Energy transformers are sometimes the principle supply of common-mode noise in remoted switching energy converters. Why? As a result of contained in the transformer, the windings on the first and secondary sides of the isolation barrier are in shut proximity—normally separated by lower than 1 mm—leading to important parasitic capacitance between the adjoining windings.
The voltages that seem on these windings usually have massive AC content material. For instance, within the flyback converter proven in Determine 1, the first winding connects to the drain of the first change, which has a voltage waveform with massive AC content material throughout many frequencies. This AC voltage injects common-mode currents from main to secondary via the parasitic capacitance, which is usually the supply of many electromagnetic interference (EMI) issues.
Determine 1 Widespread mode noise created by a flyback energy transformer. Supply: Texas Devices
Fortunately, transformer design methods equivalent to shielding and common-mode balancing can reduce the transformer impression on EMI, as mentioned within the Texas Devices Energy Provide Design Seminar paper, “Flyback Transformer Design Issues for Effectivity and EMI”. It may be fairly tough and time consuming, nevertheless, to verify how a lot your transformer contributes to EMI and how you can optimize the transformer building. For every transformer design that you simply need to check, it’s good to solder the transformer into the PCB, take your energy converter to an EMI check fixture, and run the scans. If the EMI efficiency of your transformer isn’t acceptable, it’s good to unsolder it out of your PCB and check out once more.
On this Energy Tip, I’ll present you a very simple approach to verify the EMI efficiency of your transformer earlier than ever soldering it into your board.
Utilizing solely a perform generator and an oscilloscope, you’ll be able to mimic the circumstances seen by the transformer within the circuit and measure the transformer’s common-mode EMI signature. The diagram in Determine 2 reveals how you can configure this measurement for the transformer utilized in Determine 1. Discover that this transformer has two windings on the first (WP and WAUX) and one winding on the secondary (WS).
First, use a brief piece of wire to tie the AC quiet nodes collectively on the first. An AC quiet node is any pin on the transformer that ties to main floor within the circuit, both on to or via a capacitor. On this instance, each Pin 2 and Pin 3 are AC quiet nodes on the first aspect of the isolation barrier. When you’ve got a transformer with a number of windings on the secondary, additionally, you will have to tie all the secondary quiet nodes collectively, however don’t join them to the first quiet node.
Determine 2 Transformer CMRR check setup that makes use of a brief piece of wire to tie the AC quiet nodes collectively on the first and secondary, and applies a small sinewave throughout the first winding to measure the ratio between the voltage induced between the first and secondary AC quiet nodes and the voltage injected by the perform generator, or CMRR. Supply: Texas Devices
Subsequent, use the perform generator to use a small sinewave throughout the first winding of the transformer. This mimics the first winding voltage, however now you’re testing at a single frequency with a secure and low voltage. The amplitude of the sign isn’t important, because the parasitic capacitances of the transformer are largely impartial of the voltage amplitude.
Lastly, utilizing one channel of the scope, measure the voltage injected by the perform generator. With one other channel, measure the voltage induced between the first and secondary AC quiet nodes. The ratio of those two alerts is basically the common-mode rejection ratio (CMRR) and is a sign of how a lot your energy transformer will contribute to common-mode noise at that frequency.
Determine 3 reveals the outcomes of this check at 100 kHz for 2 completely different transformers. The development used for transformer #1 leads to a CMRR of –39.6 dB, whereas the CMRR for transformer #2 is greater, measuring –31.4 dB. This means that transformer #1 will produce much less common-mode noise than transformer #2. With the perform generator, you’ll be able to examine the transformers’ traits at completely different frequencies.
Determine 3 The time area transformer CMRR check outcomes that signifies that transformer #1 produces much less common-mode noise than transformer #2 on the check frequency of 100 kHz. Supply: Texas Devices
Alternatively, you’ll be able to carry out this similar check utilizing a frequency response analyzer (FRA) to comb the frequency of the injected sign throughout your complete frequency vary of curiosity. Determine 4 reveals the FRA measurements of the identical two transformers throughout a large frequency vary of 100 kHz to 30 MHz. Discover that the acquire could be very flat over a variety from 100 kHz to round 4 MHz. The acquire at 100 kHz correlates very nicely with the perform generator check, indicating that the perform generator check at 100 kHz is enough to characterize these transformers throughout this band of frequencies. At frequencies above a couple of megahertz, it’s best to measure the CMRR of those transformers on the frequency of curiosity.
Determine 4 Frequency area transformer CMRR check outcomes for transformer #1 and #2 throughout a large frequency vary of 100 kHz to 4 MHz utilizing an FRA. Supply: Texas Devices
Determine 5 reveals the outcomes of soldering each of those transformers into the PCB of a switching energy converter measuring the performed EMI in opposition to Comité Worldwide Spécial des Perturbations Radioélectriques (CISPR) 32 Class B limits. The highest restrict line corresponds to the quasi-peak measurement, and the decrease restrict line corresponds to the typical measurement. As anticipated, the EMI outcomes for transformer #2 are worse than transformer #1. Actually, transformer #1 passes with first rate margin, whereas transformer #2 barely fails.
Determine 5 Performed EMI check outcomes for transformers the place transformer #1 passes with margin and transformer #2 barely fails. Supply: Texas Devices
Apparently, each transformers on this instance have the identical winding construction and building. The variations in CMRR are utterly attributable to variations within the manufacturing course of, demonstrating how delicate EMI will be to transformer building. Small variations equivalent to the precise placement of particular person strands of wire inside the transformer or the thickness of insulating layers can have profound results.
For the instance in transformer building, it’s clear that you may’t be assured that every one models in manufacturing will go CISPR 32 performed EMI limits. One resolution is to extend the EMI filtering within the circuit to supply extra margin. Another choice is to make use of the perform generator check to display each transformer pattern throughout manufacturing. This check is similar to the forms of assessments generally used to check and display transformer flip ratios between windings, so no particular gear is required. Within the instance, solely passing transformers with a CMRR lower than –38 dB provide a excessive chance that every one models will go EMI when assembled into an influence converter system.
The transformer’s impression on EMI
Debugging EMI points is fraught with many obstacles and difficulties. The easy measurement approach described on this Energy Tip can prevent important time and frustration on the solder bench and within the lab. To your subsequent remoted power-supply design, take a couple of minutes to measure the CMRR of your energy transformers earlier than soldering them into the circuit boards, after which examine the CMRR to the ensuing EMI. You’ll acquire a greater understanding of the transformer’s impression on EMI, and what stage of transformer CMRR will go EMI in your system.
Brian King is a Techniques Supervisor and Senior Member Technical Workers at Texas Devices. With over 26 years of expertise in energy provide design, he has supported over 1300 enterprise alternatives and has designed over 750 distinctive energy provides utilizing a broad vary of TI energy provide controllers. Brian has printed over 45 articles associated to energy provide design, and since 2016 has been the lead organizer and content material curator for the Texas Devices Energy Provide Design Seminar (PSDS) sequence.
Associated Content material
