RFQ/BOM 0 Se connecter / Registre

Sélectionnez votre emplacement

image de l'utilisateur

What is the reason for the abnormal push-pull output waveform?

Hardware design
novembre 24, 2020 by Edmund 524

As shown in the figure, the input PWM is a square wave signal with a frequency of 4M and a duty cycle of 50%. The output requires a sine wave of 120V, the transformer is PEI32, and the primary to secondary turns ratio is 2:2:8. Other parameters are shown in the figure, but in the design, the output waveform is completely deformed. What is the reason for the expert's guidance?

Tous les commentaires

user image

Alba Posté sur November 24, 2020

There seems to be a strong 7th harmonic. Probably there is a resonance point at the 7th harmonic.

0
user image

Cecily Posté sur November 24, 2020

Obviously, c7 and c6 and transformer inductance, respectively, constitute the 6-7th harmonic oscillation.

Increase these two capacitors, or increase the inductance of the transformer, it is best to increase the inductance of L1 at the same time, so that resonance occurs in the fundamental wave.

0
user image

Glen Posté sur November 24, 2020

Equivalent to IRF540 output capacitor, and 100p in parallel with you, 22 ≈ 230p. After series connection = 115p, the resonance is about 5.4MHz. In theory, you should be able to see a good top-cut resonant waveform, but in fact, the secondary side inductance and distributed parameters undergo high-order resonance, then the only possibility is the primary side In fact, there are great high-level incentives.

It depends on your driving circuit, 2 inverters, direct drive and the output capacity of the op amp is only 70mA. Just barely drive the MOS tube in a trapezoidal wave form.

Therefore, your dual pipes cannot run if they are turned on at the same time. At the same time, it will cause the primary side to enter low impedance. Not to mention the large power loss. The key is that the excitation energy of the primary side is completely short-circuited when common, and the moment of exit, it will generate a very fast current increase. , This will happen twice in each half cycle, which is the excitation source for high-order oscillation.

There are exactly 2 high-speed operational amplifiers, and the driving capability can barely be 4MHz. But the circuit can't do this anymore, it must be driven by 2 IOs independently, with a slightly larger dead zone. Or directly use a dedicated single-ended input driver chip with dead zone. Like EG3011, there are IR series to drive.

Only when the driver common problem is solved, you can see the normal waveform.

0
user image

Christa Posté sur November 24, 2020

It seems that the problem of the dead zone can only be solved first. In addition, according to the principle of the ZVS oscillator, an oscillating capacitor is connected in between. According to my calculation, it is about 220PF. The improvement is not big. For the time being, the problem of the dead zone is solved first.

0

Écrivez une réponse

Vous devez vous connecter pour répondre. Se connecter | Registre