PMOS is used as PWM switch power supply. The output PWM voltage signal has negative voltage overshoot voltage for too long. Please solve it.

flag2010

PMOS is used as PWM switch power supply. The output PWM voltage signal has negative voltage overshoot voltage for too long. Please solve it. [Copy link]

 

See the schematic diagram below: The VBAT voltage in the circuit is 3-4.2V, and the D401/D4 Zener diode is a 4.3V Zener diode.

Q6: NPN tube LMBT4401LT1G

Q7: PNP tube: LMBT4403LT1G

Q5: NMOS tube: LNTS4409NWT1G

Q403 is AOS material: AONR21117

VBAT is the power supply for the load, the supply voltage is 3V-4.2V, and the load connected to the lower end of Q403 of PMOS is a 0.4 ohm resistor. The PWM input signal of the microcontroller is a 3.3V square wave signal with a frequency of 20KHz.

This circuit comes from: a pdf file on TI's official website, which is also a classic circuit, page 27

https://www.ti.com/lit/slua618

This is the circuit I designed:

But I found out that when the PMOS_PWM signal is input into the PWM signal, a serious negative voltage is measured at both ends of 0.4R, and the time is very long. See the oscilloscope waveform, the PWM switching frequency is 20KHz.

Can anyone help me solve the problem of how to optimize and improve the negative voltage so that the signal outputs a standard square wave signal? Now the rising and falling edges of the turn-on signal are still very fast.

The back-end tried to use a 10uF capacitor for C404, but it seriously weakened the rising and falling edge time, and the output signal was not standard. The signal tested by the oscilloscope in the figure above is the waveform when C404 is a 1uF capacitor.

Jacktang

It's all C404's fault.

Remove it and optimize again

flag2010

Jacktang posted on 2024-10-13 08:36 It was C404 that made a fuss and removed it, and then re-optimized it

The effect of canceling C404 is the same. I have tried it, and the negative voltage is too long. And the negative amplitude is larger.

maychang

Where are the 1/2/3 pins of the P-channel MOS tube connected to in the first post? What other components are on this NET? Is it only the VBAT2357?

maychang

The connection method of transistor Q7 in the first post is wrong anyway, and the voltage direction is wrong.

How many V does VBAT2357 have?

davidzhu210

The intention is to use Q6/Q7 to drive the MOSFET, but this drive circuit is seriously wrong. What is the VTH of the MOSFET? Also, where is the drain of the MOSFET connected? Please give me the picture so that I can analyze it.

flag2010

maychang posted on 2024-10-13 11:11 Where are the 1/2/3 pins of the P-channel MOS tube in the first post connected to? What other components are there on this NET? Is it only VBAT2357?

I have updated the post and the description is more complete. Please help me take a look. VBAT is directly connected to a 4.2V lithium battery, and the operating voltage range is 3V-4.2V. The connection method of the transistor is designed based on the information on TI's official website.

flag2010

davidzhu210 posted on 2024-10-13 11:19 The intention is to use Q6/Q7 to drive the MOSFET, but this drive circuit is seriously wrong. What is the VTH of the MOSFET? Also, where is the drain of the MOSFET connected? Give the picture to everyone...

This circuit is based on the design in TI's official data. Please help me find out where the design problem lies.

The VTH voltage of the PMOS driving resistor 0.4R is typically -0.7V. See the screenshot or the information in the post for detailed specifications. Please help me find out where the problem is. The output of both ends of 0.4R has a serious negative voltage for too long when it is turned off, but no serious overshoot and ringing are obviously seen when it is turned on.

flag2010

maychang posted on 2024-10-13 11:13 The connection method of transistor Q7 in the first post is wrong anyway, and the voltage direction is wrong. How many V does VBAT2357 have?

Does it mean that Q7 is connected in reverse 2/3, causing the voltage at the G and D poles to not be released quickly when the PMPS is turned off?
I connected the PNP tube in reverse 2/3, and the phenomenon is actually the same. Or the negative voltage lasts too long.

davidzhu210

Do you mean measuring current? Maximum 10A

davidzhu210

The negative voltage at 0.4R is probably caused by the Miller effect of PMOS. Try reducing R5 or try another tube. The Coss of this tube is 770pF, which is relatively large.

flag2010

davidzhu210 posted on 2024-10-13 19:57 The negative pressure at 0.4R should be caused by the Miller effect of PMOS. Try reducing R5 or try another tube. The Coss of this tube is 770pF, which is relatively...

I replaced it with a new Clean Energy PMOS (model NCE30P25Q), but the effect was still not ideal, so I directly raised the G-level voltage to 5V.