This month, I'd like to demonstrate the EMC risk in purchasing and using counterfeit power modules. These counterfeit modules may be obtained from numerous sources in Asia and are much lower in cost. But lower cost comes at a price, not only in product safety, performance, and reliability, but EMC compliance.
Many power modules, sometimes referred to as "wall warts," purchased through discount on-line Asian companies, lack basic power input filtering. Even Amazon has issues weeding out third-party sellers of non-compliant illegally imported power modules.
Some years ago, I purchased a batch of nearly identical clones of a standard Apple iPhone power module for experimental purposes. Even the labeling and compliance markings were the same. In the intervening years, these became intermingled with several genuine modules, so the immediate task was to measure each and separate out the clones. What I found in measuring the conducted emissions should not have been a surprise.
I'll discuss my experience with a typical non-compliant (unlabeled) power module and then describe the issue with the counterfeit Apple modules.
Non-Labeled Power Module
Figure 1. A disassembled power module showing a lack of power line filter.
Measuring the conducted emissions from 150 kHz to 30 MHz per CISPR 32 (consumer products) demonstrates the expected lack of compliance (Figure 2).
Figure 2. The module with unfiltered power input indicates more nearly 35 dB over the Class B limit for residential consumer products.
After experimenting with a conventional discrete "tacked together" line filter (Figure 3), we see the results in Figure 4. The two series inductors were removed for the measurement.
Figure 3. A conventional line filter was soldered together and remeasured. The two series inductors were not used for the measurement.
Figure 4. The module with the added line filter nearly meets the conducted emission Class B limit in peak detection mode and would meet it nicely if the two peaks were measured using quasi-peak detection.
Apple Power Modules
Figure 5. Apple modules versus the clones.
Figure 6. The test setup used for the conducted emissions measurement.
The typical results are recorded in Figure 7 (counterfeit modules) and in Figure 8 (Apple modules). These were both captured in peak detection, so the official quasi-peak measurement would be a little less in amplitude.
Figure 7. The counterfeit module measured well over both the CISPR and FCC Class B limits in peak detection mode.
Figure 8. The genuine Apple module measured much less, as expected, in peak detection mode and with quasi-peak detection would be well under the limits.
One observation was that the counterfeit modules were unable to support a full 1A load. Most would barely provide a 0.5A load, and I had to measure several at just a maximum of a 0.2A load. These would not have been good choices for charging phones.
I also measured several Belkin modules and all these met the compliance limits at a full 1A load. Unsurprisingly, several "swag" modules (freebies from show exhibitors) were poor performers as well. Modules labeled “DesignCon” and one from a major component manufacturer were non-compliant and could barely support a 0.5A load.
Obviously, the lesson here is you get what you pay for; the premium pricing for a genuine Apple, Belkin, or other name brand product will generally meet the performance specifications, as well as EMC compliance. There is also the issue of interference in the AM and shortwave broadcast bands. I'd also be suspicious of the safety and reliability of non-brand name, cloned or counterfeit products – especially those that plug into mains.