Hi! If you're wondering why this video looks a bit different, it's because I don't have my studio lights turned on both these aperture MOA and lights I usually have two of these either side of me to light up the mailbag or light up the bench that I'm actually working on and these are very nice quality high color rendering index Studio lights absolutely fantastic and a fixed color temperature. If you don't know, all of my lights in my lab here are a matched color temperature so that at least recently accurate colors in my videos. So not only do I have two of these, I've got an array of overhead studio lights above the bench as well which are those square panel lights and I Recently had all of the lights in this new lab changed under a New South Wales government scheme where they will replace your original Troffer lights which are the tubes in fact this ones are actually lens but anyway, Fluro Tube light our offers up there with these newfangled LED panel lights. They just drop in and the government will do it for almost free I Had to pay a little bit of cost or whatever, but it's heavily government subsidized and they replace them I Know I actually tested the panel's that they were going to install before they I signed up to this thing to have them replace and I check the color temperature and they were the correct two color temperatures shouldn't care about Sarah because I have higher quality studio lights that give better quality light for that, so it's not a problem.

So this is just for not shooting video just for general lab use. Unfortunately, the problem is I Didn't check out the drivers for the thing and it turns out you can probably see it here in this video. There is flicker. You should be able to see some sort of little horrible, shimmery flickery thing here.

It's really awful. and it's even worse if you use webcams on the left-hand side here that you're watching. This is my regular webcam. This is actually it's not a webcam.

It's my old Canon HF M400. It's just a little arm old Canon camcorder and it's quite good. I Do that via HDMI video capture. and on the right hand side is the Logitech C920 webcam.

One of the best webcams on the market now. I Took down the offending panel and it might look okay at the moment, but let's have a look. Let's have a look. Whoa.

Look at the webcam. Look at the webcam that it's terrible. Muriel Wow Look at that so that so the camcorder handles it Okay, but it definitely shows up on the Logitech webcam camcorders. And these webcams.

You can actually enable the 50 60, Hertz anti flicker mode and I've got that enabled. And I don't get this flickering problem with my other panel studio lights I've got or of course with these professional studio lights as well. So what's the problem? Well, let's investigate. now.

It's obviously not going to be the panels themselves because the panels are just LEDs There's no active driver circuitry inside here, so it's clearly going to be the driver. That's the problem here. and both of these are son LED Energy brand. It turns out that this is an Australian company not too far from here, so they've just rebadged as someone hung Low branding not China put their own model number on it because if you search for this model number and the model number of the driver down here, all you get is some New South Wales government database things.

So yeah, clearly they've just bolt, bought these, and rebadged, and so I think this puppy here is going to be the cheapest possible thing and that's why it's flipper flicker and then flip flipper in. I think we'll run with flipper in. So like I said, there's absolutely no data on this. These are only like twenty four white panels so they're not particularly bright.

The overhead studio panel lights that I've got in here they're from my old lab. They're actually art sixty watts a piece and you can see you can see the new ones flicker in over there. I Can't see these flickering in real life, but it might eventually be bad on the eyes. But anyway, this driver here is going to be as dodgy as and I've taken off the end caps here.

It looks like they just soldered directly on. even the mains input soldered directly on. That's a bit how you doing. They've got the holes in the case for the screw terminals down in there, but they don't use them.

Yes, not off to a good start. Well, there's your problem. now. Whilst it might look, you know, fairly neat and tidy.

The simplicity is the issue here. The primary side here. Nothing essentially wrong. here.

We've got an input fuse that's just flapping around in the breeze a little bit, but that's okay. We've got ourselves a common mode choke and some filter in, so you know that's okay. But and it's going directly into this primary side switcher chip here. And well, therein lies the problem.

There's nothing else. Look at the secondary side. there should be a little bit more on the bottom here. Of course, you know, single sided.

This isn't even if our four. this is like phenolic base. It's better how you doing there. Saving some cost there.

And anyway, we've just got a few passes around here at Asus vs.. she's doing very well today, am I? Anyway, this is all primary side stuff and this is all secondary side stuff. They've got the high voltage isolation slot there, so that's all okay, right? But the problem is the secondary side. There's our donors our do.

But secondary side is so simplistic. There is no a traditional like a regulation. on the secondary side. there is no optocoupler feedback to the primary side.

So how are they actually regulating this constant current? Because this is a constant current driver at 500 to 550 milliamps constant current. Anyway, there's no way that they're doing secondary regulate current, constant current regulation. there. It's a primary side regulation function.

So this transform, they're gonna have to be getting a tap back on the transform. If we have a look on the bottom. there you go. they're sure there's one whined it.

That's the switching winding and this would be a feedback winding here. The reason that they're doing that is because it lowers the cost. You don't need any secondary side regulator IC You don't need any Opto coupler. That's why you're certainly going to get flicker on the output.

It depends on the type and size, if your output cap of you know how much you're actually owning it, and if that's just a 100. Well, at least it's 105 degree rated. and there's nothing inherently wrong with having transformer feedback instead of like opto, a couple of feedback or whatever that's actually fine. The problem with this is is that it has no secondary side current, constant current regulation.

and they've got a piss-weak amount of output filtering as well. So yeah, they've really cut costs on this. That's why you're gonna get a metric crap ton of ripple and hence flicker on this thing. And unfortunately, these model numbers here are not telling me anything.

I Get Diddly-squat searching for that. But look, they tell you what type of material is leak m1 So the best we can do is look at the data sheet for this puppy. It's almost certainly you know, directly pulled from the data sheet or app. Note: First of all, I'm just curious to know what the compliance voltage of the LEDs happens to be.

They're thirty eight volts. of course that will depend upon the number of lead LEDs you've got in the string. the type are LEDs the drive current and the voltage. Drop the configuration all inside there.

but yeah, and if I use my AMT Toa 520i Pro be here with the toroid attachment and put it in wire mode. There's our current waveform. It's jumping around like a jack-in-the-box Let's fix that trigger in. that's it's the noise rejection.

There you go. Bob's your uncle so it's mostly a 100 Hertz ripple there because it's full wave rectified. That's why your fifty sixty Hertz anti flicker filter on your webcam is typically not working. And of course if you get that flicker on your regular camera then your depends on the frame rate issued in at and all that sort of stuff and the beat frequencies as to how it actually flickers so that magnitude they're two volts.

That means nothing because I haven't current scaled it to match what this thing is. So what you've got to do is look up your manual and in wire mode with toroidal attachment amp / output volt. So I'm gonna set that back to one to one probe and 240 million, 240 milliamps. But of course, if you want to do it nicely, you can actually change that to amps.

So it's now two hundred and forty millions. It's now scaling in milliamps. Nice. Most digital scopes will have the feature to change the units and the scaling, and in this case, you can see one volt per ampere.

It's just nice to send up like that just so you don't mentally goof it. And if you want to see the voltage waveform, I'm just probing that. And yes, it is safe because the secondary side is a transformer isolated. I've done a whole video on that how not to blow up a scope and ten volts per division.

Ten Twenty Thirty. There was the thirty eight volts that we saw before and that's the current waveform. So obviously this thing is just a real pile of turd. and that's why you get the flicker.

If those playing along at home, you want to know what the high frequency switching is. Sixty Two Point Seven Kilohertz. they're a little bit tricky to trigger on that. If we wind that out there, you go.

So let's have a look at another panel that I got some time back just as a trial. It's a non flickering series as I said. kind of a non flickering panel because it's just the LEDs inside. It's all about the driver.

Anyway, this is a forty eight What job' crap CRI but I didn't care about the CRI but the color temperature I cared about. supposed to be fifty five hundred but it wasn't I think it was like sixty seven hundred or something. Ah, completely blue balls and it came with yet another of our Chinese driver and a GT a brand one. and but it's supposed to be ripple free.

It's advertised as a flicker free one and wealth flicker comes from the ripple. There you go. So let's take this apart and see the construction design difference. So right from the get-go you can see the difference.

Well, we've got some large heat sinks on here cuz this is a 48 what job'? You can also see that they haven't tried to cut costs because they've got the screw terminals on either end and they cost money. That's why they didn't have them on the other one. That'll pin a pin penny-pinching every cent to the size of this transformer does. it is a higher wattage job' a double the power and look.

They got three output ripple caps I think the others were 330. Sam's on. Yeah, okay, whatever. Look at the large output diodes they got but on the bottom side is where you can see like primary side.

Here's our driver over here. What's very similar to before. We've got a bridge rectifier. we've got the input filter in and the fuse in and we've got a smaller So8 control up primary side.

But of course we've got a primary side switching large switching transistor because as I said, it's higher power. but the secondary side is the TEL. Uh-huh that's actually conformal code. Yeah, they got it on the primary side as well.

Looks like they haven't done the whole board, they're just almost mastered off. Nice touch just to prevent moisture causing issues in there. So you know you can get moisture in these things are up in the roof that heats up, cools down all that sort of jazz. So, but we've got a secondary side constant current controller here within of course that will be driving our switch in Trenee down there with the heatsink, but it's basically just a bridge rectifier over to the main filter cap.

Then it uses a condo proper secondary side constant current controller and you'll notice just like the other one, there is actually no opto coupler feedback from the secondary back to the primary. so they're also using a winding. Hence, yet, they've got the two pairs on the side. there.

One pair will be for the switch in one pair will be for the feedback. But the difference of course is that the constant current regulation is done on the secondary side and they've got big-ass filter caps. That's why this one is going to have what happened. It won't be entirely ripple free, but it'll have bugger all.

that essentially won't flicker. Big difference, and there's probably what. double triple the bond cost there. That could be a CEM to it is there markings on there? What type of board? material? concentration only for you? What? That's hilarious.

Anyway, uses a similar cheap camera type for knowledge base material, but it looks be a bit better than the other one. But yeah, I Reckon there's probably like three times the bomb cost in that compared to the other one. I Know it is higher wattage and you know you got to have the metal work and the big external power transistors and stuff. But yeah, that's like there's a big difference all right.

So let's just measure this puppy and see what we get. I haven't touched the scope. That was the same scale as before. big difference.

So we're 50 milliamps per division there. so like it's not much. remember this is a actually a 1.1 amp so over an amp output compared to less than half that for the previous panel. So as a percentage of the total current, it's not much.

Of course this is AC coupled here. you can see that there is some hundred Hertz there. Of course if you have a look at your time base 10 milliseconds there you go. it's a hundred Hertz And go into the high frequency stuff.

We can single-shot capture that. hey you guys switching frequency 47 kilo Hertz But you can see there's just a huge difference. It's we're getting hundreds of millions of ripple before. Yes, that's why it's flickering.

That's why you can see it. You're not gonna see any. You might be able to measure a teeny tiny bit of flicker somehow if you had sensitive enough stuff to measure that. But yeah, basically, and you're not gonna see that on camera or on, it's not gonna be a problem on the eye as well.

All right, let's take a brief look at the chip used on this sucker, shall we? It's not on on Brights our website, but I was able to get this is a got on Bright Confidential. Yeah right. Nothing Google can't find. Anyway, it tells you right here can achieve low system costs for isolated lighting by primary site control in a single stage converter significantly simplifies the LED lighting system designed by eliminating the secondary side feedback components and the optocoupler penny-pinching all the way to the bed.

It's got high power factor they claim quasi resonant operation, all sorts of stuff fast startup, blah blah blah, and it's got comprehensive protection as well, which you can do via the feedback coil on the transformer Armacao LED short-circuit protection cycle by cycle, current, limonene building, leading edge playing in under voltage lockout, all sorts of stuff. So yeah, but it's a great chip for this low-cost application. The side-effect of course, is horrible amounts of ripple. And here's the schematic, which is basically what we've got on the board.

here. we've got our bridge rectifier on the input, we had a few more you know, a fuse and some filtering stuff and things like that. and we've just got a resistive divider here powering the chip itself and we've got our external transistor was that on there? But this is interesting. We don't have a switch in transistor on the board, so I I got to assume it's the same part.

the part numbers the same but of course but this one says it's in a stop. Twenty and Twenty three six package our ones not. it's in an Esso eight package. so I'd Maybe it's a slight variation on the part that actually in a bigger Esso package that has the built-in power transistor.

So anyway, this is the best and only datasheet confidential one I could find. So it we assume it's got must have a building switching transistor. So but it's basically the same as this and you can see here's our feedback coil as well. They should actually show that going all the way down there like that to show that it's actually coupled That anyway, the exhilarating feedback and that allows them to sense with all the compensation and the output is just that.

it's a single wave rectifier, a one lousy output filter, cap one Hunger brand and straight to the lead. And Bob's your uncle like this? There's no secondary anything, it's it's doing the current regulation via the feedback coil here. So yeah, it's pretty how you're doing, but that's it's cheap and it works if you don't mind the flicker. so there you go.

We might be able to improve the ripple a little bit by increasing the output filter capacitance here, but it's pretty how you're doing. It's a hundred percent ripple that is just yeah. awful. No wonder it flickers like buggery.

But you can't just increase the capacitance, you know, Willy nilly. You got to make sure that the capable of it and the thermals are there and everything else and it's just. and I Don't hack them. just get a one that's properly designed with secondary side current limit and a decent amount of filtering.

And of course it's dependent upon you the shutter speed of your camera. It's Auto You know it's always a changing or ever you can't fix it to avoid these things. but even my high-end Sony camcorders. You can see even with my main studio lights on.

even with these lights up there. I've got like eight of them on and they sort of. The light filters through and mixes with my other non flicker lights and you can still see a little bit of flickering and shimmering in there so it's really annoying. So I'm even gonna have to try and experiment and fix these or just toss them in the bed.

They're horrible. I Don't mind the panels. The panels are okay, so there you go. Let that be a lesson.

Beware of these cheap-ass LED drivers and make sure you get ones that specifically say flicker free or low ripple or whatever it is. Otherwise, you're most likely going to end up with the lowest common denominator like we've got here. Anyway, hope you found that useful. If you liked the video, please give it a big thumbs up.

As always, discuss down below in the comments or over on the Eevee blog forum. Catch you next time.