Using Arduino to control 3 LED strip lamps for video recording

Other Hardware LEDs and Multiplexing
1

Hi, I'm looking for a solution to get better LED strip dimming for video application. I currently use this dimmer on a 24v LED strip:
https://www.aliexpress.com/item/32873844140.html?spm=a2g0s.9042311.0.0.f3f64c4dQZFK0m

But this causes terrible banding in my videos:

(I turn the camera around to show it's bad in both orientations)

Is there an easy circuit I can DIY to make a dimmer that uses a better dimming method?
Eventually I want to make three different lamps with the LED strips, so they need to be dimmed separately. I was thinking maybe an Arduino could help doing this, I would have to add three trimpots, I can reuse the analog trimpots that are in the dimmers I bought from Aliexpress.

I did do some research, but I got stuck there:
(not allowed to post more than two links sadly, making my story incomplete, anyway)

"Our recommendation is to utilize LED and DMX dimmers that have sufficiently high PWM frequencies of 25,000 Hz (25 kHz) or above. By utilizing a truly high frequency PWM dimming method, even high frame rate and high shutter speed applications can be operated without any detrimental effects of flicker."

From what I gather the Uno has enough capacity to reach 25 kHz PWM, but I don't know where to go from there. Do I have to build a circuit with a MOSFET? Will the MOSFET be fast enough to be turned on and off with the 25kHz frequency? And how do I make sure I don't damage the LED's?

Could it be possible to actually hook up the arduino to the circuit inside the dimmers that I bought? I can supply some imagery of the insides of these dimmers if needed.

I am not new to electronics, I have built a lot of stuff already, but I'm not skilled enough to come up with my own circuits yet. I hope someone has some pointers for me.

Thanks for your time so far,
Marinus

2

Yes it has the capacity do do the PWM at 25kHz. The MOSFET will definately be fast enough do do what you want, some of them can switch in the MHz range. A schematic would be much better, pictures are hard to figure out as are frizzy drawings. Here is a simple circuit from Stackexchange.com. This circuit is used a lot, R2 is connected to the GPIO as it should be. If it were connected to the gate it acts as a voltage divider decreasing the voltage to the gate of the MOSFET. It is very possible you can change the PWM frequency of your existing unit by changing a capacitor internally to it. The ones I have looked at used a NE555 as the PWM controller with a cap and resistor controlling the PWM frequency.

image

1 Like
3

Do note the absolute requirement to use a true "logic-level" FET. You may need to research a little to fully understand what that means. :grin:

4

Thanks for the useful comment, I was planning on using the IRLB8743 MOSFET. I have checked out the chart for it and it seems it's a logic-level FET:

5

I have opened up the case and took a few photos. It's a fairly simple board with indeed an NE555 timer.

It is however all SMD components, I'm not entirely sure if I can add a cap to change the timing, it seems you know more about it than I do. Would you have a link to the Stackexchange by any chance? I see this circuit in the image is used for 12v, do you know if it can also handle 24v?

Here's an image of the PCB in my dimmer:

IMG_20211005_153232-small
IMG_20211005_153232-small1116×628 118 KB

6

It is rated at 24V, 8 Amps so it should be ok. Here is a link to electroschematics that goes through a tear down of one similar to yours. The schematic was labeled from them as well. The schematic I posted is probably using a zener instead of a voltage regulator as shown in the schematic. I think it is the thing with the blue band. Hopefully this works.

image

1 Like
7

Also known - in a different package - as an IRLR7843 - note the digit swap!

Well, you need to remove the present one and replace it with a smaller one.

image
image918×581 16.2 KB

C2, presently 100 nF would need to be reduced to say, 10 nF. How well the FET, driven by the comparator with a resistor pullup would cope, I cannot say.

2 Likes
8

I have no idea what resources you have but if you need to purchase a soldering iron get a temperature controlled one that you can change tips. Be sure it is at least 50W. The extra watts will be there when you need them. I had a 30W unit and after getting a bigger one I trashed it. Extra tips and sizes always help.

1 Like
9

Thanks a lot for the useful comments, I managed to 'fix' the banding (only slightly visible on the dimmed down levels as seen in the beginning of the video) by replacing the 100uF cap with a 10uF cap! Now all that's left is to encase it in epoxy or something to make sure the component won't bend over and touch any of the other electrocities.

I think that I can continue on my project with this small change.

And I do have a temperature regulated soldering iron with a tiny tip. I do a lot of hacking electronics but never on this level and only when I'm confident that I can make it safely and only for electronics for personal use.

Once the lamps are finished I'll share it here as well. I'll also give the DIY mosfet dimmer a try to see if I can pull that off.

10

Well, not quite. :grin:

replacing the 100nF cap with a 10nF cap

Only out by a thousand! :upside_down_face:

11

Haha, sorry, that's what I meant to say! :slight_smile:

12

Good Job, thanks for letting us know.

13

Meanwhile I've also designed a much smaller enclosure for the electronics:

IMG_20211021_224739-01
IMG_20211021_224739-011920×3413 638 KB

It has a simple through-hole in the bottom of the case that fits a tie-wrap so it can be fastened to anything I want.

I did however notice that this component seems to get very hot:

IMG_20211005_153232-small_diode
IMG_20211005_153232-small_diode1116×628 90.6 KB

Is this a diode? Can it be that the diode will fail because the frequency is higher?

14

It is indeed a Zener diode, used to regulate the supply voltage to the ICs instead of the voltage regulator chip shown in #6/ #7. It would be normal for it to become hot as it could be dissipating significant power, particularly as you are using the module at its maximum voltage of 24 V.

That said, I calculate it should be dissipating no more than 60 mW at most with that 1k5 resistor feeding it so it should not be hot. That resistor value does not quite make sense to me however.

I suggest you carefully measure the voltage across it.

This is merely in the DC regulation part of the circuit. Quit unrelated. :grin:

15

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.