Constant current source + NMOS and PMOS (MCU control on and off) + LED solution consultation

BlueInvestment

Constant current source + NMOS and PMOS (MCU control on and off) + LED solution consultation [Copy link]

 

Design principle and goal:
The constant current source chip outputs constant current to the LED, and the single-chip microcomputer controls NMOS+PMOS to achieve LED high brightness for 200us, cycle

Problem: From the waveform of the sampling resistor in series with the LED, the current value has declined. Why is this?

BlueInvestment

Boss, please take a look @ qwqwqw2088

qwqwqw2088

BlueInvestment published on 2020-6-29 09:10 Boss, please take a look @qwqwqw2088

There is a problem with the MOS tube drive. There is a ringing phenomenon on the G pole.
A driving resistor should be added to the G pole, between a few ohms and tens of ohms. Debug it to confirm

BlueInvestment

qwqwqw2088 posted on 2020-6-29 09:46 There is a problem with the MOS tube driver. There is a ringing phenomenon on the G pole. A driving resistor should be added to the G pole, between a few ohms and tens of ohms. Debug it to confirm

Boss, this is the circuit diagram

BlueInvestment

qwqwqw2088 posted on 2020-6-29 09:46 There is a problem with the MOS tube driver. There is a ringing phenomenon on the G pole. A driving resistor should be added to the G pole, between a few ohms and tens of ohms. Debug it to confirm

Should a driving resistor of several ohms to tens of ohms be added between the IO port of the microcontroller and the G terminal of the NMOS?

qwqwqw2088

BlueInvestment published on 2020-6-29 10:30 Is it necessary to add a driving resistor of several ohms to tens of ohms between the IO port of the microcontroller and the G terminal of the NMOS?

R3 is a bit big

BlueInvestment

qwqwqw2088 posted on 2020-6-29 11:41 R3 is a bit big

I used a 22R resistor for R3 and found that the waveform did not change. I don't know why.

qwqwqw2088

BlueInvestment Published on 2020-6-29 12:22 R3 I used a 22R resistor and found that the waveform did not change. I don't know what the reason is

It is mainly affected by the Q1 switch tube. There are spikes in the switch, and the Q1 drive is a bit powerless, but it is not due to insufficient power.

Or reduce the switching frequency

Or directly drive Q1 with low level to observe whether the waveform at the end is normal.

BlueInvestment

cruelfox

Why do you try to use PMOS as a power switch here? According to your diagram, if you want the PMOS tube Q1 to be turned on, the gate voltage must be negative. How is this possible?

When the LED is on, the current flows through D5, R1, and Q2, but not Q1.

BlueInvestment

qwqwqw2088 posted on 2020-6-29 13:28 BlueInvestment posted on 2020-6-29 12:22 I used a 22R resistor for R3 and found that the waveform did not change. I don't know what the reason is...

I have a question, when I use this circuit to drive one light source or four light sources in series, the instantaneous current collected is decreasing and smooth. Why is this? Please see the picture.

qwqwqw2088

It is difficult to judge by just looking at this waveform. It may be related to the constant current source loop design, because it is not overshoot.

In fact, the picture below is quite normal.

The drop is not a problem, the stability of the LED temperature after it rises is the main thing

XIAOHU88666

The design is quite innovative and worth learning. I give a thumbs up to the OP! ! ! ! ! ! ! ! !

BlueInvestment

qwqwqw2088 posted on 2020-7-3 06:49 It is difficult to judge just by looking at this waveform. It may be related to the constant current source loop design. Because it is not overshoot, the figure below is actually quite normal. ...

Good morning, boss. I found a problem during the test. Under the same conditions, the waveforms captured at both ends of the sampling resistor and the waveforms captured at both ends of a single light source using an oscilloscope have different trends. Is this caused by the characteristics of the resistor and the diode? Can an oscilloscope current probe (expensive) help solve this problem?

qwqwqw2088

BlueInvestment published on 2020-7-7 08:39 Good morning, boss. I found a problem during the test. Under the same conditions, the waveforms at both ends of the sampling resistor and the waveforms at both ends of a single light source captured by an oscilloscope are the same.

It depends on the test method.

In fact, the voltage conversion across the sampling resistor can be tested to be the same

Also, the final result of the lamp is to test whether the LED luminous intensity is sufficient, whether the intensity decreases after lighting, and whether there is flicker, which is also related to the constant current stability.