

Joined Jan 29, 2010
9,574
hi,
More questions.
What are the two remote devices, ie: are they designed for a differential input.?
Do they also share a common ground with the L/C bridge.?
E
 

Thread Starter

or789

Joined Oct 15, 2019
19
hi,
More questions.
What are the two remote devices, ie: are they designed for a differential input.?
Do they also share a common ground with the L/C bridge.?
E
Hi,

First device - Electronic control system is designed for differential input by using AD7796. already working perfectly.
right now i need to send the data to other EMG(electromyography) - device. My part is to only send raw data of the load cell to this device.

So i am looking for interface which send load cell - differential output raw data to both this devices which is synchronized.

No they do not share common ground.
 

MrSoftware

Joined Oct 29, 2013
1,676
Are you not linking the grounds for a specific reason? i.e. for safety, or because the two systems will be operating at different potentials?

If the grounds cannot be connected, then you cannot share the raw load cell output, you're going to need to add something. You can convert the load cell signal to differential, but with load cell output as tiny as it is you may do better to amplify the signal first, then run the amplified signal through a differential driver. Can you explain more about your system and the reasons for your requirements? That might helps us come up with better ideas.
 

Thread Starter

or789

Joined Oct 15, 2019
19
Are you not linking the grounds for a specific reason? i.e. for safety, or because the two systems will be operating at different potentials?

If the grounds cannot be connected, then you cannot share the raw load cell output, you're going to need to add something. You can convert the load cell signal to differential, but with load cell output as tiny as it is you may do better to amplify the signal first, then run the amplified signal through a differential driver. Can you explain more about your system and the reasons for your requirements? That might helps us come up with better ideas.
Hi, our product is muscular traning device which consists of force sensor. Right now we have to give the raw values of the force sensor to another device - EMG.

In the attached image of EMG device, it needs load cell raw data to display as one of function as show in the figure.
 

Attachments

Thread Starter

or789

Joined Oct 15, 2019
19
I can connect the ground wire of both the devices and loadcell. 5V power is supplied to loadcell from ECU.
 

Reloadron

Joined Jan 15, 2015
5,140
The output (sensitivity) of your load cell is 2.0 mV / volt which is pretty common. The recommended excitation voltage is 10 Volts with a maximum of 15 volts. I saw where you are using 5.0 volts which is fine but will result in a lesser full scale output from the cell. What that all means is that your load cell has a full scale value which could be 50, 100, 150, 200, 300, 500, 750 kg. When at full scale in tension or compression the load cell output voltage will be 0.002 volt per volt of excitation. If you have 5.0 volt excitation that means at full scale the Vout will be 0.002 * 5 = 10 mV. Using that you can convert to engineering units of force.

You don't mention if the load cell is being used in a tension or compression mode, or both? That is sort of important. The output of the load cell should run to a differential input, meaning the load cell output can be positive or negative with respect to its common.

Something else to consider is your excitation voltage should be stable, keep in mind your sensitivity is 2.0 mV / Volt so you need a stable excitation.

Ron
 
Last edited:

ericgibbs

Joined Jan 29, 2010
9,574
hi,
I see that the L/C has sensor wires for the Vexcitation.
If the 5V Vext cannot be guaranteed to be 'stable' there are circuit options which measure the Vext at the L/C Bridge.
By measuring the bridge Sense voltage, with say an MCU, it is possible to make the L/C amplified output ratiometric with reference to the Vsense wires.
Which will compensate for changes in Vext.
E
AAA 135 08.11.gif
 

Thread Starter

or789

Joined Oct 15, 2019
19
hi,
I see that the L/C has sensor wires for the Vexcitation.
If the 5V Vext cannot be guaranteed to be 'stable' there are circuit options which measure the Vext at the L/C Bridge.
By measuring the bridge Sense voltage, with say an MCU, it is possible to make the L/C amplified output ratiometric with reference to the Vsense wires.
Which will compensate for changes in Vext.
E
View attachment 197476
Hi Eric,

Sorry for asking late. Could you let me know how can i connect to 2 devices from Load cell with common ground ? I can only give 5v Vext to loadcell. Also have to provide galvonic isolation.

Thanks in advance
 

ericgibbs

Joined Jan 29, 2010
9,574
hi 789,
There are linear Opto isolators available, ref attached PDF.
Are you allowed to amplify the data prior to isolating and attenuation.
E
 

Attachments

No connected grounds means isolation is needed. Isolation in the analog domain is difficult to do well, even the "high linearity" optocoupler has lousy gain matching, not to mention gain shift with temperature. Isolation in the digital domain is much easier. Also, any load cell is likely measured ratiometrically, so its output for any fixed force is a product of the force and the reference. It connot be considered to be "X volts per unit of force" but is "X volts per unit of force for a supply voltage of Y". This complicates sharing a load cell signal, as two different references are in place. If I had to I'd consider building a circuit that "spies" on the AD7796 output and, via some microcontroller, converts that to a simple PWM signal driving a fast (10Mb/s) optocoupler. Use the isolated optocoupler's PWM output to convert the second device's Vref into a (small) voltage that appears like a load cell. All in all it would be easier, cheaper and faster to just stack 2 load cells, one for each device.
 
Top