VOOHU switching power supply current transformer selection: turns ratio, load resistance and overcurrent protection verification

In PFC, half-bridge, full-bridge and larger power flyback or forward power supplies, the main switch current is often a narrow pulse. Engineers hope to turn this pulse into a voltage that the controller can recognize without increasing conduction losses too much. Therefore, current transformers are often placed in primary current sampling or cycle-by-cycle overcurrent protection channels. Just because the prototype can start up does not mean that the protection point is reliable: if the value of the load resistor is too small, the signal seen by the controller may be too low; if the value is too large, it may amplify the spike, raise the reset voltage, or make it more difficult for the core to reset during the off time.

The real starting point for selection is not "just close to the maximum current", but the controller's current detection threshold, target peak current, switching frequency, maximum duty cycle, reset method, isolation boundary and PCB layout. The official website of VOOHU can currently check the existing page of current transformers such as EE5, EP7, EP10 and EF12.6. This article uses four public material numbers to establish the primary selection direction, and strings calculations, waveforms and protection actions into a reproducible board-level verification process. The page fields are only used to narrow the scope. The mass production conclusion must still be based on the latest specifications, controller reference design and physical samples corresponding to the complete suffix.

1. First convert the primary current to the controller threshold, and then select the turns ratio

Turns ratio determines signal scaling, but the controller sees more than the ideal scaled value

If the primary conductor passes through the magnetic core to form a turn, the 1:N marked on the product can be used as the initial conversion relationship. Under ideal conditions, the secondary current can be approximately written as Is≈Ip/N; when the secondary side load resistance is Rb, its terminal voltage can first be estimated as Vb≈Ip×Rb/N. Conversely, if you want the target peak current Ipk to correspond to a certain target detection voltage Vtarget, you can use Rb≈Vtarget×N/Ipk to get the first version of the resistance value. This formula is suitable for establishing direction, but it cannot be directly used as a BOM conclusion.

The true voltage on the CS pin of the controller is also affected by the rectifier diode voltage drop, winding resistance, excitation current, load resistor tolerance, RC filtering, clamping network and trace parasitics. Different controllers may also have minimum, typical and maximum thresholds with leading edge blanking or internal ramp compensation. Therefore, the calculation table must retain at least two columns of "transformer secondary terminal voltage" and "controller pin voltage" at the same time. It cannot end with just writing a turns ratio on the schematic diagram.

Higher turns ratio does not mean safer, signal-to-noise ratio and reset margin should be looked at together.

Under the same primary current and load resistance, increasing N will reduce the secondary current and output voltage; if Rb is increased year-on-year in order to restore the detection amplitude, the load terminal voltage, reset voltage, parasitic capacitance charge and discharge, and spikes will change. A resistance value that is too small will suppress the effective signal and increase the threshold error and noise proportion; a resistance value that is too large may cause the secondary side voltage and power consumption to exceed the original assumptions. The power consumption of the load resistor should be calculated according to the RMS value of the actual secondary current waveform, rather than simply replacing the peak value multiplied by the duty cycle.

2. The current transformer cannot transmit DC, and the reset condition must be entered in the selection table.

Only looking at the on-time and not looking at the shutdown reset is the most common misunderstanding in switching power supplies.

Current transformers are suitable for detecting AC or pulses that return to near zero per cycle, but cannot maintain the DC component. During the switch conduction period, the secondary side maps the primary current pulse to the detection network; after the switch is turned off, the magnetic core must release the excitation energy through the path specified by the reset resistor, diode, zener or controller reference circuit, so that the next cycle starts from the controllable flux state. If the maximum duty cycle is increased, the off-time is shortened, or the reset path is restricted, magnetic flux may accumulate cycle by cycle, ultimately manifesting as waveform droop, threshold drift, or even core saturation.

There is no fixed resistance value for the reset network that applies to all topologies. The volt-second balance of the on and off phases, diode reverse withstand voltage, maximum duty cycle and abnormal operating conditions should be checked according to the data sheet or reference design of the controller used. In addition to observing the sampling pulse amplitude on the oscilloscope, we also need to look at the reset voltage after shutdown, whether the waveform returns to the stable baseline before the next cycle, and whether there is a cycle-by-cycle offset at high temperatures and the lowest input voltage.

3. Put the public fields on the official website back into the real electrical boundary for explanation

The maximum current field is used for preliminary screening and cannot be directly equated to the over-current point of the project.

The VOOHU product page discloses the IR(A) Max field, which can help distinguish the 20 A and 40 A directions, but the web page does not give all the pulse duration, frequency, temperature rise and waveform conditions required for this article. The Ipk in the project may be a single-cycle peak value, a short-circuit peak value, or a continuous repetitive pulse. The three have different effects on copper loss, core and terminal temperature rise. Therefore, the 20 A or 40 A on the page cannot be directly written as the controller OCP threshold, nor can the long-term carrying capacity be derived from just one peak field.

Inductance, isolation and size require separate answers to separate questions

The public page also lists the inductor L, turns ratio, isolation voltage and core size. Inductance will affect the excitation current and pulse sag, but the values ​​under different test frequencies, voltages and connection methods cannot be unconditionally compared horizontally; the isolation withstand voltage is used to establish the insulation direction, but it does not automatically equal the operating voltage of the whole machine, creepage distance or a certain safety certification; EE5, EP7, EP10 and EF12.6 are related to the board area, pins, welding method and thermal environment. The four categories of fields should be entered into electrical, insulation, structural and manufacturing reviews separately, rather than combined into one "higher performance" ranking.

4. Rectification, load, filtering and layout jointly determine whether the overcurrent protection is credible

A typical current transformer sampling chain includes polarity definition, rectifier or clamping devices, load resistors, reset network, and filtering to the CS pin of the controller. If the polarity of the device is reversed, the correct forward signal may not be obtained in the required conduction interval; the diode voltage drop and temperature change will move the detection point under small signal; an RC that is too heavy will weaken the true peak value or delay protection, and an RC that is too light may cause the turn-on spike to be triggered by mistake. All components should be identified around the controller reference circuit and threshold timing.

In terms of layout, the transformer secondary loop, load resistor and controller CS loop should be short and compact, away from the MOSFET drain, gate loop, transformer high dv/dt node and high current commutation ring. Filtering and clamping devices should typically be located close to the controller pins and return with a clear, low-noise reference return. During debugging, only by measuring the primary current, transformer secondary voltage and CS pin voltage at the same time can we distinguish whether the spike comes from the power loop, transformer leakage inductance, rectification recovery or ground bounce near the controller.

5. Create a preliminary selection list using accessible on-shelf material numbers

The four Chinese and English detail pages in the table below can be accessed on August 5, 2026, and the product site map includes the corresponding material numbers. The table only uses the L, maximum current, turns ratio, isolation withstand voltage and size fields that are clearly disclosed on the page to map the project conditions to the direction of further requesting information and samples; it does not constitute a commitment to substitution, inventory quantity or mass production applicability. Before ordering materials, you must still check the test conditions, pins, polarity, recommended pads, temperature grade, packaging and delivery date in the latest specification.

6. Split the overcurrent protection verification into four reproducible steps

Step One: Establish Thresholds and Error Budgets

Obtain the minimum, typical and maximum values ​​of the CS threshold from the controller data, and define the normal peak value, current limit target and worst-case short-circuit current. Estimate the secondary current and load resistance based on the turns ratio, and then incorporate the diode, winding, resistance tolerance, temperature and controller threshold deviation into the upper and lower limits. Normal full load must not be accidentally triggered, and abnormal current must be limited before the allowable stress of the power device, main transformer and rectifier.

Step 2: Check reset and device stress at maximum duty cycle

Check the reset volt-seconds for conditions such as minimum input voltage, maximum load, startup, and transients that push the duty cycle higher. Confirm the peak voltage and power consumption of the transformer secondary terminal, diode, zener, load resistor and controller input, and retain the measured margin after simulation or calculation. If the waveform's baseline moves or the slope is abnormal in consecutive cycles, the reset and saturation problems should be solved first, and do not cover the phenomenon with heavier filtering.

Step 3: Do waveform correlation test within hot, cold and input ranges

Use a calibrated current probe or a reference method approved by the project to measure the primary current, and collect the secondary side and CS pin waveforms at the same time. Covers minimum vs. maximum input, light load to full load, target high and low temperatures, startup, load steps, and different switching frequencies, and records peak ratio, pulse droop, leading edge spikes, reset voltage, and noise margin. The grounding method of the test probe must also be fixed to prevent false spikes from being introduced into the measurement loop itself.

Step 4: Verify protection action with recoverable fault conditions

Perform overload, output short circuit or project-specified fault tests under controlled energy-limited conditions, and record trigger current, response period, device peak stress, shutdown mode and recovery behavior. Then review the temperature rise of the transformer, load resistor, rectifier device, power switch and main magnetic parts. Only when normal working conditions do not malfunction, abnormal working conditions can be promptly limited, and repeated test results are consistent, can the transformer and resistance combination be released to mass production BOM.

7. Frequently Asked Questions (FAQ)

Can a current transformer measure DC current or steady-state average current?

DC cannot be measured directly. It relies on changing magnetic flux coupling current information and is suitable for AC or pulse sampling with a clear reset interval. If your project requires DC or bidirectional averaging, shunt resistors, Hall, fluxgate, or other suitable sensing solutions should be evaluated and isolation and bandwidth requirements redefined.

The higher the turns ratio, the more accurate the overcurrent protection?

uncertain. Higher turns ratios change the secondary current, required load resistance, output amplitude, parasitics, and reset conditions. Accuracy also depends on threshold tolerance, diode, excitation current, temperature, layout and calibration method. A combination of back-extraction from the target CS voltage and the full error budget should be performed rather than pursuing higher N alone.

Can the IR(A) Max on the official website be directly used as the current limit value of the controller?

cannot. This field is suitable for initial product screening, but the controller current limit value also depends on the actual pulse waveform, duration, frequency, temperature, threshold, tolerance and protection strategy. After selection, the current specifications must be obtained and verified in real power levels.

Can the secondary side of the current transformer be left open during debugging?

It is not recommended to open the secondary side when there is current on the primary side. Open circuits may generate higher secondary voltages and pose risks to devices and personnel. Board-level debugging should ensure that the load and reset paths are correctly connected, that power is removed before changing wiring, and that the safety regulations of the transformers, controllers, and complete machines used are followed.

Conclusion

The value of the switching power supply current transformer lies in converting high-side or large current pulses into isolable and processable detection signals; its difficulty is also in the same signal chain. First use the controller threshold and target Ipk to infer the turns ratio and load resistance, then check the reset, isolation and layout under the maximum duty cycle, and finally use the primary current, secondary voltage and CS pin waveform to complete the correlation verification. The public material numbers on the VOOHU official website can help narrow down the directions of EE5, EP7, EP10 and EF12.6, but the real mass production conclusions still come from current data, samples and failure tests.