What is an Overload?
An overload occurs when the voltage of an incoming signal is higher than the voltage range the sensor or data acquisition system can handle. When an overload occurs, the data beyond the range limit is clipped (Figure 1).
Overloads distort the amplitude and frequency content of the data.
There are three main types of overloads:
1. Data Acquisition System Overloads
This type of overload occurs when the range of the data acquisition system is lower than the range of the incoming signal. The incoming signal is clipped as shown in Figure 1 above.
Usually when this occurs, a red light appears on the acquisition system and an overload flag is set.
2. Transducer Overloads
This type of overload occurs when the physical input the sensor measures is greater than what the sensor can handle. This causes the incoming data to be clipped before it is digitized. This will happen even if the range of the data acquisition system is greater than the range of incoming data. See Figure 2 below.
To fix this, try using a less sensitive transducer. For example, instead of using a 100mV/g accelerometer, switch to a 10mV/g accelerometer. For the same amount of vibration, less voltage will be generated.
Be sure to check transducer data sheet and find the max level the transducer is rated for. For example, an accelerometer may have a max level of 500g. If the data acquisition system is reading 500g the transducer may be overloading.
3. Out-of-Band Overloads
An out-of-band overload occurs when frequency content beyond the specified bandwidth is of higher amplitude than the range limit of the system. This can be a very difficult overload to diagnose because the data displayed in the bandwidth region will be within range, yet the system will indicate an overload.
To fix an out-of-band overload:
What Does an Overload do to the Data?
Scenario: A transient signal was acquired. The range of the signal is 1V. The signal was acquired once with a 10V range and once with a 0.16V range. The signal that was acquired with a 0.16V range is clipped in the time domain (see Figure 4a below).
An overload can affect the data across the entire measurement bandwidth. It can raise or lower amplitudes and generate additional noise on the measurement.
Why is there so much spectral content introduced when the signal overloads?
In the time domain, the signal becomes clipped. Essentially turning the “sine” shaped signal into a “square wave” shaped signal.
The Fourier series of a square wave consists of all odd harmonics of the frequency of the wave, continuing to infinity. Because the clipping creates square waves at various periods and frequencies, the fundamental odd harmonics of these square waves are also created at many frequencies. The odd harmonics of these fundamentals are carried out across the entire frequency range creating lots of spectral content.
This phenomenon is called harmonic distortion (see Figure 5 below).
You can see that even with just three or four square waves at different frequencies, harmonic distortion effects the spectrum greatly.
Overloads greatly affect the resultant spectrum!
WARNING: Even if the overload is out of range you will still get harmonic distortion across your entire frequency range. The overload will still effect your data even if it’s out of band!
LMS Test.Lab Tips!
In the LMS Test.Lab Channel Setup workbook, under the “Range” column, the dynamic range of the channel can be set. Try maximizing this range. Most SCADAS systems have a max range of 10V.
In LMS Test.Lab under the Acquisition Setup tab there is a level bar which gives an indication of whether or not the data acquisition system is in overload. A white bar indicates the range is too large. Green bars indicate the range is set appropriately. Orange indicates that the signal is within the upper limit / overhead region. A red bar indicates the range is set too low and the channel is overloading.
To auto-set the ranges, click 1) start ranging, 2) hold level, then 3) set ranges.
This will set the range to the appropriate level. Hopefully minimizing the risk of an overload.
Under the “More…” button in Figure 9, there are some additional helpful options.
It is possible to set the dB level of overhead used for the ranging.
Overhead is essentially “headroom” for safety. It is additional room in the range to protect against spikes in the data which may cause overloads. A common overhead setting is 6 dB (or 50% of the input range). You can always increase or decrease this number to suit your needs.
It is also recommended to activate the “Use full range when autoranging” box. When activated, the range is set to the full range before the autoranging procedure starts.
NOTE: Use the autorange feature to set the ranges appropriately. It is not wise to always use the max range of the system to avoid overloading. See the explanation of why in this article.
Monitor the incoming data in the Measure workbook by dropping in data from the Data Explorer. It is important to monitor incoming data to ensure the data is as expected. This will make it easier to detect what type of overload occurs.
4. Overload Logging
In the Measure worksheet, after stopping a run, the “accept or reject run” dialog box will appear (unless the message is suppressed). If an overload occurs, the dialog box will warn the user that an overload occurred and an overload report will be generated.
If no overload occurred, the normal dialog box will appear (as shown below).
5. Overload Report
To get the details of an overload report, browse to the run in the Navigator workbook. Two reports will be listed: the “OfflineOverloadLogging” and “OnlineOverloadLogging”. To view the contents of the report, right click the report and click view.
This will open a “File view…” window. Double click on the .txt file within the window to open the detailed overload report. The report contains which channels overloaded and at what points in time.
Related Digital Data Acquisition Links:
LMS Test.Lab acquisition tips: