I have been building on of my first PCBs where I want to control a few High power LEDs. The microcontroller I want to use doesn't have PWM outputs so I want to use a cheap I2C PWM pin expander so I can connect it to a high power MOSFET switch module: https://protosupplies.com/product/high-power-dual-mosfet-switch-module/

Unfortunately from what I understand PWM pin expanders don't really exist so I went to search online and found a LED driver that seems to be able to what I want. The PCA9633. But when I look at the characteristics I see this:

PCA9366 LED Driver output voltage

I power the IC from 3.3v so that means the output PWM High is only a little more then 2.3v? Why is that? Is there a way to I can convert that to 3.3V easily so I can send it to the MOSFET PCB? Or should I look for different PWM pin expander that can output 3.3V?

This may seem like a basic question but I don't really know what the best practice here is, so any help is appreciated!

  • \$\begingroup\$ You should power the PCA9633 from 5V. Most chips will not output the full supply voltage, so in order to get at least 3.3V into the MOSFETs the supply has to be higher than 3.3V. \$\endgroup\$
    – Paul
    May 23, 2023 at 15:08
  • \$\begingroup\$ Looks like this is the minimum (usually these are written as min/typical/max and there's only a number in the min column). That means the manufacturer is guaranteeing you will always get at least 2.3V if VDD is 3.0V and IOH is -10mA. What you actually get in reality is a different question and it could be quite a lot higher especially if IOH is a lot lower. \$\endgroup\$ May 23, 2023 at 18:04

1 Answer 1


This may seem like a basic question but I don't really know what the best practice here is, so any help is appreciated!

Your MOSFET gate realistically should be driven at close to 4V or higher:

enter image description here

(taken from here https://aosmd.com/sites/default/files/res/data_sheets/AOI4184.pdf )

If your microcontroller can output 5V, then I would run off of 5V. If it outputs 3.3V, you could use 4.5V for the PWM chip, which would allow the inputs to be compatible with 3.3V logic (it needs at least 0.7*VDD) while still easily driving the MOSFET gate at 4V.

  • \$\begingroup\$ Hmmm :/ That kind of sucks. My I2C lines are also 3.3V so If I feed 4.5V then logic high is 3.15V That's dangerously close to 3.3v I think. I think I'm considering building a I2C level shifter and running the PWM chip of 5v and also have 5V I2C signals. \$\endgroup\$ May 23, 2023 at 15:31
  • \$\begingroup\$ @HarmvanVliet Any reason you can't pull up the i2C to 3.3V while running the PWM chip at a higher voltage? \$\endgroup\$ May 23, 2023 at 15:35
  • \$\begingroup\$ The bus is pulled up to 3.3v as required. Still I think 3.15v is skirting dangerously close to 3.3V. I'm not experienced so maybe I'm wrong but I expect small noise may trip below 3.15V. And also generating 4.5V would require a LDO whereas a simple level shifter can be done with a few transistors. \$\endgroup\$ May 23, 2023 at 15:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.