instruNet

¹ Heating Considerations with Resistors
When current is pumped through a resistor, it heats up. When a resistor heats up, its resistance changes, and this can cause errors in your measurement. The current though a resistor is calculated via:

Current (Amps) = Volts Across Resistor / Resistance in ohms

The power dissipated by a resistor is:

Power Dissipated (Watts) = Volts * Volts / Resistance
= Current * Current * Resistance

The amount a resistor heats up is:

Change in Temperature (Celsius) = Thermal Resistance (C/Watt)
* Power Dissipated (Watts)

The amount a resistor changes for a change in temperature is:

Change in Resistance (ohms) = Change in Temperature (Celsius)
* Temperature Coefficient (ppm/C)
* Resistor Value (ohms)

For example, a 100ohm resistor with a 100 ppm/C temperature coefficient and 30C/Watt thermal resistance that is passing 50 milliAmps would enjoy the following situation:

5Volts across resistor = 100ohms * 50mA
0.25Watts power dissipation = 5V * 5V / 100ohms
7.5°C temperature change = 30C/Watt * .25Watts
0.075 ohms change due to temperature change
= 7.5°C temp change * .0001ohms/ohm/C thermal resistance * 100 ohms

² Excitation Voltages for Bridge Circuits
If you type an unreasonably high value into the Vout field of the Constants area and press Enter, instruNet will set the output voltage to the highest possible value without allowing the internal output amplifier to saturate (e.g. ≤ 4mA for #iNet-100/100B and ≤15mA for #iNet-100HC). Setting the highest possible Vout, causes the highest possible voltage to be read by the Vin terminals, which increases the signal to noise ratio and therefore increases accuracy. The downside to having a high excitation voltage is that it increases the power dissipated by the resistors, which increases their thermal heating, which increases the drift from their resistance's at ambient temperature (e.g. typical resistors offer 100ppm resistance drift per degree C change in temperature). Resistors with low temperature coefficients (e.g. 25ppm/C) are helpful if this is a problem. Also, if noise is a problem, it is sometimes helpful to install a capacitor (e.g. 1µF) between the Vout terminal and GND, at the bridge (possibly far from instruNet), to hold the excitation voltage steady.

³ Shunt Resistors and Bridge Completion Resistors
Bridge completion resistors and shunt resistors should be as accurate as possible (.1% is often ok, .01% is very good), and have a low temperature coefficient (25ppm/C is often ok, 5ppm/C is very good). If you use a less accurate resistor, we recommend that you measure it with a DVM, and then type this more accurate value into the Rshunt field in the Constants setting area. To determine the effect of a resistor inaccuracy, calculate your engineering units for a typical case, and then increase the resistor value by its expected error, and note the change in the resulting engineering units output. For example, if a 100ohm shunt resistor is used to measure a 1mA current source and it changes 1%, then the measured reading would change 1%, or .01mA. For a list of precision resistor suppliers details, click here.

⁴ Bridge Completion Resistors in Strain Gage Bridge Circuits
In a bridge, all 4 resistors must be the same value, within 10% or so (1% is better, .1% is excellent), in order for the bridge to operate properly. In some bridge circuits, all 4 resistors are supplied by the sensor manufacturer; whereas in others (e.g. 1/4 or 1/2 bridge circuits), the user must supply "completion" resistors of the same value as the gage to complete the bridge circuit. This can be done by installing precision resistors (e.g. 0.1%), or by installing fixed unstrained strain gages of the same ohmic value. 3

⁵ Filter Settings for Low Level Measurements
Strain gage, thermocouple, and RTD voltages are typically very low and therefore often require low pass filtering to reduce noise. Low pass filters cause high frequencies to be rejected, while low frequencies are passed. Visually, the signal becomes "smoother". instruNet provides several low pass filter options:

The Low Pass popup menu in the Hardware settings area can select a variety of analog low pass filter options (e.g. the Model 100 provides the following analog one pole low pass options: off, 40Hz, and 4KHz).
The Integrate field in the Hardware Settings area selects how long the signal is averaged before instruNet returns one number. This "averaging", in effect, is a low pass filter. Careful, this averaging fully consumes the instruNet controller, and therefore reduces the maximum possible sample rate, as noted in Sample Rate Vs Integration Vs. Noise. A 0.001 sec integration time is often very helpful at reducing noise and increasing accuracy.
The Low Pass settings area provides a means by which one can digitally filter a signal, post acquisition, with tremendous accuracy.
The user can manually place a capacitor across the Vin+ and Vin- input terminals with any bridge or voltage divider circuit to provide a 1pole low pass filter where the cut-off Frequency in Hertz is equal to 1 / (2 * π * R * C); where R is the source resistance in ohms, C the parallel capacitance in Farads, and π is 3.14159
For more details on reducing noise, click here.

⁶ Selecting a Voltage Divider Shunt Resistor
Shunt resistor values are typically chosen to cause a large voltage (several volts maximum) to be measured by instruNet, without heating up the resistor significantly to cause its resistance to change or causing the excitation voltage source to over shoot its maximum output current (e.g. 4mA on the #iNet-100/100B and 15mA on #iNet-100HC). If the Rshunt value is low, then the voltage across it is low, and this decreases the signal to noise ratio of the measured signal. Also, Rshunt must be selected such that the voltage across Runknown does not exceed the instruNet maximum input voltage (e.g. +/-5V with the Model 100). Due to these limitations, instruNet might not let you set some of the fields too high or too low.

⁷ Voltage Range Settings For Strain Gages
Since strain gage voltages are often very small, a small input range (e.g. +/- 10mV) works best for most measurements. Increasing the voltage range increases the range of strain that can be read, while sacrificing accuracy with small voltages (e.g. instruNet can read 5mV more accurately with a +/-10mV range, than with a +/- 100mV range). Please refer to your equation for details on how strain relates to voltage measured.

⁸ Balancing your Bridge with the Vinit Correction Voltage
Vinit is the voltage measured across the intermediate nodes of the bridge when the strain gage(s) are unstrained in a bride circuit. This is measured by putting instruNet into Voltage mode, Differential Wiring, with a low voltage Range (e.g. +/-10mV), and then measuring the resulting bridge voltage (e.g. via the value shown at the bottom of the Probe dialog). You must then enter this voltage value into the Vinit field in the Constants settings area, reset your Sensor field to Strain Gage, and reset your Wiring field to where you had it. Subsequently, all reported "strain" values will reflect resistance changes from the "unstrained" scenario. Vinit is used as an offset correction factor to "balance" the bridge. If you do not want to go through the trouble of "balancing" your bridge, simple set Vinit to 0.

⁹ The Strain Gage "GF" Factor
All strain gages are manufactured with a specific Gage Factor (GF), which relates a change in resistance, to strain. The GF is often printed on the strain gage package, and must be correctly entered into the instruNet GF field within the Constants settings area. This is used to calculate the "strain" value returned by instruNet.

¹⁰ Accuracy of Measurements
Accuracy measurements are affected by the noise pickup on the leads, the accuracy of the sensor itself (i.e. thermocouple's are typically accurate to +/- 1C to 3C) and the instruNet measuring system. A noisy environment and long sensor leads are often the worst threat to accuracy. Integrating (via the Integrate field) a signal over a period of time will give a more accurate measurement by filtering out noise at the expense of a lower maximum sample rate.

An example of how to calculate accuracy is as follows:

Suppose you are doing a current measurement where the current is calculated as the voltage drop across a shunt resistor divided by the resistance in ohms of the resistor.

Current (Amps) = Volts across shunt resistor / shunt resistance in ohms

Suppose the measured voltage is accurate to 1mV and the 1K ohm shunt resistor is accurate to 1%. Subsequently, the accuracy of the measured current would be

Max Current Error = 1mV / (.01 * 1K) = 100microAmps

¹¹ Alternating Positive and Negative Excitation Voltages
To reduce the burden on one side of a power supply (e.g. +12V or -12V), excitation voltages often alternate positive and negative. For example, when powering 350ohm strain gages, the excitation voltages are typically set to {+5V, -5V, +5V, -5V...}. The alternating polarity evenly burdens the +/-12V supplies. Please note that in low current cases (e.g. <2mA), this is not necessary.