Summary of problems related to single-ended flyback switching power supply circuit debugging

Single-ended flyback switching power supply circuit debugging

conclusion of issue

-

01

-

Solve the problem of excessive temperature rise of MOS tube

Taking the example as an example, the main heat source found in the actual debugging is MOS tube chip, high frequency transformer,

Output diode.

For the MOS tube chip, the actual debugging found that the temperature rise reached 45 degrees, because TOP246YN provides two switching frequencies, 132KHZ and 66KHZ, by changing the switching frequency to 66KHz, the effect is not obvious, so the main guide from the MOS tube After the loss is started, the MOS tube is replaced with the lower RDS TOP247YN, and then the external current limit point is set to reduce the temperature rise, and the temperature rise is obviously improved. Finally, the external heat sink heat dissipation area is appropriately increased, and finally the MOS tube temperature rise control is performed. At around 25 degrees

-

02

-

Solve the problem that the output rectifier temperature rise is too high

Still taking the example of this topic as an example, the output diode was found in the debugging, and the temperature rise reached 42 degrees. Since the first prototype used the MUR1060 diode of the TO220 package, the conduction voltage drop was about 1.1V, and then it was changed to conduction. The MBR10200 with a voltage drop of about 0.7~0.8V, and then the TO220 heat sink is changed to a wide-body heat sink, the output rectifier temperature rise is obviously improved, and the final output rectifier temperature rise is controlled at about 22 degrees.

-

03

-

Solve the problem that the output overshoot is too large when power is turned on.

When testing the output overshoot amplitude indicator, it is found that the output overshoot amplitude is a bit large when powering up, which exceeds the safety voltage of the power supply module of the latter stage. The analysis reason may be caused by the dynamic response of the system feedback loop is too slow, so feedback from TL431 By starting the loop, by changing the position and size of the poles in the TL431 feedback loop, the problem of excessive output overshoot at power-on can be significantly improved. After the improvement, the output voltage rises steadily to the output voltage level and stabilizes the output. I overshoot.

-

04

-

Solve the problem that the output voltage 100HZ power frequency ripple is too large

When testing the output power ripple, it was found that the 100HZ power frequency ripple was relatively large, and the following tests were performed:

1. Increase the output electrolytic capacitor capacity, select a smaller ESR electrolytic capacitor, the test results are not significantly improved;

2. Increase the capacity of the input electrolytic capacitor, the test results are not significantly improved;

3. Increase the DC gain of the feedback loop to increase the low-band gain, and the test results are not significantly improved;

4. Connect a 103-size Y capacitor between the output voltage ground and the chassis ground (the chassis ground is connected to the ground). The test result 100HZ power frequency ripple is significantly improved, basically can be controlled at 0.5%~1% of the output voltage. ;

-

05

-

Solve the problem of high temperature transformer over temperature rise

When the transformer temperature rises during the test, it is found that the temperature rise of the transformer surface reaches 50 degrees, which makes it difficult to meet the application requirements under the 85-degree working environment. The following tests were carried out:

1. Replace the core material, the core material of PC40 material was selected first, and the magnetic core material of PC44 material was replaced later, the test result was slightly improved;

2. In order to reduce the impact of the skin effect, the number of strands of the windings is increased, and the test results are slightly improved;

3. The soldering pins of the transformer on the PCB board all increase the copper surface area and expose the copper. The test results are not improved, but there is a slight improvement;

4. Slotting treatment under the transformer on the PCB board, increasing the convection speed of the upper and lower air of the transformer, the test results are not improved, but there is a slight improvement;

After the above comprehensive treatment, the temperature rise of the transformer is covered, and it can basically be controlled at about 40 degrees;

-

06

-

Improve the efficiency of switching power supply

When testing the power efficiency, the efficiency was found to be 85.3%. Although such efficiency can be applied to the product application, the experiment was carried out in accordance with the principle of excellence:

1. The feedback proportional voltage divider resistors are initially used with 37.4K and 3.19K, and now the partial pressure remains unchanged, and changed to 73.2K, 6.19K;

2. Remove the power output indicator circuit;

3. Under the premise of not affecting the stability of the loop, appropriately reduce the loop DC gain;

4. The output diode adopts a low-recovery, high-reverse-voltage fast recovery diode, and at the same time, the RC absorption circuit at both ends of the output diode is removed on the premise of ensuring the reverse voltage withstand of the diode;

5. Reasonably adjust the parameters of the RCD clamp circuit of the primary winding to avoid unnecessary energy absorption;

After the above tests, the power efficiency has improved and has now increased to 87.6%.

-

07

-

Solve the problem of high frequency howling of transformer

After the completion of various theoretical calculations, the contact proofing transformer was started. In order to facilitate the modification of the transformer parameters during the commissioning, the supplier was specially contacted to proof the two types of transformers: one that was immersed in the insulating varnish and one that did not. Dipped paint treatment. During the commissioning, it was found that the power board had a high-frequency whistling sound of several KZ during the working process, and the following tests were carried out:

1. Debug the feedback loop parameters, confirm that the loop parameters are designed reasonably, and the high frequency howling sound always exists during the debugging process;

2. Test the DS waveform and output waveform of the MOS tube switch to confirm whether the design parameters are reasonable;

3. Replace some of the ceramic capacitors into plug-in point de-capacitors to eliminate the howling caused by the piezoelectric effect of the ceramic capacitors;

4. Replace the transformer with the dipping treatment and the non-impregnated paint;

5. Change the switching frequency of the MOS tube, the howling sound is improved, but it cannot be completely solved;

After the above test, it was found that after the replacement of the varnish-treated transformer, the howling sound was basically gone, and the preliminary judgment was due to the whistling caused by the high-frequency oscillation of the transformer coil.

• end •