By giving thought to where filtering is best applied, it is possible to avoid much of the engineering effort associated with certification, as well as keeping BOM costs down and avoiding interaction as well as current leakage concerns.
Even though, as discrete units, power sources will normally be able to comply with relevant regulations concerning electromagnetic interference (EMI), it is within the context of a complete system that compliance must be validated. If a bought-in AC-DC power supply unit (PSU) needs to be integrated into a system design, it will mean that the engineering team involved has to deal with the EMI aspect themselves, leading to dealing with numerous challenges. This article will look at what needs to be addressed and how the workload can be minimized.
Depending on the market that the end product is intended for, there are certain EMI standards that engineers need to be aware of. In the United States, there is FCC Part 15 Class A and B, while globally the IEC’s CISPR 32 Class A and B standards will apply. Although it should be appreciated that these standards are not completely identical, they are still fairly closely aligned. Therefore, provided that one of these is met with a reasonable degree of tolerance, it will be able to comply with the other one as well.
For both of these standards, Class A relates to the equipment being used in commercial, industrial, or business settings. Conversely, Class B deals with residential usage. The latter has greater stringency because the equipment is assumed to be operating in a less controlled environment, with greater potential for cross-interference occurring.
At the point of production, PSU manufacturers will typically ensure that their units are compliant with one of the EMC standards just described. This is when they are tested in isolation though, not in conjunction with the system in which they will eventually feature. Since statutory electromagnetic compliance (EMC) certification applies to the complete system, the noise generated by the PSU must be added to that emanating from the data lines, clocks, and suchlike. The total then needs to be below the figure that is deemed acceptable by the standard.
When there is only a single PSU involved, there is a good chance that the system will attain the necessary EMC compliance. Engineers can thus avoid the difficulties associated with including external filtering into their design. As we will see, things start getting more complicated once multiple PSUs have to be taken into account.
The AC-DCs fitted as modules within a system will normally have a wired connection to a chassis mains connector or to a flying lead where the EMI measurements are made. This means that there could be pick-up on this cabling, which can heighten the risk of EMC compliance not being met. Provided the AC-DC module’s EMC compliance has a reasonable margin, then the system should be able to gain certification, as long as care has been taken with the grounding and cable routing.
Figure 1 The EMI pick-up is commonly witnessed on cables. Source: CUI Inc.
Single external filter
There are several reasons why specifying an AC-DC with internal filters may not be the best approach, however. If, for instance, full EMC compliance cannot be achieved with an internal filter alone, then an external filter may still need to be added to the design. Not only will this add to the overall cost of components, it will also take up additional engineering resources. The ohmic losses that the extra filtering brings are sure to have an impact on overall system efficiency. Furthermore, in some situations, it is possible that the internal and external filters might interact with one another, causing resonances at certain frequencies that could increase EMI levels.
Rather than this over-engineered and inefficient arrangement, having a single external filter can prove to be a far better solution. Such a filter would be located close to the AC inlet and matched better with the actual system load. Internal filters tend to have specifications that deal with worst-case scenarios in terms of output load and temperature levels. The actual applications for which they are intended may not have these exacting demands though. So, if an external filter is sourced and added to the system after incorporation of the PSU or PSUs, it’s likely that this can be significantly less expensive and also have a smaller size format. Both of these are advantageous from engineering and commercial perspectives.
In a design where there are several PSUs placed in an N+n configuration, with at least one of these idling continuously, having an external common filter rated for just the actual number of PSUs powered at any one time would be possible. The need for internal filters for each PSU would be negated and notable bill-of-materials (BOM) cost reductions could be achieved.
Figure 2 Individual filtering on multiple PSUs (left) compared to a more streamlined approach with a single filter (right). Source: CUI Inc.
Optimal filtering strategy
There are other issues that can arise when a system features multiple PSUs. Additive effects will mean that the EMI characteristics witnessed are more complex, with external filtering being mandated as a result. This will lead to the potential for interactions between all the different filters involved, particularly ones not loaded at that particular time. The effect that the earth leakage current which each filter contributes to the system’s overall efficiency will need to be given serious consideration.
CUI can provide design engineers with a variety of prospective options when it comes to PSUs, allowing different use case scenarios to be attended to. Solutions can be sourced with either EMI filters included or absent, depending on what application circumstances dictate. Take the PSK-15W Series PSUs for example. These comply with CISPR 32 Class B limits as a PCB module with internal filtering. They are Class II safety rated—additional insulation having been included—with very low leakage current, so the system’s operational parameters are not impinged upon.
Another example is the 15-W-rated VOF-15B Series. The AC-DC units incorporated within this series have filtering included, thereby allowing them to meet Class B EMC requirements. As these are open-frame PSUs with inherently low EMI levels witnessed, the filtering levels needed are very modest. This translates into minimal additional expense being called for.
To keep overall costs down still further, the PBO-15C Series PSUs come without internal filtering included. Instead, these 15-W output units rely upon support from a system filter, which in many cases, would already need to be present anyway. Again, low current leakage is maintained.
Though ensuring that EMC requirements are met is of paramount importance, this can be achieved in several different ways. Depending on the criteria, the optimal filtering strategy can take quite different forms. By giving thought to where filtering is best applied, it is possible to avoid much of the engineering effort associated with certification, as well as keeping BOM costs down and avoiding interaction as well as current leakage concerns.
This article was originally published on Planet Analog.
Ron Stull is power systems engineer at CUI Inc.