Home     Forum     Projects     Experiments     Circuits     Theory     BLOG     PIC Tutorials     Time for Science

### Author Topic: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS  (Read 4323 times)

0 Members and 1 Guest are viewing this topic.

#### gustavo

• Newbie
• Posts: 6
##### High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« on: January 02, 2013, 14:07:05 PM »
Hi evreyone:

I open this post to follow the discussion of "High Efficiency High Current LED Buck Driver using the A6210" questions. I will summarize my circuit and my problem, so people can discuss.

My circuit is the one suggested in the datasheet with my own values, extracted from A6210 design tool.It means:
• Ten Leds in series
• set of values R1:390k, R2: 75m, L1: 33uH. This means 2,6 A through the leds
• second set of values :R1:390k, R2: 150m, L1: 33uH, which gives 1,3 , in order to be more conservative, as a second attempt
• DIS pin input enable during 100 us, disabled during a 10 ms or more, depending on settings for input DIS signal

Well the behaviour of the signal output, after inductor, is not what I expected. It means that:
• When I disconnect inductor from leds string (I have a jumper there), in order to test voltage output before final test my signal is not a PWM output voltage.  I expected a voltage PWM output. After talking with giorgios lazaridis i didn't expect any particular voltage value, but I expected a PWM voltage, square wave, well defined. But what I had was  40 Volts when DIS is ON and then the voltage goes very very slowly down. The higher voltage (40V) is flat during 100 us (as I expected) but then I have is a slow down curve when dis is off. Such slowdown curve goes to Zero 1,5 ms after the pin dis is turns off (15 times more). I didn't care about the voltage value, but I do about the shape
• Then I connect the jumper( so I have the load connected) , closing the circuit, and look with osciloscope what is happen. Maybe in this way I could check what happend in the led string and see voltage and intensity (with datashhet help). BUT, once I connect the jumper, signals (voltage output form the inductor and even DIS signal) goes realy noisy, with peaks of some volts(ten or more). Leds are obviously lighting very hot. So i'm unable to check anything else, or my leds will burn out

This is my problem summarized. Giorgios Lazaridis' answer was the following:

think that you've mistaken with the dis pin. Pulsing the dis pin will not pulse the output accordingly. The dis pin will change the valley current of the output (almost linear). You need to make a simple circuit first! Very important. Like, 12-24 volts voltage with one or two leds in series. Plus!!!! IF YOU have the circuit on breadboard, all your calculations are wrong. On breadboard you can only approximate values.

So, bottom line is:
1. Do not pulse dis pin for first circuit
2. Pulsing dis pin will NOT pulse the output. Do not expect any sort of PWM output
3. Dis pin controls the output current. 10% PWM at dis pin means 10% of the current selected
4. The output current will seem to you like mountains (that is why the chip operates as valley current driver). Looking thought the oscilloscope you expect to see an output voltage of something similar to triangular waveform, pulsing for a few hundreds of millivolts around your set value

Then, I have now more questions, because I think I have missing the point, maybe because of my lenguage mistaken:
- I don't understand why I dont have to expect any sort of PWM output. If you look A6210 Datasheet you will find graphics from the osciloscope, where voltage is a perfect pwm voltage signal. Moreover, I copy some words of the related blog entry : "The name "DIS" comes from the "Disable/Enable". You may consider this pin as a switch. Leaving this pin unconnected, the SMPS does not operate and the LED is completely off" & "This pin can also be used to inject PWM pulses into the circuit and have brightness control over the LED". So, I know I missing a point, but I don't really don't know where. I know maybe is a stupid question, but I can't find where is my mistake

- So, If my my target is to pulse leds in a "stroboscopic way" (sometimes on, somtimes totally off)........ this buck driver it is not a good idea?

Thanks for everyone help in advance. Thanks giorgios for your help

Best regards
« Last Edit: January 02, 2013, 14:09:20 PM by gustavo »

#### kam

• Administrator
• Hero Member
• Posts: 1849
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #1 on: January 02, 2013, 15:03:15 PM »
Yes, you can indeed inject PWM pulses and have brightness control, but the output will not be PWM!!!!!!!

Many leds accept 3 ties the nominal current when pulsed to achieve higher brightness. That is true and i understand what you want to make. But this driver does something different. I uses PWM pulses NOT to PWM the output, but to regulate the voltage of the output so that the current will be a proportion of the max current, relative to the Duty Cycle proportion. You will never see PWM on the output... So, power your LEDs with the nominal current and not the max pulsed current.

#### gustavo

• Newbie
• Posts: 6
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #2 on: January 02, 2013, 15:55:00 PM »
Ok, I understand what you say. Thanks.

But, but, To be a little more boring this time, I've attached a plot

Looking at this plots, which is in A6210 Datasheets (Page 9 plot 4), seems to be what I'm looking for. This was another reason why I thougt I could do that with this driver.

I can see also the first two plots, what are exactly what you are telling me,so which is the diference between these plots(2-4 ) in page 9? I think the only diference is DIS pin, I im unable to reproduce it. But maybe I'm wrong

Regards

#### gustavo

• Newbie
• Posts: 6
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #3 on: January 02, 2013, 16:50:08 PM »
Another "basic" or "stupid" question.

As long as PWM "input" is optional...........Which is the difference on having a particular input pwm signal on DIS pin or not? I mean: As far as I know, the buck driver works controlling the current, keeping it constant. But also works as "valley current controller", so the current never goes below this valley current, keeping the current with high peaks and low valleys around an average value (value obtained from my own component setting choose).  This peaks and valley gives a "frecuency" to the current signal. This is the frecuency they talk in datasheet. This should be a normal behaviour with DIS disable . Am I right??

Then, what happend if I input a PWM signal? what is the difference in behaviour terms?

Thanks
« Last Edit: January 02, 2013, 16:58:21 PM by gustavo »

#### gustavo

• Newbie
• Posts: 6
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #4 on: January 02, 2013, 21:04:17 PM »
I attached 3 Pictures of what is happened

First is the signal with no load. In blue is the DIS signal, In yellow the output of the inductor.  YOu can see the time enabled is 100 us, but you can see the slow down curve out from the driver.

The second one is with the led String attached. I have try to get one led out the PCB, and leave it alone, so I can check if could be any PCB problem. But I can see that the led alone has the same behaviour (obiuosly a less noisy signal).

Note that EVEN DIS pin signal gets noisy, but I have look carefully the desgin and it seem correct

These are the images

#### kam

• Administrator
• Hero Member
• Posts: 1849
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #5 on: January 14, 2013, 09:20:25 AM »
As long as PWM "input" is optional...........Which is the difference on having a particular input pwm signal on DIS pin or not? I mean: As far as I know, the buck driver works controlling the current, keeping it constant. But also works as "valley current controller", so the current never goes below this valley current, keeping the current with high peaks and low valleys around an average value (value obtained from my own component setting choose).  This peaks and valley gives a "frecuency" to the current signal. This is the frecuency they talk in datasheet. This should be a normal behaviour with DIS disable . Am I right??
Then, what happend if I input a PWM signal? what is the difference in behaviour terms?

I am not sure what you ask here. The DIS pin has 2 functions. 1st is that you can use it as a switch, to turn on or off your lights. 2nd is that it can control the output current by injecting PWM. But the output PWM has no linear response to the input PWM.

I am not sure if i understand what circuit you want. A understand that you don't want just to dim the LEDs, you want something different. I think that you want to overdrive the LEDs for a short time, right? If this is correct, then the chip is wrong. Here is what you need: You need an SMPS or other sort of voltage regulator to achieve maximum voltage for over-driving the LED (for example 15volts for a 12V LED string). Then you wanna make a linear constant current regulator which has faster response, and drive this linear driver with your 1mSec pulses.

#### gustavo

• Newbie
• Posts: 6
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #6 on: January 14, 2013, 11:18:29 AM »
Well I will try to explain better myself

I am not sure if i understand what circuit you want. A understand that you don't want just to dim the LEDs, you want something different. I think that you want to overdrive the LEDs for a short time, right?

That is correct. I wanted to overdrive my leds  for a short time, and then let the leds off for a longer time, and then pulse the leds again, witha low duty cicle, so mi leds doesnt burn. I thought a current constant driver could be suitable to do that, if I could PWM the current in the output somehow.

You need an SMPS or other sort of voltage regulator to achieve maximum voltage for over-driving the LED (for example 15volts for a 12V LED string).
understood. I knew I had this posibility to do it, but in this way I will always have my current slightly "uncontrolled", specially when the leds are "forced " to be in high intensities of the I-V curve

The DIS pin has 2 functions. 1st is that you can use it as a switch, to turn on or off your lights. 2nd is that it can control the output current by injecting PWM. But the output PWM has no linear response to the input PWM.

I think I undestand, but then, I don't understand some graphics shown in datasheets as I told before in the post(Page 9 plot 4). I mean, looking at those plots, I see the circuit can "act" as a current switch, leaving current flows when DIS is ON and stopping the current when DIS is OFF very quicly.  This is the main thing I don't undertand. I feel conflicted at this point.

I was also thinking that if I put a simple MOSFET between leds string and driver output I will get what I'm looking for,  but I don't know if I could do that

Anyway, despite the circuit would not be able to do what I expected, I have a problem in my circuit: I dont know why the signals goes tottaly uncontrolled , even DIS pwm signal, when I conect the leds string. It is frustrating, because I'm not able to check anything

« Last Edit: January 14, 2013, 11:26:30 AM by gustavo »

#### kam

• Administrator
• Hero Member
• Posts: 1849
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #7 on: January 14, 2013, 13:38:18 PM »
Quote
I think I undestand, but then, I don't understand some graphics shown in datasheets as I told before in the post(Page 9 plot 4). I mean, looking at those plots, I see the circuit can "act" as a current switch, leaving current flows when DIS is ON and stopping the current when DIS is OFF very quicly.  This is the main thing I don't undertand. I feel conflicted at this point.
The graph states that CH1 is the current (indeed) and CH2 is the main switching voltage, NOT the dis pin voltage. Is this what confuses you?

Quote
Anyway, despite the circuit would not be able to do what I expected, I have a problem in my circuit: I dont know why the signals goes tottaly uncontrolled , even DIS pwm signal, when I conect the leds string. It is frustrating, because I'm not able to check anything
During my tests and prototypes, i burned some 5-6 of the A6210 and i can tell that as long as they are exposed to tests and soldering jobs, they are very sensitive. Maybe something bad has happened to your chip? OR maybe the input voltage needs more filtering? Do not forget that these regulators tend to add noise to the line especially when the load is high

#### gustavo

• Newbie
• Posts: 6
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #8 on: January 14, 2013, 14:58:53 PM »
The graph states that CH1 is the current (indeed) and CH2 is the main switching voltage, NOT the dis pin voltage. Is this what confuses you?
No, I feel confused because , for example, in plot 4 , i can see:
1.- CH1 output current
2.- CH2 output voltage (in Lx node) before the inductor

Both channels have that waveform because of 10 Khz PWM signal at DIS pin. So what I see in that plot is that I have a current of 1.MhZ as I expected of a current valley controller , but  I can see also that current and voltage goes to 0 when DIS pin is Off. So that particular circuit shows in the plots high current values just for a short time (when DIS is on) and then the current goes quickly to 0. This behaviour is kind of similar of what I'm looking for, because I can then overdrive my leds during a short time and then let them rest.

What I mean is that I see that I can cut off the current with DIS pin totally, and I maybe misunderstand your previous explanations, but I interpret that this can not be done whit the driver

Quote
During my tests and prototypes, i burned some 5-6 of the A6210 and i can tell that as long as they are exposed to tests and soldering jobs, they are very sensitive. Maybe something bad has happened to your chip? OR maybe the input voltage needs more filtering? Do not forget that these regulators tend to add noise to the line especially when the load is high

I will make another tries with a new allegro and diferent values.
« Last Edit: January 14, 2013, 15:03:02 PM by gustavo »

#### kam

• Administrator
• Hero Member
• Posts: 1849
##### Re: High Efficiency High Current LED Buck Driver using the A6210: QUESTIONS
« Reply #9 on: January 14, 2013, 15:50:47 PM »
I think that using this driver for your purpose is overkill, if not impractical. A simple mosfet-mosfet current driver set at the current you want with a PWM input is what you need. Maybe you want to use an SMPS to get the voltage down to your requirements and thus increase the system's efficiency. But i certainly do not think that you can switch on and off the the A6210 to get the short pulses you want.