169Views15Replies

# Dimmer Question about Amperage? Answered

I've searching for a dimmer that its flicker won't be visible. This dimmer is the one that I came up with

It claims to be 3-35V, and 90W MAX, but also 5A MAX.
Does this mean that the maximum power that it can supply is either 5A or 90W? That means that for 3V it would be able to supply 15W (5A), and for 35V it would be able to supply 2.5A (90W). Am I correct? Is it the first one that reaches MAX the maximum?

I'm going to use it for 12V. That means that I can use 5A max, right?

Thank You!

Tags:

## Discussions

I do not recommend using pulse width modulation (PWM) for driving LEDs.

PWM is, in my mind, essentially the same thing as constant voltage. I mean the time-averaged voltage of a PWM signal is essentially just the input voltage multiplied by some constant duty cycle.

For LEDs it is better to drive them with constant current, rather than constant voltage. To get constant current requires using a regulator intended for delivering constant current.

Such things exist. Such things exist for sale on eBay. Although I cannot immediately point to one, because I am not certain about the specs for your LED, or LED module, or string of LEDs, or whatever it is.

I mean, the way this game is played is you look at the specs for your LED or string of LEDs, and then ask: What is approximate forward voltage drop? What is the maximum allowable current?

Then try to come up with a constant current regulator that can comfortably supply a current less than that maximum value, at that forward voltage drop.

The way LEDs work, the forward voltage drop does not change much, over widely different amounts of forward current. So power dissipation is equal to a roughly constant voltage, multiplied by whatever the current is; i.e. I expect power dissipation to be roughly proportional to current.

Moreover, the way to implement dimming for LEDs is simply through current control.

If you want less light output, change the current set point; e.g. if the maximum is 4A of current, set the current to 2A to get approximately half the light output, or set it to 1A to get one quarter the max output.

Thanks, I've heard of constant current dimmers, but what I don't understand is this:

If I've dimmed the LED to the point where it's off, will I have 12V @0Amps? How is that possible?

The problem is that I still don't know enough in electronics to know how many amps I'll need, what resistor will be needed, and so on...

I also always put LED's in parallel, which I know is a mistake. With a CC one, I MUST arrange them in series, right?

---------------------------------------------------------------------------------

I also want to find them for cheap, and the only place I know is eBay. The problem is that many of the sellers don't provide enough information, and my Chinglish isn't that good.

If I have a 12V 10A power supply, want to drive 5 10W LED's (4 cool white, one warm white), I can't put them in series. If they were all cool ones, I would have to bring up the voltage to 60V, and then limit the current. I'm guessing I would a Step-Up Constant Current Regulator? Is it something like these, except for the fact that they don't reach 60V?:

http://www.ebay.com/itm/250W-10A-DC-Step-up-Boost-...

http://www.ebay.com/itm/HOT-5A-Auto-Step-Up-Down-R...

Circuit elements wired in series, share the same current.

Circuit elements wired in parallel, share the same voltage.

The reason for wiring LEDs in series, is this is a simple way to guarantee each LED in that series string has exactly the same current.

The goal after all, is to build a circuit that makes sure each LED has some constant, predictable, current flowing through it.

The fear with wiring LEDs in parallel is that they might not share the current equally, in particular there is fear caused by the fact that the current-voltage characteristic of LED is temperature dependent, and maybe also some fear that LEDs are not exactly alike in their current voltage characteristic, even if they are same part number and sisters from the same batch, born the same day, although that is a lesser fear. I think the main fear is regarding temperature difference.

Specifically, for the case of just 2 identical LEDs in parallel, there is fear that one of them (e.g the left one) will be slightly warmer than the other, and this warmer LED will take more than its share of the current (i.e. more than half) and maybe get even warmer still.

Of course, taking current away from the right LED (assuming the currents through both have to sum to a constant) should cause its voltage to drop, and thus the shared voltage of both to drop, so that's kind of a stabilizing influence.

So it is not the kind of game where all the current flows through one LED or the other, making one dark, the other twice as bright. Instead I think there is going to be some sort of equilibrium where the warmer LED is just somewhat brighter.

By the way, you might have noticed LEDs wired in parallel in other people's designs, even in mass produced devices found in the wild, that you kind of suspect are well-designed. I mean how dumb could a mass-produced design be?

As an example, your 10 watt LED module itself may consist of 9 smaller LEDs, as 3 parallel strings of 3 in series, and the story I have heard is that they can get away with this because all those LEDs are so close together, on the same die, so that all these LEDs have almost exactly the same temperature.

However, if you wire several LEDs in series, then you know for certain that there is going to be exactly the same current flowing through each LED. You know, things wired in series share the same current. They kind of have to, as an unavoidable consequence of Kirchhoffs circuit rules.

Anyway, kind of a long story, but I suspect this is why designs with LEDs in series are favored, and designs with LEDs in parallel are frowned upon.

Regarding the eBay converters you linked to, I kind of like the first one,

http://www.ebay.com/itm/250W-10A-DC-Step-up-Boost-...

and I think it could comfortably drive 4 (but probably not 5) of your big 10 watt LED modules, assuming the forward voltage drop across each is about 12 volts, and the drop across all 4 in series, 4*12=48 volts.

By the way, without modifying the power converter board, the method for dimming, or conversely bright-ing, this string of 4 LEDs would be by turning that little screw on the pot that adjusts the constant current set point.

And that method, turning the tiny little screw, might or might not be what you had in mind for your dimmer. I mean, to adjust it more than once or twice, it would be nice if it were a knob big enough to turn with your fingers.

Another nice feature would be if the knob stopped turning at a point before over-currenting the LEDs. With the existing little trim pot on the board, you have know when, erm rather where, to stop turning the screwdriver.

Replacing the little trim pot on the board, with a larger finger-turnable pot that stops at an angle before the LEDs burn, would require some skill.

In the little diagram I drew of this setup,

https://www.instructables.com/file/FUNXT1PISAUPKU8/

I imagined there would be an ammeter in series with the LEDs, that way I could stop turning the screwdriver when meter reads 800 mA, or 0.8 A.

Actually there are some tricks to setting up these little single board, eBay bought, Chinese made, power converters, the most important of which is to make sure it is working the way you expect before connecting your expensive LEDs; i.e. get its set points adjusted correctly first, using some other, similar load, like a big, appropriately sized, resistor.

In answering a previous question, here,
https://www.instructables.com/answers/Led-buck-conv...

I set up a eBay-bought, CC-CV regulated buck converter, to power a single 3 watt LED, with constant current, at about 3.0 volts at 0.3 amperes, or 3.0*0.3 = 0.9 watt of power, and also took some pictures of this.

I link to this because it is a setup somewhat similar to the thing you want to do. Although, an important difference is the thing you want to is a power level of about 50 times greater, around 50 W compared to 1 W.

Thank you so much! It probably would have taken me a month to type all of that!

I think that in massed produced items, the companies are looking to save a tenth of a cent... Why would they care if an LED on a child's toy would last 500 or 5000 hours?

I'm planning to place the LED's on the biggest heat-sink that I can find, and I might also add a fan.

To stop yourself from turning a knob too much you can do it either with a resistor (I think) or mechanically, in a way that just stops it...

Think about it. What is a switch ? Across its pins you have 12V, through its pins you have 0Amps.

A diode is a switch. It has a very low resistance when its on, and a high resistance when its off. The switching action is the polarity of the applied voltage AND the voltage in the forward direction. If you put 11 V on a diode that switches at 12, it is OFF and 0 current flows (this is an approximation of the truth, before I am corrected)

The key parameter is called the diode forward voltage drop, and its the number you need for an accurate resistor value when calculating a suitable one.

Now THAT makes sense. Only the first paragraph :)

The CC stuff is a bit too complicated for me now, I'll need to learn quite a bit more until I reach that level... I think I'll use a regular PWM dimmer, unless someone tells me that this is a horrible decision.

Thank you!

WHAT LED's ! If they have no current limiting resistor a PWM dimmer will NOT work.

I know this is a really simple questions for people who know. I just want to make sure...

I also might be able to get a bit more if I keep the dimmer cool...

I also found this dimmer, so I think it will be better

Although I see the virtual rocks flying in my direction already:
A PWM dimmer in the simple form does not provide an average voltage to the LED.
It flashes the LED at the selected voltage too fast for the naked eye to see.
Through the pulse width modulation the on-off times vary and if the off time is much longer than the on time the LED appears quite dim.
If the on time is almost 99% or even full the LED will light up at full brightness.

A simple but effective LED dimmer could work like this:
PWM generator, adjustable to control the brightness.
Constant voltage, constant ampere regulator, like LM317 combos for example.
Regulators are set to what the LED(s) need. e.g. 3.2V and 30mA.
The PWM generator now uses a simple transistor to switch these set values to the LED.
So in no case the voltage or the amps will be higher than the set values.

LED's are quite sensitive to overcurrent but can tolerate overvoltage in a PWM controlled situation.
This mean with slightly higher voltages than required for a LED you can use less amps to get the same brightness - but only to about 60% duty cycle.
IMHO the better option is to select LED's with a higher brightness than required and to drive them at lower levels.
This increases lifetime by several factors compared to using the stated values.

A LED dimmer works by chopping the input voltage into small segments - The more the voltage is chopped the lower the overall brightness (Pulse Width Modulation - look it up).

This means that the full current will be available all the time its just that the full voltage is not. On average the power supplied is lower but at any 'on' part of the cycle the full power is applied.

Yes- That's also what I'm confused about.

I know how a PWM dimmer works (at least I think I do) If I have a 1Khz dimmer, and I want to dim the LED's brightness to 90%, every 10th of that the dimmer's cycle will turn off the LED to the point where it reaches 90% of its brightness. Something like that, I think...

Neither of those looks to me like a constant current LED driver, whatever the blurb says. You will still need a big fat current limiting resistor on a naked LED, or your LEDs will die instantly.

What LEDs are you driving ?