In this emission noise evaluation, we used a CISPR 11 radiation noise measurement environment as a reference to measure the radiation noise of a commercially available articulated robot set.
Furthermore, the robot operated any single joint (repeated left-right motion).
We evaluated the radiation noise.
Two noise tolerances are set in CISPR 11.
Here, Class A covers equipment that is not directly connected to a low-voltage power grid (interpreted as a commercial AC power supply network). For example, the case where AC power is supplied to equipment through a cubicle or other high-voltage receiving equipment. By contrast, Class B covers equipment that is directly connected to a low-voltage power grid (interpreted as a commercial AC power supply network). This would be the case where AC power is supplied in a shared pole-mounted transformer or other public distribution network.
Articulated robots are thought to be used mainly in factories. However, the radiation noise of the commercial robot prepared for this evaluation ended up exceeding the CISPR 11 Class A tolerances, and we found that it did not satisfy CISPR 11, which will become the basis of the standard applied in the future.
Because there are currently no noise regulations applied to robots, this is not a problem under these conditions. However, it is thought that new forms of noise suppression will be needed when the future noise regulations are applied.
We investigated the radiation noise mechanism using the following methods.
(An example of the investigation that was carried out)
As a result of investigating the radiation noise generation mechanism of the purchased articulated robot, we learned the following.
The noise source was the switching noise of the DC-DC converter module. The switching frequency of the DC-DC converter is 470 kHz.
There are three noise-conducting paths. The first path conducted switching noise to the negative line of the DC-DC converter output. The negative line was connected to the controller housing, and the switching noise was conducted to the housing.
The second path conducted noise to the housing due to the fact that the crystal heat dissipation plate of the DC-DC converter (heat dissipation plate that the switching noise is coupled to) touches the housing.
The third path conducted noise to each cable connected to the control board (AC power supply, motor, encoder, and teaching pendant cables) with superimposed switching noise.
The antennas leading to external radiation were the controller housing and each cable.
Suppressing the radiation noise of the robot requires that the noise not be conducted to the controller housing and each cable. As shown in the figure below, it is believed that shielding the DC-DC converter module and inserting a filter in the noise-conducting path is an effective approach.
However, because it is difficult to shield the DC-DC converter due to its construction (unable to completely remove the noise that is coupled to the crystal heat dissipation plate with a retrofitted shield, and it inhibits the heat dissipation, which causes the DC-DC converter to malfunction due to heat), even if measures are carried out for (a) and (c), noise is conducted along path (b), so we were unable to verify the noise suppression effect.
To carry out a measure for (b), it is believed that the wiring pattern and the DC-DC converter module selection must be reviewed along with reconsideration of the method of heat dissipation, and it was determined that there are limits to retrofitted noise suppression.
Because we could not sufficiently shield the radiation from the DC-DC converter module in this evaluation, we were unable to verify the effect of inserting a filter. However, the key points for selecting a filter are shown in (1) through (3) below.
Because robot radiation noise is prone to becoming a problem in many cases between 30 MHz and 500 MHz, an effective approach is to propose a noise filter that focuses on the frequency band.