Strain Gage Shield output range questions

Help & Support Robot Parts
1

I’ve got a couple questions regarding the RB-Onl-38 Strain Gage Shield (SGS).

  1. It would appear that the excitation voltage is 3.3VDC rather than 5VDC. I found this when putting a multimeter across the outside pins (#1 and 4). In the default setup it does not appear that a full 10-bit output is possible. My application is a full bending bridge setup on a cantilever beam. The aluminum beam is only 1/8" thick and about 10" long so I can bend it to induce pretty high strain levels. The raw output from the example sketch (v1.10-16) of my beam under no load is about 384. Under a high load, displacing the beam about 1" at the end, is about 670. A similar bend in the opposite direction reads about 1. The gages are about 2" from the fixed end of the beam.

By using a suggestion on the SGS w/Load Cell blog page, I connected the 3.3VDC to the AREF pin as well as added the [font=Courier New]analogReference(EXTERNAL)[/font] command in the [font=Courier New]setup()[/font] section of my Arduino Uno sketch. This still has an excitation voltage of 3.3V measured across pins #1 and 4 of the SGS, but the output goes from 1 to 1017 (pretty close to the theoretical limit 0-1023) with a no load reading of 584. Given my application with + and - loads, I’m guessing I’d want to adjust the voltage offset to be about 511, correct? This would give me the largest range on either side of my midpoint to measure loads.

If my loads create strains which hit the limits 0 and 1023, it means that I would want to reduce the gain provided by the SGS, correct? The documentation for the SGS show which resistors on the board can be swapped out to another value to create a different gain. Can they be switched out with a non-surface mount resistor? Is there any way to have a gain which can be adjusted more easily?

  1. My second question, slightly opposite of my first: Is it possible to use 5VDC as the excitation voltage without damaging anything? To do this I’m assuming I’d connect the 5VDC pin to the AREF pin as well as using the [font=Courier New]analogReference(EXTERNAL)[/font] command. I’d like to do this to boost the output signal of a set of strain gages that has a low output with the loads I’m applying. In this lower output case I have a full axial bridge that has relatively low output and I’m hoping to get a better signal by using 5VDC excitation. Would an alternative to this be to again change the gain by changing out the gain resistor for the channel?
2

Hi,

]You are correct. The excitation voltage on the load cell strain 1 and 2 on the SG Shield is 3.3V. Therefore connecting the 3.3V DC to AREF pin on the Arduino and using the analogReference(EXTERNAL) will let you increase the load cells readings accuracy. We also recommend to add a decoupling capacitor for noise reduction between the AREF pin and GND (a typical value of the capacitor is 100nF). Correct again, adjusting the voltage offset to 511 (without load) should provide you with the full ADC range 0 to 1023 to measure load cells. As for the gain resistor, we don’t have an easy/straight method to swap the resistor to change the gain. It should ideally be a surface mount resistor, that’s why we don’t really recommend changing the resistor unless it’s mandatory to have accurate measure and unless you have the soldering skills to solder the new SMD resistor./:m]
]The 3.3V Arduino pin on the SG Shield is connected to Strain 1 and Strain 2. Therefore the excitation voltage of the load cells 3.3V cannot be changed. A workaround for this, is to bend the 3.3V pin of the SG Shield and with a jumper wire, connect the 5V pin from the Arduino to the 3.3V stacking header pin. This way, the load cells excitation voltage will be 5V rather than 3.3V. You won’t have to connect the 5V Arduino pin to the AREF and to add the analogReference(EXTERNAL) function in your sketch, because the Arduino ADC analog reference is 5V by default. /:m]

Regards,