![](https://csdnimg.cn/release/download_crawler_static/1619539/bg14.jpg)
TWE F/T Transducer Installation and Operation Manual
Document #9620-05-TWE-03
Pinnacle Park
•
1031 Goodworth Drive
•
Apex, NC 27539
•
Tel: +1-919-772-0115
•
Fax: +1-919-772-8259
•
www.ati-ia.com
•
Email: info@ati-ia.com
- 18 -
4.2 Signal Conditioning Circuitry
Signal conditioning circuitry has several purposes:
• To minimize noise by amplifying the transducer signal. Signal conditioning circuitry should be
located near the transducer for best results.
• To provide high-impedance signals for a data acquisition system.
• To maximize the transducer’s resolution by optimizing the range of the output signals. This
usually means amplifying the output and applying any necessary offset so that the transducer’s
output range matches the input range of the data acquisition system’s analog-to-digital converter
(ADC).
4.2.1 Strain Gauge Amplification and Hardware Bias
The interface circuits to all six strain gauge signals should have excitation, amplification,
and hardware bias.
4.2.1.1 Excitation
We recommend an excitation of 0.000V and +5.000V using a low-noise, low-
drift voltage source. Any drift in the output voltage will cause a gain error in
your output equal to the ratio of the drift in volts divided by the nominal power
supply output voltage. You can compensate for this drift by measuring the
power supply voltage and using it along with the strain gauge readings rather
than using a constant for the power supply voltage.
Since the noise present in this source directly impacts the noise seen in your
system it is important to use a low-noise source. The amount of noise transferred
to your outputs will be approximately equal to the power supply noise divided
by the excitation voltage and then multiplied by the output range.
For example let’s use a +5V excitation power supply with 300µV noise and a
system output range of ±5V (i.e., a 10V range) that is fed into a 16-bit A/D
converter. The output noise directly attributable to the power supply is 300µV ×
5V ÷ 10V = 150µV. This is approximately equal to one bit of noise for the 16-
bit A/D converter. (One bit of noise for the converter is the voltage range
divided by the number of quantization points. For the 16-bit A/D with a 10V
range this is equal to 10V ÷ 2
16
= 152.6µV.)
4.2.1.2 Amplification.
The optimum amplification for each gauge is calculated by the TWE Calibration
Spreadsheet that is included on the system CD. These target gains will optimize
the transducer’s output range. It is not necessary to match these gains exactly.
However, setting the gain too high could cause early ADC saturation (decreased
range), and setting it too low could cause a decrease in resolution. Although the
ideal gains for the six channels are all different, they are typically similar,
enabling you to use the same type of amplifier circuit for each, if you choose.
4.2.1.3 Hardware Bias.
The nominal unloaded output of each of the half bridges is the voltage halfway
between SG
LO
(upper excitation voltage) and SG
HI
(lower excitation voltage) or
2
LOHI
nom
SGSG
SG
= .
All strain gauge output voltages will have some offset bias voltage present. This
offset bias voltage is the shift of the strain gauge output from ideal unloaded
output (SG
nom
). The hardware bias values for your transducer were measured
during calibration and can be found in the TWE Calibration Spreadsheet.
Removing this offset voltage prior to amplifying the signals will allow you to