PWM module for CC LED driver - am I doing anything silly/dangerous/etc with this circuit?

Thread Starter

Eddy Current

Joined Jan 25, 2017
25
Hi, I have a lamp with a constant current LED driver, which won't dim on the AC side, and has no ports for controlling output. So I just made a 555 PWM controller in parallel with the COB LED to drive a portion of the current onto a wire-wound resistor load. As I've never made anything like this before, and don't have an oscilloscope, is there anything I should be worried about between simulation and reality?

Electrolytic cap is rated 25V, and the load 10W. Ceramic caps and metal-film resistors for the rest.

CC LED driver voltage range is 4-16V, output 350mA

COB LED is running at 15.8V connected alone to driver

Diodes used were 1N4148, not 1N4149 as indicated

Cheers,
Ed

*edit - helps if I upload the circuit diagram
 

Attachments

GopherT

Joined Nov 23, 2012
8,012
Hi, I have a lamp with a constant current LED driver, which won't dim on the AC side, and has no ports for controlling output. So I just made a 555 PWM controller in parallel with the COB LED to drive a portion of the current onto a wire-wound resistor load. As I've never made anything like this before, and don't have an oscilloscope, is there anything I should be worried about between simulation and reality?

Electrolytic cap is rated 25V, and the load 10W. Ceramic caps and metal-film resistors for the rest.

CC LED driver voltage range is 4-16V, output 350mA

COB LED is running at 15.8V connected alone to driver

Diodes used were 1N4148, not 1N4149 as indicated

Cheers,
Ed

*edit - helps if I upload the circuit diagram
If you have an LED driver that can drive 350mA constant current, then the 50 ohm resistor is not needed unless there is some unknown limitation to your LED driver (constant current source).

Also, the two resistors (1k resistor and pot) with the 0.1uF timing cap looks like you will have a PWM frequency of only 40Hz or so (check my math - I did it in my head). Normally those drivers want LED PWM of about 200 HZ.
 

Thread Starter

Eddy Current

Joined Jan 25, 2017
25
The 50 Ohm (minimum) resistor is to bleed current away from the LED. Otherwise there is no way to dim it. AC side dimming doesn't work with this style of driver. I tried. The CC controller will keep on outputting 350mA up to 16V. That current needs to go somewhere. The module just sits parallel with the LED and CC Driver to bleed current.

Period is 4ms, or 250 Hz.

The 470u cap and >50 Ohm resistor means the LED doesn't turn off, or ripple much.
 

Thread Starter

Eddy Current

Joined Jan 25, 2017
25
Cheers, but the PCB is already milled and soldered :) Everything works fine, just wondering if anything is a hazard.

Heat dissipation through the load seems fine.
 

djsfantasi

Joined Apr 11, 2010
6,974
Just curious?

If you were confident enough to have made the PCB, why bother asking? Anything any member might communicate has been made moot by the fact you have a PCB board made and (apparently) do not plan on making any changes to it?

What if your design was truly dangerous or silly in that it would not work? Would you still respond "Cheers, but the PCB is already milled and soldered"? Or would you respond differently?

As I said, just curious.
 

Thread Starter

Eddy Current

Joined Jan 25, 2017
25
No, I would remake it. It works fine with the resistor, so if it isn't a safety issue, why on earth would I change it? Just because I can partially design a circuit, and do some analysis, doesn't mean I know every bloody thing about it!


Moderator's Note:
Snip the insulting words.
 

GopherT

Joined Nov 23, 2012
8,012
No, I would remake it. It works fine with the resistor, so if it isn't a safety issue, why on earth would I change it? Just because I can partially design a circuit, and do some analysis, doesn't mean I know every bloody thing about it!
So, someone mentions your mode of operation and, instead of entertaining an interesting discussion, or simply ignoring him, you say "Go F#€& yourself"? Do you need a time out?
 

Thread Starter

Eddy Current

Joined Jan 25, 2017
25
So, someone mentions your mode of operation and, instead of entertaining an interesting discussion, or simply ignoring him, you say "Go F#€& yourself"? Do you need a time out?
Someone mentioned I didn't need a particular resistor. I said "cheers" -meaning "thank you" - meaning "I appreciate you pointing that out" - meaning "I will try that in future simulations". I then commented on where I was up to, so if it wasn't a problem I wouldn't be replacing it. Geezuz.

I may have responded a bit harsh to djs, but in no way did he comment on the circuit. He was just condescending.
 

GopherT

Joined Nov 23, 2012
8,012
Someone mentioned I didn't need a particular resistor. I said "cheers" -meaning "thank you" - meaning "I appreciate you pointing that out" - meaning "I will try that in future simulations". I then commented on where I was up to, so if it wasn't a problem I wouldn't be replacing it. Geezuz.

I may have responded a bit harsh to djs, but in no way did he comment on the circuit. He was just condescending.
I agree that he didn't comment on the circuit but, I disagree that he was condescending. I believe he was being inquisitive and trying to have a human to human interaction with you - a conversation. If you are not interested in conversation, ignore the question. It requires the least effort from you. If you have the energy to answer, engage the conversation and answer the question.
 

Thread Starter

Eddy Current

Joined Jan 25, 2017
25
I disagree that he was condescending. I believe he was being inquisitive and trying to have a human to human interaction with you - a conversation.
I answered his question in a civil manner. I finished my reply with something worthy of the way he finished his. I don't treat passive aggression with politeness.
 
Top