Hi. This is just going to be a quick follow-up video to my previous one on the Raspberry Pi three photosensitivity issue because quite a few commenters us. and I think rightly so that it might have been caused this lockup issue when you use the photo flash on a camera on the Wi-Fi module here on the back, which is a new issue which wasn't on the Raspberry Pi 2, then it might have been caused by, you know, some people saying like EMP It's basically magnetic coupling from the large amounts of current in the photo flash here onto the board. so it may have been an electrical coupling interference issue rather than photo sensitivity of the bare die on the bottom here.

So that's a fair enough comment because the photo flash on here. There's a charged capacitor inside here, a photo flash capacitor which charges up with a massive amount of energy and then dumps it very very quickly into the photo flash bulb here. and there's wires running up the sides here like this: I Like the capacitors inside the camera and the wires running up here, this means there's a big loop area here. and what this loop area means is that any corresponding loop area on the board, you can get effectively magnetic transformer coupling loosely coupled.

From this into this, you've got a high discharge current very quick that can generate EMF in loops inside here. primarily our power supply loops and other things. Some people might have said the antenna, but the antenna is not actually a loop as such. Anyway, we won't get into the details of that.

I thought yeah, that's a fair enough comment. So let's actually test it now. What I've got is I'm just pinging here to our double eight double eight and you can see that our time there is. you know, 20 odd milliseconds or something like that.

Occasionally it jumps up to 30 or something like that and that's just fine now. I've got some black electrical tape over the Bairdi Wi-Fi chip down here so let's use our flash and see if we can see any variation in that pin time. No, nothing whatsoever. and let's try it a few times just for good measure, just to show that it wasn't luck.

And for reference, I've got this about 10 centimeters away, actually 11 centimeters away from the actual white ship itself. Now, I won't do anything different except physically remove the black electrical tape. So our Bairdi Wi-Fi bluetooth chipset is now exposed exactly the same camera position, all the same settings and bingo, we got the thing to lock up now. Oh there, No it, no there we go.

it actually recovered I was going to say I saw this a couple of times. it did actually cover set went up to 95 milliseconds there and but I had to reshoot this video. by the way because I lost all my data because I don't ask the memory card issue. Oh there we go.

a thousand milliseconds and before I got it straight away but it was actually 10 centimeters away before. so I'll do that. But anyway, that is confirmed. It is a photo flash issue and not I'll put it back there.

we go so we've moved a bit closer. it is. That is definitely confirmed. It is a photo flash issue photo sensitivity issue with the bare die.

the energy particular frequencies is getting through that die and let's try it again. Now see if I can get a complete lock up I have actually had it recover like that that you saw before, but then I have actually had it completely lock up. Hopefully it'll do it this time where it just does not recover at all and it completely times out and we get no more internet and I think maybe it's going to do that this time. So there you go.

that is definitely confirmed. It is not any sort of electrical interference coupling issue and definitely photosensitivity of the die. Absolutely confirmed and will actually start to see some messages There we go. No buffer space available so we're completely locked up.

Our internet didn't recover. It's no good. You have to reboot the board and know it doesn't matter how close I have it here. I've got it like three centimeters away.

now. It's right. it's clacker basically and we can. There we go.

We can flush that thing and it does absolutely nothing with the electrical tape in place. It's completely opaque. We take the electrical tape off and this puppy is going to lock up. a guaranteed Bingo confirmed.

And there were also were several people who wanted me to use the D message command. is it or whatever it's called I don't know I've never used it before with the /w follow option so wait for new messages basically. so let's actually try that Here we go. and I don't know where all that means it's all Wi-Fi chipset Bluetooth data whatever.

Anyway, it's waiting for a response. Oh, just put the tape over that and flash it. We're back at like 11 centimeters distance or anything. nothing and take this off.

and there we go. And we got absolutely nothing out of that whatsoever. Let's go down a bit more so we're close. I So we definitely know that it's going to lock up.

and nope, we didn't get anything out of it. What's the way about sorry and we will see that that is definitely dead. There we go. So of course that's going to be exactly the same issue for the use 16 power supply chip on here.

just like on the Raspberry Pi 2. There was basically no reason to suspect it wasn't Now sorry to anyone who was watch my previous video. seemed to be quite a lot of people bit a bit of an uproar, really got the impression that I was making out that this was a somehow like a really big deal and it's not I Thought that was pretty obvious from the video that you had to have it so close that it's an unrealistic scenario. and that raspberry pie seemed to possibly I don't know, we need confirmation from them maybe purposely tried to fix the you 16 issue by putting that black coating on the top from the roster pi2 because it wasn't really a big deal for the Raspberry Pi - either.

it was kind of one of those nice things that if you use this exposed board and you know for an art installation or something and somebody came up and took a photo flash photo of the thing then yet it might reset and shut down. In any case, it wasn't a big deal back then because the fix was easy and there are the odds of it happening to your average person or borderline zero and but it now has a black coating on it's probably an order of magnitude better ie. less sensitive to that reset issue on that chip as you saw in my previous video. it had to be like, you know, an inch or two away right up its clacker in order to you know for it to reset and be an issue and I thought that was pretty obvious.

Okay, I didn't sorry if you got the impression that I made out that it was a big deal, it definitely isn't I just did it as a for a a follow-up video to the previous one because nobody had tested the Robert Raspberry Pi 3 that I could find anyway to see if this was still a potential issue or not. Regardless of how small the issue is, academically, it is quite interesting and potentially hey, I don't know the output of the flash on this camera. if you had a big speed light thing on your digital SLR or something and you happen to have the bottom of your board exposed like this with the bare Wi-Fi died and you were relying on the Wi-Fi or Bluetooth or whatever and it could remotely be an issue that you might have to take into consideration, but most people will not have to worry about. It's basically a non-issue it's just interesting and I thought it was worthy.

Just do or pointed out to people that you know potentially that issue might exist regardless of how small it is. Anyway, the fix is easy. Black electrical tape, blue tack, some sort of, you know, hope, anything opaque basically and non conductive, some sort of silicone or something. It doesn't matter.

all whack it. In a case, as most people do, it's not a big deal. so sorry if you've got the impression that I was making out that it wasn't Anyway, hope you found that interesting. Definitely confirmed It is a photoelectric issue on that Wi-Fi chip.

Still a small problem and hey, raspberry Pi, don't have to fix it. It's a low-cost non-for-profit board. I'm totally aware of that and you know, no big deal, it's just an interesting video anyway. I Hope you found that conformation interesting and I haven't done extensive tests with exactly how far the Y fiber is saw.

It was interesting that it sort of imposed the Wi-Fi and upset it a little bit and then it actually recovered and got the ping. I Don't know if there's any other issues I Don't know if Bluetooth that fail doing that or whatever, but you know I mean the odds of it being that EMP thing were quite low. I Think because the chip designers, for example, they're gonna like this chip, they would have been basically concerned about any you know, interference to it on the power supply and things like that. That's where they would have concentrated.

But like the million odd transistors in this side side this thing or how many it is and how many and which ones get affected by the photoelectric effect from this flash here at particular frequencies practically impossible to no predict, protect against or something like that. So you know that's just the thing with these bare dies. A lot of people are. Start: Why do they make these packages If they're photosensitive? Well, it applies to all bare dies like this.

And because they're basically chip scale packages, they are really, really incredibly small. And these packages are what make your mobile phone and other ultra miniature products these days possible if you put them in standard epoxy packages and everything, they're bigger. and maybe only a little bit bigger. But all that space, extra space adds up.

so they're You know, That's just the thing with these things. A Bare Die is technically photo sensitive. It's It's nothing new. it's a very well know enough phenomenon.

And that's you know. Usually they go into cases, the mobile phone case or whatever, and they're not exposed to anything, so they're just fine. So there's nothing wrong with these chipset packages. They're not bad design, they're just exposed die.

That's the nature of the beast. Anyway, you found that follow-up video interesting. Catch you next time.