PWM Buck Regulator Question

Thread Starter

tvist21

Joined Jun 11, 2021
20
Questions:
1. Why do switching regulator ICs have dedicated PWM inputs pins?
2. Can you use PWM on a low side switch at the output of a constant voltage buck regulator module? For instance Remcom Power's R-78E5.0-1.0. My concern is that dedicated PWM inputs operate differently than using an external low side switch on the output of a module. Concerns include:
a) Would the buck regulator module be able to switch completely on and off quickly enough?
b) Would buck regulator module have inductor current oscillation issues from turning on and off so quickly?


Context:
I have a single high power RGBW LED with each channel requiring between 2.7 and 3.5 Vdc at 150mA per channel.
I have an Bluetooth RGBW controller that outputs 12 Vdc.

I want to be able to connect the two together but I do not want to step down the voltage by using linear regulation (resistors or voltage regulator) because it gets way too hot: 12V - 2.7V = 9.3V -> *0.150A = 1.4W (similar calculations for other channels).

Instead, I'd like to use some kind of switching regulator to step down the voltage.

My potential solutions include:

1. Finding a 4 channel constant current buck regulator with 4 PWM inputs that control each channel independently. I'd use a highly resistive voltage divider on the output of the RGBW controller to control the 4 PWM inputs of the 4 channels. I do not know if a chip like this exists? Anyone?

2. Could I get a way with using a 1-channel constant voltage buck regulator + four low side switches for PWM on the output of each RGBW channel of the LED? My concerns with controlling a buck regulator module (that doesn't have a PWM pin) are listed at the top of my thread.

What do you guys think about this?
 

Marley

Joined Apr 4, 2016
470
Answers to your questions:
1/ Can you give me an example IC that has this?
2/ No. Not for the device referenced. Needs a reasonably constant DC voltage in to give a constant DC voltage out. This device probably has an internal filter capacitor on the input.

Some buck regulator controller ICs have an enable/disable input that can stop the switching action very quickly. Possibly immediately. The output voltage will take some time to drop to zero due to the output filter capacitor discharging into the output load. The switching regulation action will again take time to restart and get to a stable output voltage.
 

Papabravo

Joined Feb 24, 2006
18,102
It sounds like you want to control both the current AND the voltage on each channel. Is that correct? It may be possible to control the voltage and limit the maximum current, but AFAIK you cannot control both simultaneously.
 

Ian0

Joined Aug 7, 2020
4,863
1. Why do switching regulator ICs have dedicated PWM inputs pins?
They don't, but Constant current LED Drivers do.
https://www.diodes.com/part/view/AL8862?BackID=10012
and for a wider selection
https://www.diodes.com/products/pow...ers/medium-and-low-voltage-dc-dc-led-drivers/
2. Can you use PWM on a low side switch at the output of a constant voltage buck regulator module?
Yes. It will have a large output capacitor which will buffer the changes in current.
1. Finding a 4 channel constant current buck regulator with 4 PWM inputs that control each channel independently.
Not seen one, but there are some single-channel devices in 5-pin SOT23 packages: https://www.diodes.com/part/view/AL8861?BackID=10012

What is your power supply voltage? You may well find that you don't need the intermediate buck-regulator.
 

Thread Starter

tvist21

Joined Jun 11, 2021
20
They don't, but Constant current LED Drivers do.
https://www.diodes.com/part/view/AL8862?BackID=10012
and for a wider selection
https://www.diodes.com/products/pow...ers/medium-and-low-voltage-dc-dc-led-drivers/

Yes. It will have a large output capacitor which will buffer the changes in current.

Not seen one, but there are some single-channel devices in 5-pin SOT23 packages: https://www.diodes.com/part/view/AL8861?BackID=10012

What is your power supply voltage? You may well find that you don't need the intermediate buck-regulator.

Thanks for your reply Ian.
You're right, I'm seeing the PWM on constant current LED drivers, not CV. Thanks for pointing that out.

My supply voltage is 12 Vdc.
I've attached a preliminary schematic. Would you mind giving me your feedback?
1. I added capacitors at the output of the buck regulator (two 47uF) for each channel.

2. The bluetooth controller and the RGBW PCB already exist but I made little blocks representing each of them. The bluetooth controller has an internal 5V regulator so the bluetooth controller and the buck regulator will both share the 12Vdc input.

3. I have an LCL circuit at the input of my buck regulator for EMC compliance (as suggested by the buck regulator's datasheet). Another concern is that if I am using low side switches in my circuit, will the switching frequency of the regulator and switching frequency of my low side switches interact to produce hard-to-deal with harmonics that could find their way back to the input of the buck regulator (bypassing my LCL filter)? What about the different switching frequencies causing different ripple frequencies thus causing ripple to stack on top of each other causing the peak inductor current to potentially reach over it's maximum rating? The inductor should be rated for at least 1A and I'm only drawing 60% so I think I have some room for that kind of issue.
 

Attachments

Ian0

Joined Aug 7, 2020
4,863
Conspicuous by their absence from your diagram are the output inductors between the "drain of lowside switch" and the LED.
Small differences in the PWM ICs will cause their frequencies to differ so, whilst they may be in phase from time to time, they won't stay that way for long.
If the frequency of output 1 is 100kHz and output 2 is 100.01kHz, then they go in and out of phase ten times a second.
The 5V regulator is superfluous - just use an LED driver that will cope with 12V input.
 

Thread Starter

tvist21

Joined Jun 11, 2021
20
Conspicuous by their absence from your diagram are the output inductors between the "drain of lowside switch" and the LED.
Small differences in the PWM ICs will cause their frequencies to differ so, whilst they may be in phase from time to time, they won't stay that way for long.
If the frequency of output 1 is 100kHz and output 2 is 100.01kHz, then they go in and out of phase ten times a second.
The 5V regulator is superfluous - just use an LED driver that will cope with 12V input.
Why would I want inductors between the drain of a lowside switch and an LED?
Would an LED driver be any different than using a buck converter?
You sound smart, would you be open to a zoom call if I can't figure this out?
 

Ian0

Joined Aug 7, 2020
4,863
Why would I want inductors between the drain of a lowside switch and an LED?
Because the low side switch will switch the entire 5V power supply to the LED, and the LED will blow up at about 4V.
(Look at the voltage-current graph for the LED.
Would an LED driver be any different than using a buck converter?
I assume you mean buck regulator, because they are both buck converters.
The difference is that a buck regulator regulates the voltage, and LED drive is a buck converter that regulates the current.
LEDs must be driven with a constant current
 

Thread Starter

tvist21

Joined Jun 11, 2021
20
Because the low side switch will switch the entire 5V power supply to the LED, and the LED will blow up at about 4V.
(Look at the voltage-current graph for the LED.
But I would use resistors at the input of each channel of the RGBW LED to drop the voltage and limit the current.

I assume you mean buck regulator, because they are both buck converters.
The difference is that a buck regulator regulates the voltage, and LED drive is a buck converter that regulates the current.
LEDs must be driven with a constant current
Yes I meant the regulator (thank you for the distinction between the two).
If I were to use an LED drive that supplied constant current, could I use a single driver that supplied 0.6A so that each channel would get it's nominal 0.150A current? Wouldn't switching one channel with a PWM low side switch cause a decrease in how much current is flowing through that channel, thus cause an increase in current going through the other 3 channels (KCL kirchoff's current law) - thus potentially burning out the other 3 channels? Or does the 0.6A current remain equally distributed regardless of different PWM duty cycles on each channel?

I have been refraining from using a constant current source because I thought I would need 4 different constant current devices because of this KCL issue.

I found an the LDD-600S meanwell driver which is FCC approved that would be perfect for this if the current remains distributed.
 
Top