LED Dimming: PWM vs Current
HTML-код
- Опубликовано: 3 сен 2015
- In this video, I demonstrate two different ways to dim an LED, along with electrical analysis of the circuits using a multimeter and an oscilloscope.
Skip to 9:39 for the bit about Arduino PWM.
3d printed lithophanes: www.createthis.com/products/li...
Here is the Arduino gist, as promised in the video: gist.github.com/anonymous/cfb...
These are the products I use or mention in the video:
LED light strip (not the one in the video, but similar and less expensive - this is what I would buy today, personally): [amazon affiliate] amzn.to/1IOkHBN
Arduino Uno Ultimate Starter Kit -- Includes 72 page Instruction Book: [amazon affiliate] amzn.to/1N7pVyB
SainSmart UNO, ATmega328P USB CABLE Included for Arduino: [amazon affiliate] amzn.to/1IOkiz6
Fluke 87-V Digital Multimeter: [amazon affiliate] amzn.to/1JGSwIG
Extech EX430 True RMS Autoranging Multimeter with K Type, Capacitance, Frequency, and Duty Cycle: [amazon affiliate] amzn.to/1N7qb0D
RIGOL DS2072A: [amazon affiliate] amzn.to/1N7qf0p
Rigol DS1054Z: [amazon affiliate] amzn.to/1IOkBtR 
Хобби
![Roddy Ricch - Survivors Remorse [Official Music Video]](http://i.ytimg.com/vi/J5IeFzr2GgE/mqdefault.jpg)
I have absolutely no experience of electronics but your tutorial is fantastic. Very well explained and easy to understand - if you're not a teacher, you should be!
way to go my friend, teaching your kids is very good, and you sound really excited with your new toys, I also have the same model scope ( mine is software enhanced ) it is a great scope for the DIY people
Great video, which has led to interesting discussions in the comment section. Who cares if you’re not a professional engineer - the video got me thinking and understanding better. 👍🏼👍🏼
Very good. Very explanatory. Very interesting. Thanks.
Cool ! Is an Arduino Nano able to do the same? This red led is 2v. How does this Arduino support a much higher voltage LED ?
The (mis)pronunciations of potentiometer and multimeter make me chuckle.
I thought he video was excellent. I do have 1 question. I have a capacitive touch sensor that I would like to drive a digital potentiometer or pcmcia dimmer thus a touch dimmer for led lights. are you aware of any components ? Basically when you touch the sensor the dimmer would increment up to a set value and then reset
Call me crazy but I prefer current limiting. I realize it's much less energy efficient but I hate the flicker from the PWM modules, you can especially see it through cameras.
+TheLegendBullet Fluorescent bulbs always bothered me (flicker), but led lighting is fine to my eyes.
Flicker bothers me a lot, too, but if you get the frequency high enough, it isn't noticeable. Diffusers also help.
if you wave your hand infront of a pwm-dimmed ledstrip you might see stropo-effects, therefore i want to use current limiting myself.
Especially important in TV/broadcast applications - when the lights are out of sync with the camera shutter speed, artifacts and flickering occurs.
BTW: current limiting can be very efficient - but you should use a true constant current regulated switching power source in such applications, not a resistor in series. It can be relatively easily built off a well-known LM2576-ADJ with a bunch of driving circuitry and external regulation (even by means of PWM - you can basically integrate a PWM signal (using a simple well-tuned RC LPF) to create a reference voltage compared against an output shunt voltage and further driving an LM).
This approach is used on some professional regulators, like Netled (an Ethernet-controlled smart 6-channel LED driver).
@sgfhk321 PWM is like switching electricity on and off very fast.
You could use an buck boost converter with current limiter. You don't require any more components then that converter+you get more operating voltage with less battery voltage. If you are lazy go with an buck boost or step down converter, both are available with CC/CV.
Triac dimmers are using PWM ! They modify how long the bulb stays on during the cycle, it's pulse width ! And you can use a switching power supply to control the current of an LED or anything without wasting a lot of power.
If i use pwm will it be visible when I record it. Like recording with a 60hz camera somewhere where there is 50hz current (flickering)?
+Michael Wiedemann I don't think so, but it's possible.
This will definitely happen. PWM dimmed lights fuck with cameras really bad.
Indeed. Recently, I even saw it happen in a studio-recorded TV show (where you'd expect they know what they are doing and choose the right light source). Very bad PWM flicker from the LED lights in the studio setting when dimmed.
After playing around with Arduino (and other boards), I actually found out that you can turn your PWM frequency up so much that it's not nearly as bad on camera.
Just avoid frequencies in the audible range. ;)
You can pass the PWM signal through a simple RC low-pass filter... typically you'd already have a current limiting resistor on the pin output so all that's needed is a capacitor tied to ground after the resistor. The higher the capacitor value, the more flicker that will be filtered out. You can repeat the RC-network for a steeper-cut-off.
The only other factor is the response time (only reason not to use the highest value CAP you can.) If you want to do strobe-FX if your low-pass filters at too low a frequency say, below 30Hz or so, you would also filter-out intended flicker.
I am trying to make a dimmer for 120v at 50Hz Flicker free. Any advice would be appreciated because I am stuck.
It's for quasar science 2ft crossfaders from the US and I am powering them with a step down from 240v to 120v at 100W but still at 50hz.
The manufacturers Quasar Science only recommend Lutron CL LED dimmers but they are all 60 HZ.
I've looked a Drill speed controllers but it's the same problem and they only offer brightness adjustment over 50%
I think people may still be able to detect 50hz in certain circumstances. There was a reddit thread about this recently on a gaming forum. Gamers are using 144hz now. VR headsets tend to use 90hz and above. I think if you want the flicker to be less noticeable you should try increasing the frequency.
@@createthisdotcom I meant that 50Hz is what comes out of the mains 240v AC and in America it's 120v 60hz AC.
A line ( series) Triac or 2 anti parallel SCR's is bare minimum that gives good control but not linearity... also no guarantee about flicker..... however if the led element has to be in series,
with mains, another option ( d.i.y.) is a full bridge rectifier also in series followed with a
high frequency pwm 1kHz-50kHz controlled switch( across +/-)..
** caution here..LIVE CIRCUITRY...
NEED FOR ISOLATED switch driver logic..etc.
You could also see the difference using an Ocsil-eye-scope.
Could eye?
Buys a Fluke 87 but doesn't know why a potentiometer has 3 leads....whats wrong with this?
I'm not an electrical engineer. I've read a basic electronics theory book and played with circuits as a hobbyist, but I'm not a pro. I learned early on to always buy the highest quality tools I can afford. It just saves a lot of time and frustration. For the record, my field is software.
+createthis i understand what you mean but thats like buying 3k$ gaming PC to do simple tasks like word or web browsing
+Omar Carrero the poor man pays twice
Why the Rigol then? By that philosophy you should have a Tektronix scope! Just kidding! Kinda... As much as I love quality tools and always try to buy the best, I could never justify the cost Fluke 87 for even an advanced hobbyist. Most of us need more than one meter to measure different things at different places in a circuit, so owning ONE 87 makes less sense to us than say TWO Extechs or other cheaper meters.
If you have the extra dough and want a amazing, over the top meter its worth it, but you DO NOT need to spend that much to get a good, LONG LASTING meter.
You are not going to pay twice just cause you didn't buy an 87. Don't want newbies getting put off by that comment.
Good video though, and keep it up!
You're right, I did skimp a little on the Rigol. I don't use it very often and I figured it would work for 99% of what I need. So far it hasn't let me down. I can always apply the hack if I need a little more power. I've been wanting my own oscilloscope since 1999 when I went to college and was trying to build my own laser tag clone using a 555 timer. I was pretty happy to be able to afford one finally. It sits on my shelf most of the time, but that's fine because it's there in case I need it.
I'm a pretty big fan of the Fluke 87. I've bought an Extech first and now I use it as my second meter. The Extech is ok, but it let me down one day so I bought the Fluke. The Fluke is a very nice tool. Every time I need a meter and I grab the Fluke I'm happy I bought it. It just exudes quality. That's generally how I want to feel about all of my tools. I don't really care too much about the price years later when the cost is forgotten. I just don't want to grab the tool and find myself wishing I had a better one. That's the worst.
I also have a Fluke 376 FC that I rarely use. Talk about pricey.
Financially, I ease the pain by making these RUclips videos. I don't *need* any of this stuff. I'm a programmer by trade. All I really need is a laptop and my brain. However, I enjoy exploring the world through tools and techniques, so I make these videos sharing what I learn and recoup some of the cost. It's not for everyone, but it works pretty well for me.
Have a good one!
I want to have a long line of LEDs powered by single input PWM.
What kind of problems could you face with this approach?
@createthis
That's pretty vague. Those LED light strips are basically what you describe, however, and PWM works fine with them.
Yes. I mean if we use different types, what could be potential issues?
It is just a general question!
Way too broad a question for me to answer intelligently, sorry.
the simplest current source in electrical engineering is the series resistor. Current is the "y" value of the function given by ohms law, the causes are voltage and resistance (independent variables) or in other words, what you can control in the system. The current would just be the response of the election you made. Now talking about different types of leds in series, the most powerful leds need more current than the less powerful ones, this is true for current technology (not to mention voltage, but this is more complicated). If you have them connected in series, the current would bottleneck to the least powerful leds since if put them under the current needed to power the most powerful one in the system, it would burn out quickly since it cant dissipate that amount of energy. You need to make a parallel connection so you can control the current independently to each led or led types.
i forgot to say that power leds cant be used with series resistor as they suffer from thermal drift and consequently from positive feedback gain, the more heat it generates the less voltage it needs to operate therefore more current and into this loop until it brakes, the currrent needs to be controlled in a more sophisticated way.
Thank you
Great video. Ignore all the haters. I appreciate that you took your time to do this.
does pwm also work on motor base systems or just light like dimmers
What kind of motor? I don’t think it works on brushed DC motors, but brushless DC motors require a more complex form of PWM in order to operate. Stepper motors also use pulses.
pwm WILL WORK for brush motors too ....in fact easier to apply...there's no exception. 👍
when i dimm with current i mostly just use a transistor (and the resistor to the base of course)
PWM causes many folks a great deal of eye strain and severe headaches. I'd rather use more energy and not hurt my eyes and endure the headaches. PWM in the case of a CCFL back lit monitor is that the bulb is not instantaneous on and off like the LED, but is still glowing in the off cycle, thus greatly reducing the detriments of LED PWM back lighting. If LED is necessary (newer laptops), I'll take a small amount of lesser battery time for DC controlling versus the awful PWM controlling. Where folks get so worried about energy savings in a vehicle, and thus PWM runs rampant is ridiculous.
It is pronounced po-ten-she-OM-eter. The the outside terminals are connected to the ends of the resistor and the center terminal is connected to the wiper. The name refers to potential or voltage and is normally used as a variable voltage divider. It is connected across a voltage source and the load is connected to one end and the slider so it can feed a percentage of the total voltage from zero to 100%.It .When you only connect one end and the wiper and put it in series with the voltage source and the load, it is being used as a rheostat, which is a current limiting device. The do sell rheostats, which only have two pins.. Potentiometers were often used to control audio volume. The audio from the source is connected to the top terminal and the bottom terminal, which is ground The slider and ground are connected to the amplifier, so you can feed a percentage of the input signal to the amplifier. On the other hand, the float in your car's gas tank adjusts a rheostat, hooked in series with the battery and the gas gauge which is actually an ammeter. As the gas is used the float drops, uses more of the rheistat and less current flows through the gas gauge/ammeter.
Thank you. I will attempt to correct my pronunciation from now on and I appreciate the additional information.
Potentiometers as in voltage potential..
'
cute pretty baby on the picture..
the dimmer control is better and easily than full switch
what will happen when you turn up the voltage and keep duty cycle at 20%.
Since the current trough an led rises exponentially with the voltage you would quickly fry it. If with a very low duty cycle, lets say 1us pulses, if the current is momentarily to high, the led will be damaged even though the apperent brighteness will be very low
@@nielsdaemen ok, if it's all about watts, if I lower the duty cycle but equally raise the voltage. That should be another way of making the led brighter while keeping a low duty. I notice the faster I strobe an led, The dimmer it gets do to time. One idea is to increase the current to allow the led to get near it's full brightness in short rising edge trigger flash. Any thoughts or recommendations.
Leds are current controlled devices...& it's current can be controlled only by sensing/measuring it and using it in a feedback loop..
The led current being controlled must be within the device " safe operating range"
I thought leds where highly sensitive to there voltage rating and even small over under voltage was bad for them?
LEDs have a voltage requirement expressed as fV. Red, orange, yellow, & 570 nM green usually have an fV of 1.8-2.2 volts while white, blue, & 525 nM green have an fV of 3-4 volts. The LED REQUIRES a minimum voltage to make it light up. LEDS also have a MAXIMUM constant current rating which if exceeded will shorten the life. 5mm & 3mm radial leaded LEDs typically are rated at 20 mA max. A simple circuit would require a current limiting resister whose value and wattage rating are determined by the supply voltage and current rating of the LED. There is also a maximum current rating which can be used when PWM dimming is used whereby the LED is rapidly switched on and off which allows the heat to be dissipated. Hope that helps.
I am very sensitive to PWM flickering, too.
I can see it in almost all Samsung displays (no matter if OLED is used or LCD with regular LED backlighting), but also Philips Hue LED lights. For Hue lights, it is just acceptable, but still noticeable when I'm tired or moving quickly.
That must really suck.
It kind of does, since it looks like even big manufacturers do not really seem to care about flicker-free lighting.
I think that from around 1000hz, I do not really notice PWM any more but, for example, Samsung seems to be around 250 hz for the S7 and that I can notice very clearly, especially when someone is moving with his smartphone in my peripheral vision.
Luckily, there are also enough light manufacturers (even for nicely warmer dimming LED lights) that seem to do the job very well and also enough screens that have good PWM (or using the current method). But I can not influence what other people use in their houses or what cars they drive .
@@EymertVersteegt What about the 50Hz in most of the household lights?
@@RubenvanLeeuwen It depends on the type of household light. The old-fashioned bulbs or (eco-)halo bulbs also are said to 'flicker', but they have a short 'after glow', which makes the light look 'static' to the eyes.
As for LED lights, there really are -a lot- lights available in stores that have visible PWM flicker. I really try to avoid those. Luckily, there also are a lot of lights for sale that either use another method for dimming, or use PWM at frequencies high enough to not really bother me, or they do use PWM but in stead of a 100-0% duty cycle, there are some (very dim) led's that are always on, making the PWM effect less dramatic.
When non-dimmable LED household lights flicker, they are just blatant bad quality (you see it often in those uber cheap christmas tree light strings, where they either use very high resistant leds (about 200 of them) at immediately 230v, or they have a power brick that brings the voltage down to 12v, but then don't 'rectify' the current AND do not use a capacitor to get a more stable 12v. Same goes for retrofit lights. Some very bad specimen out there, with no rectification and no stabilization.
The potentiometer Only reduce the Voltage Not the Current. Whats is happen, when you give more voltage with the potentiometer they go brighter and they Need more Current , That is what you see on your meter. The current is not limited
by the potentiometer.
You can test this if you like by set the potentiometer for example on 8 volt With 10 leds , they need for example 200 mA
Let the potentiometer on 8 volt and place 10 Leds more , now you will see 20 leds Need 400 mA and they will get 400 mA
from the potentiometer. The volt stays on 8 volt.
Sorry for my bad English I am from The Netherlands :-)
+Kees groot A potentiometer varies resistance only.
But by Ohm's Law, as the resistance goes down, the current thru the pot goes up. As the resistance goes up, the current thru the pot goes down. Voltage across the pot depends on the rest of the circuit.
Correct. There are experiements showing that CCR dimming are much more energy efficient than PWM.
The LEDs don't *need* more current. LEDs are nonlinear electric loads, with the I=f(U) curve similar to positive exponential half. When you put higher voltage across it, the current will also lift up, because the point of operation would change. And that's how current limiting works - basically pulling down the supply voltage when an overcurrent is sensed. And the brightness is determined by current itself, so pulling down current reduces the brightness.
You lost me at: I dont know why a potentiometer has 3 leads
As one side turns off the other side turns on
You could also get a variable voltage output.
Thanks isnt it multi - meter?
+vvdv pronunciation? Mul-Tim-iter is proper, but who cares?
Ahh I have been calling it a multi - meter my whole life. Thought that made sense as it measures a multitude of things. I was checking out your video researching lights for plant growth. I wonder how PWM affects plant growth? I would assume it would be negative.
I doubt it matters. Plants mostly use chemical processes in photosynthesis. I doubt any part of that process cares whether the light comes in pulses or is constant. It's the quantity that matters. Just my personal theory having researched ATP and Photosynthesis a little bit. I really know next to nothing about plants. I'm curious what you find.
Hi there you are right to call it a Multi-meter, but also createthis is right to call it mul-tim-iter it is the same as Sodder or Solder, in Europe it is called a Multi-meter, because as you say it measures multiple things.
RIP RadioShack
Sorry, but a little concerned about the statement, "I don't remember why a potentiometer has 3 terminals..."
Also, there are a number of statements about efficiency that are not quite correct.
At equivalent brightness, PWM or DC driving will be roughly equally efficient.
The difference in efficiency comes from the actual current being fed into the LEDs rather than the fact that the PWM has "off" times, and the source of the difference has to do with the efficiency characteristics of the LEDs (which is whole another topic). Said another way, DC driving is just PWM with 100% duty cycle, so, 50% PWM is roughly equal to driving DC at 50% of the PWM's "on" current.
In fact, if you drive the LED very hard, PWM will be less efficient than DC because you have to drive the LED harder (during on time) in PWM than in DC driving!
If you are "generating" heat during DC and "wasting" energy, you are also "generating" heat and "wasting" energy during on time of PWM. The problem is because you drive the LED harder (by 1 / dutyfactor) you may in fact be generating more heat during PWM. You need to understand LED efficiency vs. current to talk about efficiency in PWM vs. DC.
I’m not an electrical engineer. Your concern is warranted.
Concern troll. And your comment does not make any sense.
@@jamescole3152 Yes, I can understand this kind of stuff is very difficult for you. Study harder.
Sorry I deleted my comment a few mins after I posted it, but I still got your reply notification. Felt my comment was a bit too harsh.
You seem a little new to electronics. But atleast you're getting things to work, rather then merely being an observer such as myself.
My initial rude comment and perhaps a few others are referring to the same thing. You put too much confidence in what you say and sound like an expert when you're clearly not.
Three things:
1. I dislike this video intensely. It is very low quality compared to my normal work. I leave it up in the hope that someone finds it interesting and/or learns something.
2. I’m highly disturbed by the general public’s assumption that anything in a video must come from an expert source. I don’t know where this attitude comes from. I look at a random RUclips video and I think “here’s a guy like me who made a video about something he’s interested in”. However, people like you see a video like that and seem to think “this person is an expert because he made a video” and then you get disappointed if I make a error. I don’t understand where that attitude comes from and I consider it dangerous. Scientists don’t look at each other’s work and think “this person is an expert I bet they’re correct”. They immediately try to find fault or flaw in the other person’s thinking or procedure or results. THAT is the attitude you need when viewing a video, whether it be RUclips or national TV or from the Pope. Skepticism.
I’m not intentionally trying to deceive anyone. I’m here to learn. If you found a flaw in my procedure, logic, or anything else, I want to know about it because that’s the only way I can learn.
3. Shortly after I made this video Smarter Every Day made a video about persistence of vision. It communicates the point I was trying to make in this video much more eloquently. I suggest you all check it out. Sorry this video missed the mark. All I can do is try and improve with time.
Flicker is good for the environment and okay for you.
Cringe worthy
Yeah, I feel that way about most of my videos. What bothers you the most?
400$ "Mul-timitter" to check a wrong connected pot.
Stop doing videos about things you don't know about.