Selective Dimmer for Two Light Sources

[Or Entity?]

A "selective" dimmer

Hi!
I need some help to solve a problem I can’t get my head around. I would be very grateful If someone could help me to get started.

I need a circuit for a dimmer connected to two light sources. When the rate of change of the current is within a pre-determined value, I want the system to behave as if the two light sources where just parallel connected to a power supply through the dimmer. => Brighten and dim equally.
But when the rate of change of the current in the “positive” direction passes this boundary I want the current to be shifted to make for example Light source 1 to shine brighter than Light source 2.
In the same way brightness is shifted to Light source 2 if the rate of change is passed the boundary value in the negative direction.

To make it clear…
If I slowly twist the dimmer in any direction both lights turns equally brighter or dimmer. But if I make a sudden move in one direction, the brightness is shifted towards the corresponding light.

I am pretty sure it can be done with Inductors, but a solid state semi-conductor circuit is preferred.

 

[berkeman]

Hi!
I need some help to solve a problem I can’t get my head around. I would be very grateful If someone could help me to get started.

I need a circuit for a dimmer connected to two light sources. When the rate of change of the current is within a pre-determined value, I want the system to behave as if the two light sources where just parallel connected to a power supply through the dimmer. => Brighten and dim equally.
But when the rate of change of the current in the “positive” direction passes this boundary I want the current to be shifted to make for example Light source 1 to shine brighter than Light source 2.
In the same way brightness is shifted to Light source 2 if the rate of change is passed the boundary value in the negative direction.

To make it clear…
If I slowly twist the dimmer in any direction both lights turns equally brighter or dimmer. But if I make a sudden move in one direction, the brightness is shifted towards the corresponding light.

I am pretty sure it can be done with Inductors, but a solid state semi-conductor circuit is preferred.

Welcome to the PF. What are the light sources? Can they be LEDs driven with PWM DC voltatge sources? Or do they need to be AC Mains powered incandescents? Do you have experience programming microcontrollers (uCs) for this type of function? Or, do you have experience designing CPLDs or FPGAs for this type of function?

 

[Or Entity?]

Thanks berkeman.

Cant say I have much experiece at all. Know some basic electronics; but I am a quick learner!
Had to search wikipedia a couple of times to understand your post..
Didn't think it had to be that complicated..

Sure it could be LEDs

 

[berkeman]

Complicated is a relative term... For a person who has uC evaluation kits laying around, and is practiced at whipping out projects with them, this project would just involve some C code. Or for a person who is practiced at building CPLD-based state machines and logic, and has a CPLD evaluation board laying around, it would only involve some VHDL or Verilog code.

But if neither of those things fit, then no problem, there should be some ways to build what you want in a fairly simple way. Driving LEDs definitely helps simplify the whole power supply thing.

Since the rotary control will most likely be a potentiometer, and you want the circuit to be simple, I'll assume that we won't digitize the pot output voltage. You could build the circuit with about 3-5 opamps and a couple transistors, I think. The first opamp would be a buffer for the pot voltage (call its output V1), the next opamp would be a differentiator stage, to give you a voltage that is proportional to how fast V1 is changing (call its output V2), and the next two opamps would be used to convert V1 into a drive current for each LED, with an error term from V2 either added in or subtracted out of that LED drive. Make sense?

But even this solution involves some basic opamp circuit familiarity (or willingness to learn). Have you studied opamps at all?

Maybe others reading this thread would have even simpler suggestions... ?

 

[berkeman]

Now that I think about it more, there may be an even simpler solution with just a few transistors and discretes...

But I need to ask what the context of your question is. If it's for a school project, then we are limited in how much help, and what kind of help, we can give. For school projects, you must do the bulk of the work. What is the context of your circuit question?

 

[Or Entity?]

berkeman,
If this was a school project I would probably have solved it myself. I made the post because this isn't really my field. I study physics at the University of Stockholm; and can probably say some t the electromagnetic foundation of this stuff; but electrical engineering, circuits and stuff is somewhat a different path.. But as I said, I am a quick learner so; hit me!

I'm looking for a way to build a desktop lamp with two groups of LEDs where I can control the color temperature and brightness with the same dimmer. Its just a potential hobby project.

 

[berkeman]

Fair enough. But maybe I misunderstood the use of the adjustment velocity. Do you want the relative brightness to only be different at the time of the adjustment (and dependent on how fast the control is turned), or do you want it to be persistent, where the difference is held after the control is stopped?

 

[Or Entity?]

I want the shift to be persistent.. till the next {faster than predetermined value}-movement.
So for example with a slow twist I brighten up both groups of LEDs (relative brightness unaffected). With a quick twist the relative brightness is shifted to the corresponding group of LEDs.

Now let's say that one group contains white/blueish cold colored LEDs and the other group contains LEDs that are warm white; then both intensity and the color temperature of the light can be controlled with the potentiometer.

 

[berkeman]

Well, for persistence, I think you will need to digitize the control signal (potentiometer signal), and control the LED brightness with either a uC or a CPLD circuit. I think the easist one for you (and your thirst for new knowledge and expertise) would be to get a uC evaluation board, and learn to program it in Assebly, Basic or C. We're discussing uC controlled projects in this nearby EE thread:

https://www.physicsforums.com/showthread.php?t=307526

.

 

[Or Entity?]

I think the easist one for you (and your thirst for new knowledge and expertise) would be to get a uC evaluation board, and learn to program it in Assebly, Basic or C.
.

..A project for this summer then!