Introduction
When designing weight measurement or force sensors using load cells, it is not uncommon to be troubled by factors such as offset error and temperature drift. In particular, when trying to measure minute signals stably over long periods of time, these errors can have a non-negligible impact on the measurement results. One known approach to addressing these issues is AC Excitation. While you may have seen the name in datasheets and technical documents, surprisingly few people have actually experienced how effective it is, or how different it is compared to DC Excitation.
In this article, we will use Analog Devices' high-resolution ΣΔ AD converter AD4170-4 and its evaluation board to compare the operation of DC Excitation and AC Excitation with an actual load cell connected, and observe the differences. We will confirm the effect of AC Excitation on load cell measurement not only through theory but also through actual measured waveforms and numerical changes.
For those of you with questions like "Does AC Excitation really work?" and "Is it worth using in my design?", we hope this article will be helpful in helping you make a decision.
What is AD4170-4?
The AD4170-4 is a high-resolution ΣΔ AD converter from Analog Devices, designed primarily for use in measuring minute signals from load cells and bridge-type sensors with high precision.
In load cell measurements,
- The signal level is very small
・Easily affected by offset and temperature drift
-Long-term stability is required
There are strict conditions imposed.
The AD4170-4 is a device that integrates an AD converter, analog front end, and excitation function to meet these requirements.
AD4170-4 Block Diagram
The advantage of being compatible with both AC and DC excitation
DC excitation is generally used for bridge-type sensors such as load cells. While this method has a simple configuration and is easy to use, it has the problem of DC component errors such as thermoelectric power, temperature drift, and 1/f noise being superimposed on the measurement results.
In contrast, with AC Excitation, measurements are performed while the polarity of the excitation voltage is inverted between positive and negative in a square wave pattern. When the excitation polarity is inverted, the sensor signal is inverted between positive and negative, but DC error components such as offset do not invert. Therefore, by averaging and calculating the positive and negative outputs, the DC error can be canceled out and only the true sensor signal can be extracted.
The detailed theory of AC Excitation will be explained in specialized explanatory articles, but it is sufficient to understand the concept of "error cancellation through positive/negative reversal and averaging" as shown in the diagram below.
Error cancellation by positive/negative inversion and averaging
reference: Transducer/Sensor Excitation and Measurement Techniques | Analog Devices
The AD4170-4 supports both DC and AC excitation, and a major feature is that the use of an evaluation board makes it easy to perform comparative evaluations in the same environment by switching only the excitation method.
Evaluation configuration and experimental method
In this test, we used the evaluation platform provided by Analog Devices to compare and evaluate the DC and AC excitation of the AD4170-4 under the same conditions. The evaluation system configuration is as follows:
EVAL-AD4170-4: AD4170-4 evaluation board (differential bridge input, built-in PGA, excitation settings supported)
・ SDP-K1: Evaluation board control/data collection interface board (USB connection to PC)
・ ACEsoftware: Intuitive GUI operation for measurement settings, excitation method, PGA gain, filter selection, data acquisition, and waveform visualization
- Load cell (we used model number LCB03K006M in this example.)
・PC
This environment allows switching and comparative evaluation of excitation methods without the need for complex external circuits.
Try connecting
First, connect the SDP-K1 to the EVAL-AD4170-4 via the 120-pin connector on the evaluation board. Next, connect the SDP-K1 to the PC with the included USB cable. Finally, connect the load cell to the EVAL-AD4170-4. This process requires no special tools other than a precision screwdriver.
The load cell wires and J2 terminal numbers correspond as follows:
-No.2: Power +
-No.4: Output +
-No.5: Output-
-No.6: Power supply-
-No.8: Shield
These sources were used as references when making the connections.
・Hardware Guide [Analog Devices Wiki]
In the Hardware Guide, the load cell power supply is split into two and connected to the J2 terminal, but since an equivalent circuit can be created by shorting LK6 and LK9 on the EVAL-AD4170-4, J2 terminals No. 3 and No. 7 are not used this time.
First, try measuring with DC Excitation
Set up the evaluation board for DC Excitation. Basically, follow the Hardware Guide, but note that LK8 must be shorted and connected to AVSS. Also, as mentioned in the previous section, LK6 and LK9 must also be shorted. Furthermore, leave LK7 open.
Now that the hardware configuration is complete, let's move on to configuring the ACE.
ACE Configuration (DC Excitation)
Once the connection and evaluation board settings are complete, the next step is to configure the software using ACE. This section explains the process by following the actual screen flow.
1. Start ACE and check the recognition of the evaluation board
When you start ACE, it will automatically recognize the connected hardware and display the AD4170 Board.
Once you have confirmed that the evaluation board is recognized correctly, select AD4170 Board and proceed to the settings screen.
2. AD4170 Board Configuration screen
Next, open the AD4170 Board tab to display a GUI that gives an overview of the entire evaluation board configuration.
Here you can visually check the power supply configuration, input system, reference configuration, etc.
At this stage, no special setting changes will be made, just select the AD4170 and proceed to the settings.
3. Channel Settings
Next, you can set the channels. Here, you can specify the combination of channels and analog inputs to be used.
First, select MUX.
Select the channel and analog input. In this example, select Channel 0 and set AIN+[0] to AIN5 and AIN-[0] to AIN6.
4. Reference Settings
Next, set the Reference Input. In this evaluation, the reference configuration on the evaluation board is used, and AVDD/AVSS is selected as the reference input for the AD converter.
First, select Reference MUX.
Next, select AVDD and AVSS.
5. Digital filter settings
Next, set the digital filter and data rate on the Filter setting screen.
First, select DIGITAL FILTER ENGINE.
This time, set it to Sinc Filter5+Avg and the Data Rate to 500. Please note that this Data Rate number represents the Filter Word.
For the relationship between Data Rate and Filter Word, see the AD4170-4 data sheet. Here, 500 means 1 ksps.
Reference: Relationship between Data Rate and Filter Word (excerpt from the AD4170-4 datasheet)
6. Memory Map Settings
Finally, use the Memory Map screen to ensure that any chopping or AC Excitation related settings are disabled.
First, select "Proceed to Memory Map".
Type "misc" in the search bar to search for related registers. Since it is set to Channel 0, select Misc[0] and set the bottom row Chop_lexc[0] to "No Chopping".
This completes the DC Excitation measurement setup.
Let's measure DC Excitation!
Now that the ACE setup is complete, we will perform measurements using DC Excitation.
Select "Proceed to Waveform Analysis"
After setting the Sample Count (we set it to 1000), press Run Once...
The results were as follows:
The average output value during DC excitation is approximately 170 μV.
At this point in this chapter, we will not evaluate whether this value is large or small, as this measurement is merely positioned as a reference measurement to understand behavior during DC excitation. Note that the measurement was performed in an environment where no load was applied to the load cell and where the effects of surrounding vibrations and disturbances were minimized. Therefore, this result shows the output when the load cell and measurement system are in a static state.
Finally, let's try measuring with AC Excitation
Next, we switch the excitation method to AC Excitation and perform measurements. The settings are slightly different from DC Excitation, so we changed them using the following procedure.
AC Excitation Settings
First, change the settings of the evaluation board. Basically, follow the Hardware Guide, but please note that LK8 must be shorted and connected to AVSS, just as when performing DC-Excitation. Also, short LK6 and LK9. Here, LK7 is set to A, just like the Hardware Guide.
Next, set up the ACE. Using the Memory Map screen, select "Proceed to Memory Map", type "misc" in the search bar, and search for related registers. Select Misc[0] and set the bottom row, Chop_lexc[0], to "Chop lexc AB CD".
This completes the AC Excitation settings.
It's time to finally experience the effects of AC Excitation!
Measurements were carried out using the same procedure as for DC Excitation, and the average output value for AC Excitation was approximately 58 μV. This is approximately one-third of the output obtained with DC Excitation (approximately 170 μV).
However, the load cell used in this evaluation originally had a small offset in an unloaded state, and the evaluation was carried out under conditions in which disturbances such as vibration and temperature fluctuations were minimized as much as possible in the measurement environment. Therefore, we believe that the measurement system was already sufficiently stable at the time of DC excitation, and as a result, the offset cancellation effect of AC excitation may not have been noticeable.
Therefore, in the next phase, we decided to intentionally generate an offset and compare the results in a situation where the difference between DC and AC excitation would be more clearly apparent.
Add an offset factor
We considered the following three methods for generating offsets:
①How to warm up the load cell
Although it is expected that an offset will occur due to a temperature gradient, it was determined that it would be difficult to see the difference because it was difficult to keep the load cell at a uniformly high temperature and the temperature characteristics were good.
②Method of adding an offset in the order of mV using a dry cell battery
Although it is possible, the wiring and configuration would be somewhat complicated, so we decided not to adopt it this time for simplicity's sake.
3) Method of disrupting bridge balance by external resistor (adopted)
We adopted this method because it has a simple configuration, high reproducibility, and the offset amount can be adjusted by changing the resistance value.
A 47 kΩ resistor was selected as the external resistor. Each resistor that makes up the load cell bridge is 350 Ω, and by connecting 47 kΩ in parallel to some of them, the equivalent resistance value is intended to change by approximately 1%. The external resistors were connected to No. 5 and No. 6 of the J2 connector as shown below.
Measurement results after adding offset
DC/AC excitation measurements were performed with the bridge unbalanced by an external resistor.
As a result, the following values were obtained:
At DC Excitation: Approx. 7 mV
At AC Excitation: Approx. 57 µV
With DC Excitation, the intentionally added offset appeared directly in the output, and an output on the order of mV was confirmed. On the other hand, when switched to AC Excitation, the output was significantly reduced to the order of µV, despite the same wiring and evaluation conditions. This is thought to be the result of the excitation polarity reversal and internal processing by AC Excitation canceling out the DC component caused by bridge imbalance.
Summary
In this article, we used the AD4170-4 evaluation board to compare the behavior of DC excitation and AC excitation in load cell measurement.
First, we confirmed that under conditions where the load cell offset is small and the measurement environment is stable, the difference between DC and AC excitation is limited, and that sufficiently stable measurements are possible at DC excitation. On the other hand, under conditions where the bridge balance was intentionally disrupted using an external resistor and an offset factor was added, an output on the order of mV (approximately 7 mV) appeared with DC excitation, whereas with AC excitation this was significantly reduced to the order of µV (approximately 57 µV).
These results show that AC Excitation is particularly effective under conditions where DC component errors such as offset and drift become apparent, but that sufficient performance can also be obtained with DC Excitation under favorable conditions.The AD4170-4 supports both DC and AC Excitation, so a major feature is that it allows the user to flexibly select and switch between excitation methods depending on the measurement target and usage environment.
We hope that this evaluation will be helpful in selecting a load cell measurement method and designing a high-precision measurement system.
Click here for manufacturer site/other related links
Inquiry
If you have any questions regarding this article, please contact us below.
Analog Devices Manufacturer Information Top
Analog Devices Manufacturer Information If you would like to return to the top page, please click below.