Howling reasons for switching power supply
Those who have done development work staff have such experience, the test switching power supply or in the experiment have heard similar products to fight bad sound or high-pressure leakage arcing sound uninvited: their sound, large or small, or from time to time; its rhythm or deep or shrill, or volatile persons in there.
1. transformer (Transformer) dipping bad: Includes non-impregnated varnish (Varnish). Whistle and caused waveforms spikes, but generally normal load capacity, in particular: greater power output by howling even more, the low-power performance is not necessarily clear who is. I worked in a 72W charger products have had bad experiences with the load, and find the core material has very strict requirements in this product. Vibration generating abnormal sound (this product to customer requirements more stringent) add that, when the poor design of the transformer may also be at work.
2. PWM IC ground trace errors: Normally products will be part of the performance can work, but there are some products but can not load and start-up may not be a failure, especially when the use of certain low-power IC, and more It may not work properly. I have used the SG6848 test board, because when there is no thorough understanding of the performance of the IC, with experience hastily layout, the result can not even do a wide voltage test during the test. Misery!
3. optocoupler (Opto Coupler) operating current point alignment errors: When the optocoupler operating current position of the connection resistance before there will be a secondary filter capacitor may be howling, especially when the tape a long time even more.
4. the reference voltage regulator (Regulator) IC TL431 grounding wire mistakes: the same secondary voltage regulator IC ground reference and ground, like primary IC has similar requirements, it is not directly and transformers connected to cold Geothermal . If even the consequences together with the capacity to drop and is howling and the output power is proportional to the size. When the output load is large, close to the power power limit, the switch transformer may enter an unstable state: the previous cycle switch duty cycle is too large, the conduction time is too long, too much high-frequency transformer energy transmission ; DC rectifier inductive energy storage period is not fully released this week, the judge did not produce PWM cycle next make the switch turns the drive signal or duty cycle is too small; the whole cycle after the switch is turned off or conduction time is too short; inductor through more than one full cycle of energy release, the output voltage drops, the switch duty cycle within a cycle turn big ...... again and again, so that the occurrence of low frequency transformer ( intermittent regular full-cutoff period of dramatic changes in the frequency or duty cycle) of the vibration, low frequency sound emitted the human ear can hear. Meanwhile, the output voltage fluctuation also increased compared with normal operation. When the unit of time intermittent full cutoff cycle number reaches a substantial proportion of the total number of cycles, or even make the original work to reduce the frequency of ultrasonic vibration in the transformer band, into the human ear audible frequency range, a sharp high frequency "whistle called. " At this time of the switching transformer work in a serious overload condition, the time may have burned - this is a lot of power before burning "screaming" the origin, I believe that some users have had similar experiences.
Load, or load switch is also likely to occur when very light intermittent full cut-off period, the switch transformer overload state in the same work, the same is very dangerous. To solve this problem, it can be preset at the output load of false solutions, but in some "saved" or the power supply is still occasionally occur. When not load or the load is too light, the back EMF generated by the transformer at work are not well absorbed. Such coupling transformer will be a lot of clutter to your 1.2 windings. The noise signal includes an AC component of many different spectrum. Of which there are many low-frequency waves, when the natural oscillation frequency of the low-frequency waves is consistent with your transformer, then the circuit will be formed low frequency self-excited. Core transformer no sound. We know that the human hearing range is 20--20KHZ. So we designed the circuit, usually coupled with frequency selection circuit. In order to filter out low frequency components. Your schematic point of view, you'd better be in the feedback loop plus a bandpass circuit, in order to prevent low-frequency self-excited, or is your fixed-frequency switching power supply can be made.
Power switching power supply short circuit whistle
I believe we encountered such a situation, the switching power supply in power after the sudden short-circuit test fully loaded, sometimes have the power to hear the howling of the case; or when setting current protection, when the current debugger to a Dan, there will be howling, howling sound of its cadence, it is very annoying, mainly for the following reason:
When the output load is large, close to the power power limit, the switch transformer may enter an unstable state: the previous cycle switch duty cycle is too large, the conduction time is too long, too much high-frequency transformer energy transmission ; DC rectifier inductive energy storage period is not fully released this week by the PWM judge, the next did not have to make the switch drive signal conduction cycle or duty cycle is too small; the off switch in the entire period after state, or conduction time is too short; inductor through more than one full cycle of energy release, the output voltage drops, the duty cycle of a cycle switch will big ...... again and again, so that the occurrence of a low frequency transformer (intermittent regular full-cutoff period of dramatic changes in the frequency or duty cycle) of the vibration, low frequency sound emitted the human ear can hear. At the same time, the output voltage fluctuation also increased compared with normal operation. When the unit time When the proportion of a considerable number of intermittent full cutoff period to achieve the total number of cycles, or even make the original work to reduce the frequency of ultrasonic vibration in the transformer band, into the human ear audible frequency range, a sharp high frequency "whistle call." At this time of the switching transformer work in a serious overload condition, the time may have burned - this is a lot of power before burning "screaming" the origin, I believe that some users have had similar experiences load, or load is light When the switch may also be intermittent full cut-off period, the switch transformer overload state in the same work, the same is very dangerous.
To solve this problem, it can be preset at the output load of false solutions, but in some "saved" or the power supply is still occasionally occur. When not load or the load is too light, the transformer produced at work EMF are not well absorbed. Such coupling transformer will be a lot of clutter to your 1.2 windings. The noise signal includes an AC component of many different spectrum of which there are many low-frequency waves, when the low-frequency wave and you transformer When the natural oscillation frequency of the same, then the circuit will be formed low frequency self-excited magnetic core of the transformer will not sound. We know that the human hearing range is 20--20KHZ. So we designed the circuit, usually plus option frequency circuit. to filter out low frequency components from your schematic point of view, you'd better be in the feedback loop plus a bandpass circuit, in order to prevent low-frequency self-excited, or is made of your fixed-frequency switching power supply that is can.