 I have a new spider. I found her the other day walking across the pavement outside my apartment. They live underground and they're only active during the night. I will need a camera that works at night. If I can build myself an infrared lamp and if I can remove the infrared filter from this camera, which will allow me to record what the spider does during the night. Like that. I believe that I now have a working infrared camera, which is nice. The next step is to light the spider's enclosures somehow and for that I will need an infrared floodlight. So let's build ourselves an infrared light. Now I was originally going to sacrifice this, which is a lighthouse unit for my Roomba robot vacuum cleaner. I have three of these and they've never been much use, so I thought that I could remove the infrared LEDs from this. However, I then had a better idea and instead visited my favorite junk shop, which sells things that other people have thrown away and got these. These are remote controls. In each one of these should contain an infrared LED, which I should be able to harvest. And there may also be other useful and or interesting things in here. So let's start ripping these apart and see what's inside. These will all be the same. There's a nasty rubber mat with carbon pads that push onto tracks in the PCB. A closing of contact and then a single chip. Wow, there's actually quite a lot of this. This must be an old one. There's quite a lot of circuitry. Normally there's a single chip, which is frequently a blob that does the work. What it does is when you press a button, it then outputs a series of pulses to the infrared LED here, which transmits to the CD player by the look of it. So let's just remove this LED. See if I can get it off intact. Excuse me, my soldering iron is acting up. The last time I used the soldering iron was to poke holes in the plastic lid of an ice cream container to produce air holes for an enclosure for crickets, which I feed to my spiders. The spider, by the way, did some more excavation and has eaten a cricket. Well, I assume it's eaten a cricket. The cricket is no longer in the enclosure. It is possible the cricket escaped. There we go. One infrared AD. What have we got here? Some diode transistor, a mysterious thing. The plastic box might be handy for a battery holder, but mostly that is now e-waste. What about this one? We've managed to get all our LEDs. We have a pretty decent hole, actually. So let me do some clear up and then we'll see if these actually do anything. I now have a nice handful of infrared LEDs and one vintage red LED. I took this one off in order to demonstrate that it is really useful for testing. If I plug it into my tester, let me just put this in a place where it doesn't get too much camera glare. Here, between pins 1 and 2, and press the button, you can see the light flash as the tester figures out that it's a diode. This also has the advantage of pinpointing the positive and negative side, which is not always terribly obvious. Let's try that with the infrared ones. I have the infrared camera here set up so hopefully we can see the flashes. There we go. Blinking nicely. And that shows the positive side is on pin 1. And I'm not sure the camera will focus adequately, but the LED does have a flat side, which in this one is pin 1. And we'll go through all the others just to make sure that it's that way around and that they all actually do infrared. And what I'm going to do is solder all these to the edge of this piece of strip board. All of them in parallel so I can drive them simultaneously. So I believe I finished the bulk of the soldering. It did in fact change the layout to make it all fit. It now uses a single track for the cathode, the flat side of each LED down the middle, and then staggers them in the opposite direction. I did have some problems where one of the LEDs was the wrong way around, but my trusty component tester here detected it. So now if I stick these two pins in and push the button, you will notice that it shows up as a single diode, which is good. All right. So the next thing is to actually drive it properly. So I will just remove these two pins I soldered on for testing. I want to run this thing of five volts. And so since LEDs need series resistors, I did the calculations or at least found a website that did them for me. And it seems I need a 38 ohm resistor. So this is indeed a approximately 38 ohm resistor. So I will just double check that. I might have mislabeled them. There we go. 38 ohms. So let us solder that in. So push through the ground and solder. So with luck, I also need a wire for the other pin. Okay, we've got some wire. So let us tin the ends. I had to guess at what sort of current these LEDs like. And it seems that infrared LEDs typically run at what, 100 milliamps? So a 105 volts at 100 milliamps requires a, it doesn't actually work, requires a 38 ohm resistor. Now it's possible that since I have LEDs in parallel, now I think of it, the current may need to be higher. I've got 10 of them. Do I actually need a 3.8 ohm resistor? I should probably check up on that. I went and read up how LEDs handle being in parallel and how to calculate the resistor. The answer is you can't put LEDs in parallel. I wish I'd thought to do that before doing all this work. So what is going on is, I'll just turn the power supply off. LEDs are not ohmic devices. So the higher the current through them, let me rephrase. With an ohmic device, the current is proportional to the voltage being dropped across the device. A classic ohmic device is a resistor. A resistor will draw twice the current at twice the voltage. However, LEDs don't do this. LEDs either conduct with low resistance or they don't conduct with high resistance. And the actual current through them goes up very sharply as the voltage goes above a particular critical threshold known as the forward voltage. When you connect multiple LEDs in parallel, then they will all have very slightly different forward voltages. And you end up with the case where one LED will turn on, its resistance will drop sharply, and then all the current will go through that one LED and not any of the others because they haven't turned on. With the result that the LED goes pop. It's a situation known as thermal runaway. And as this thing is going to be running through the night, I wish to avoid it. This is safe because this resistor here is sized to run any one LED. So even if one turns on and all the others stay turned off, it will limit the current to the appropriate value. If I put the 3.8 ohm resistor in and one LED turned on, then that would be 10 times too much current and the LED would die. What you have to do instead is use one resistor per LED. So I'm going to need a completely different set of wiring. Fabulous. So now I'm going to have to desolder this, and of course desoldering is my favourite task. How hard can wiring up 10 LEDs be, I thought? This will be a nice simple job suitable for a short video, I thought. Bah. Well. So I'm going to have to rethink why I'm doing things. Each of these LEDs, and I hope they're still going to work after I've desoldered them, needs its own separate resistor. Luckily I have 10. Just the right number. So each one will need a resistor. Each one will need one pin connected to a common ground or power. And the other pin connected to the resistor. And the other end of the resistor will be connected to the other power rail. So I'm wondering whether it would be best to do this with floating resistors. Or I've got this prototyping board. Prototyping board is similar to VRaboard, except each pad is on its own and disconnected from the others, rather than strips like VRaboard. This means that if you want to connect pads together, you have to bridge them with solder. Okay. I'm going to take a break and then come back to this fresh, I think. So it's the next day. I'm reasonably refreshed and more coherent. I think it is time for a time lapse. So now let's hook this up to our good old tester. Now this won't work quite as well because there are resistors in line with the LEDs. So who knows what it will show up as. And let's get the infrared camera set up, which should be here. Press the button. No blinking lights. It's showing up in the tester as a resistor. There are 4.4 ohms, which given that I have 10, 39 ohm resistors in parallel is about right. So let's try it with the power supply. That's on. 5 volts. Nothing. Well, it's during an amp, which is what I expect. That's 100 milliamps each. But I'm not getting any results. So I've actually cut a couple of the bus bars, isolating all except this LED here and this LED is also still not lighting. So I think the only course of action now is to desolder it and see whether it still works. Okay, I did a bit more reading and I am somewhat confused. Let me show you something. I've taken one of the LEDs off and patched on this 220 ohm resistor. And when I power it up, it lights, as you can see in the camera footage. Now, I found a datasheet for the infrared LEDs, which say that they will max out at 50 milliamps, which is way lower than the 100 I have calculated these resistors for. So really, I should have exploded the LEDs on this board. But I haven't. They just didn't light at all. But this one is fine. When it connects it up to 5 volt supply with a high resistor with a lower current, it lights fine. So what is going on with this board? Well, the answer is I don't know. So I'm going to try putting this LED back onto the board. But I'm not going to bother with that for now. Instead, I'm going to mend these broken connections. So I just need to straighten these wires I cut out so they touch and then add a solder blob like so. So now we stick some... The end seems to fall on off my soldering iron. I think this soldering iron might not be too long for this world. I might need a new actual iron part of it. So we put lots of solder on here. This is turning into a mess. I think the plastic screw thread is not in great shape. That explains some problem I've been having with soldering because the tip may not be getting hot enough. There we go. That's better than this one. I think that should do it. So now all the other LEDs should be connected up. So I am going to crank the voltage way down as far as it will go. Hook it up and then wind it up a little and see what happens. You see, it's not lit. It's still not lit. That is just not lighting. So why was this lighting? So let me just double check which side the flat is. So this is ground. I'm sure that's not shorting out and this is... And you can see lots of infrared light coming out. If I wind this up to five volts, you see it gets a bit brighter. The focus is all dreadful because I haven't adjusted it. Adjusting the focus and the camera is pretty dreadful. So let's just... I need to hold this button down and do one of these. That's the wrong way. Anyway, you can see that light is coming out. So I am deeply confused about what's going on here. The only thing I can think of is that I've somehow managed to screw up the connections in such a way that the electricity is not actually passing through the LEDs at all. That way they are not being subjected to the high current and therefore aren't failing. I wonder if one of the LEDs are shorted through, but the LED itself, once it's lit, once it starts conducting, it's got negligible resistance. So yeah, I just don't know. So I think what I need to do now is to take all these LEDs off again and check the board. So I got the hot air gun out and laboriously desoldered this. Not all the holes are clear, but most of them are. And I have rapidly figured out what the problem was. Do you see these white marks in between the pads on each side? Well, that indicates that there is a conductive track between them. So because the LEDs were attached across these, the LEDs were being shorted across. That was why it wasn't working. So after all that, I went through and I retested all the LEDs and surprisingly enough, nearly all of them work if I attempt to get this in. Unfortunately, I ended up having to cut the leads short. But if I hold that in front of the outer focus camera and run the tester, it flashes. And in fact, they all work except probably this one, assuming I didn't manage to... Let me just get these a bit more on camera. Unless I managed to get them mixed up. Yep, that one is dead. I probably ended up dumping 100 milliamps across this and fried it. So this one goes in the bin and we go back to the drawing board. I have 10 40 ohm resistors and nine LEDs. What can we do with this? Well, it turns out that one of the recommended ways to run groups of LEDs is to put them in series. I want to run this a five volt so I can use a USB power supply. So because each one has a voltage drop of... Well, the tester will tell me about 1.1 volts. That's actually quite low. There's supposed to be about 1.5 to 2 volts for an infrared LED. But this is saying 1.1. So we can put them in parallel until you reach five volts and then we need a single current limiting resistor. So at 1.1 volts, four of them gets to 4.4 volts. So we need to drop 50 odd milliamps over the remaining 0.6 volts. So that should be fairly easy to calculate. So luckily it turns out that with a string of three LEDs then in order to get a current of about 30 milliamps which seems reasonable for one of these LEDs you need a 40 ohm resistor so I can make use of these although at least only three of them. So I think I'd better solder this up now. And you know what this means. So let's plug it into the tester and see what the tester thinks it is. Okay, the infrared camera is on so if anything comes out we'll see it. Of course only three lit. But that does believe it's a diode with a forward voltage of 2.4. 2.4, interesting. Why 2.4? It should be 1.1 times 3, 3.3. So I suppose it could be two diodes but three of them are lit. Okay let's do some continuity testing. There's probably a dodgy joint to the bus bar or possibly two dodgy joints to the bus bar. Well those all make connection. Although it might be that the voltage is sufficiently low it's only causing, well clearly I have two different types of LED here the blue ones and the white ones. It may be that the blue ones have a lower forward voltage which means that the meter which is not producing very much current at all is not sufficient to light three at once. So let's get out of the power supply and wind this all the way down to try and prop it up so the camera can see it. Let's get some of these wires out of the way. This is all turning into a horrible mess. Okay what do we get? Okay two of the white ones have lit. We will just wind this up a bit. Okay there knows the light. We're at 3.2 volts, 4 volts, 8 milliamps, 5 volts. Okay that one has not lit and none of these three have lit. So the current must be going through this one because they're all in series and if that was not lighting there's a piece of gunk under the LED if I can get that out. I know why these haven't lit. I failed to bridge a joint there. Okay so that's light done. That looks clean. Let's stick it back on the vice, line the voltage back down. That should be okay with 5. I just want to see them light one at a time. Okay we've got those. And that is drawing a 150 milliamps total which I think is alright. Now for that LED to not be working maybe I've shorted through it. Is there a solder bridge? There isn't one here. It would be between these two pads. Actually I can detect that using the meter. So if we do this, that's the connection. That's the right way round. So black to black, red to red. So if we do it the other way round there's no connection. I wonder if that LED has failed through. It's actually much more likely because I can see something under it. It's much more likely that there's a solder bridge on the other side. The holes in the board are kind of prone to using capillary action to pull solder through the board. That's happened several times when I was soldering it up. Okay well, I'm just going to have to take it off again. Okay, so let's stick this thing in the tester and see what happens. The LED is fine. That means the problem is with the board. What do we have here? We've got a big blob of solder which is my fault from the work I just did. Let's get rid of this then. Get rid of some of the whiskers. Okay, it looks good. Let's use the meter to check that there aren't any unexpected conducting pathways. There are not. So look at the underside of the LED. It looks fine. Whatever that black blob was is no longer there. So we just need to arrange it. The right orientation. Come on. And slot it in. And solder it up. And just see if that's flush. It's not what I wanted to do. All right. So that should now be soldered back into place. Let's stick it in the vice again. What do we get in the camera? Great. Now none of these are working, which is not what I wanted. That's because I forgot to solder up the bridges that I had to remove in order to get the LED out. So it's fact not connected to anything. Okay, that's one. And that's two. Okay. It's getting quite late here. So I'm making stupid mistakes. So let's give this one last try. Fantastic. All nine LEDs work. And yes, I'm aware they're showing up in purple. That is just, that's the color that the infrared shows up on the CCD. I'm going to convert all the pictures to black and white anyway. And as you can see, can you see anything coming out of that? My real camera has a decent infrared filter. So very little should be coming through. But that is working. Finally, 120 milliamps, nine LEDs. Shame about the dead one. I'm going to call it a day. So next time I actually get to try and take photos with this. And I'm actually going to have to make another modification to this camera that means taking it apart again. But until then, I hope you've enjoyed this video. As always, please let me know what you think in the comments.