Recommendations for strain gauge measurement with Beckhoff modules
- Electrical connection:
- Operation with an additional sense line for the bridge supply is recommended:
Full bridge in 6‑wire mode, half bridge in 5‑wire mode, quarter bridge in 3/4‑wire mode. - The use of full bridges instead of half or quarter bridges is generally recommended in order to achieve higher temperature stability and higher measuring accuracy.
- Selection of the feed voltage UExc:
- A feed voltage of 5 V has proven to be useful in many cases.
- In general, it should be as high as possible within the permissible range according to the data sheet in order to achieve a large modulation of UBridge with the given nominal characteristic value [mV/V] and thus to maximize the electrical measuring range of the module (SNR increment).
- However, it should be considered that the heating of the bridge in the load cell increases quadratically with UExc to a first approximation. With high feed voltages and insufficient heat dissipation from the sensor to the machine, this can lead to massive drift effects after switching on.
- If necessary, select a bridge with a higher nominal parameter [mV/V] or a higher internal resistance [Ω].
- Selection of the nominal load of the weighbridge:
- It should be selected somewhat larger but as close as possible to the target load so that the mechanical and thus the electrical measuring range is utilized to the fullest extent possible.
- The overload capacity of the load cell must be observed. Fast weighing procedures in particular can lead to excessive mechanical stresses; nevertheless, as stated above, the bridge should not be over‑dimensioned (regarding Emax).
- The mechanical natural frequency of the load cell (in which the measuring bridge/strain gauge is installed!) or of the complete setup is to be considered in relation to weighing procedures (number of product changes, product speed, product weights). If necessary, a significantly larger nominal load should be selected to the target load, because sensors with a higher nominal load have shorter nominal measuring paths and are thus mechanically more rigid. With a more rigid measuring body – usually the softest part of the entire setup – the natural frequency increases. As a result, the dynamics of the weighing procedure can be cleanly captured and measuring errors due to the natural oscillations of the weighing setup are avoided.
- Calibration/compensation of the bridge:
- Regular zero compensation (tare) is recommended.
- The tare effect should be observed in order to recognize a possibly damaged measuring bridge: the signal of a damaged measuring bridge drifts; it does not return to the original value after removal of the load.
- For the compensation of a gain error, a compensation point close to the target load should be selected during commissioning and if possible, during operation, especially if this lies well below the nominal load (measurements in the partial load range).
- Possible filtering of the measurement, dynamic effects:
- In the case of fast sequential weighing procedures (several objects to be measured, e.g. products per second), it may be possible with adapted digital filters ‑ despite an obviously “poor” measuring signal ‑ to achieve a high measuring accuracy.
- Overshoot effects can often be observed for example, the pickup device for measurement objects (products) actually always moves mechanically (even if only in the µm range).
- The procedure of fast sequential weighing can also be dependent on the speed of which measurement objects (products) being moved over the weighting area; the filters for the measurement signal may then need to be dynamically adjusted.
- The optimal signal analysis is supported by Beckhoff with various products: flexible filters in the EtherCAT modules, TwinCAT filter designer, TwinCAT filter library, TwinCAT Analytics and so on.