EEVblog 1747 - Schottky vs PN Diodes & Measurement Traps

Hi, I found this very simple but in very interesting question posted on X by Blind Via who's an excellent follow by the way. Um, and I'll link it in uh down below if you want to join the discussion on X. I thought I'd shoot a video explaining what's going on here because we're going to go down a rabbit hole with diodes that you might not be aware of, especially if you're a beginner because it's nonobvious and there can be a big difference between different types of diodes and what applications you use them in. Can be a big trap for young players. So, the circuit posted is a uh simple diode flowing from a 5V USB input to power whatever circuitry is on the other side. But we don't care. It's common to use a diode as series input protection like this just in case your input voltage is reverse biased for whatever reason. You don't want to blow up your circuit. And the diode of course only allows current to flow in one direction. So it allows to flow in from the power supply but it doesn't allow it to flow out. Hence the why a diode looks like an arrow. And you can see that the arrow is pointing inwards from the 5V USB. So the question is, shouldn't the diode be preventing voltage or current flowing back to the input? And the answer is, well, yes, it should. So let's build it up and see what's going

on here. So what I've got here, I've got a 5V power supply and I've got a diode in series. I um using a surface mount one, so I had to mount it on a board here, but trust me, there's nothing fancy going on here. And if you want to know what it is, it's an STP 2SL60A. here. We'll have a look at that later. And we've got our voltmeter over here, available on evblog. store, by the way. All the best multimeters are on eVlog. store. So, we've got the diode in series. And of course, the diode has an anode and a cathode. Yes, cathode is marked with a K and it's spelled with a C. Eh, don't ask why. Just is. You see that the diode symbol is shaped like an arrow here pointing in this direction like this. So, current is only going to flow in that direction. it's not going to flow backwards. That's what a diode does. It literally stops the current from flowing backwards. So, this is actually the same as our Twitter question here, except the power supply is simulating the circuit, which would normally be powered from an external power supply here. So, the current flows in to the circuit and powers your circuit. And he was using a voltmeter to measure the USB port. So, we've got that diode in series. And sure enough, we're like our circuits at 5 volts and our diodees pointing in this direction. So why are we reading 5 vols on our multimeter? Shouldn't we be reading zero volts? You might be thinking, "Oh, Dave, you got the diode back to front. " Okay. Well, let's put it in the opposite direction, shall we? We're still measuring 5 volts. Is this diode blowing? What the hell's going on here? So, I've got another type of diode here for those playing along at home. It's an SS24. You know, these are quite decent dodes. We're getting the same thing. And you can see the negative mark

on the diode there and the positive terminal. So, I do have that diode wired backwards. Why are we still seeing 5 volts? Shouldn't that voltage be reading zero? Well, yes, it should. So, let's try a different diode. We get the classic 1N4148 signal diode. It's as jelly bean and as simple as it gets. Negative terminal there. Hopefully you can see it. And we put it in. We're reading zero. That works just fine. And then if we change the polarity of that, it lets the voltage through because it's forward what's called forward biased. And then when you're reverse bias, we read zero. So why does this diode work and the other two didn't? Let's try another diode here. You'll be familiar with the classic power diode, the 1N41. We put that reverse biased, we read 0 volts out. And if we change the polarity of that, sure enough, we're going to measure our 5 volts. No worries. There's very little voltage drop on there because there's very little uh current flowing through that diode because we got a 10meg input resistance on our meter here. But you can see it works as a diode. So, is there something wrong

with these diodes? Well, no. They're actually pretty decent diodes. In fact, they're better than these diodes, but they're not better for this application that we're using it in now. And the application matters. These are actually different types of diode to your regular 1N41 and your 4148. These are regular PN junction diodes. If you've studied diodes, you're familiar with that. A diode is a semiconductor PN junction. That's how it works. I won't go into the physics of it, but it allows current to flow in one direction and not the other direction. Well, why didn't these diodes do that? Well, they do that, but they're different types. These are what's called shockkey diodes, named after Walter Shockkey. Don't confuse it with Shock Lee, the more famous Shock Lee, who invented, co-founded and co-invented the uh transistor. No, different person, same field, but yeah, Walter Shocky. And if you go back to the original question schematic, aha, you would have noticed it's not a regular diode symbol. It's a shotkey symbol which has these little like square tails like this. And that indicates that that's a shotkey diode. And it goes by its more formal name of a shotkey barrier diode. So some people call them just a barrier diode, but probably more common just to call them a shotkey diode. And they're different to a regular PN junction diode like your 1N 41 or your 1N4148 here. These are just a PN junction, but the shock key diode is what's called a metal semiconductor junction. It has an extra metal layer in there, and that gives you a huge advantage in terms of

voltage drop. So, let's measure these diodes with a diode tester on a multimeter, and let's have a look. You can see that it works fine. and we put it in the other direction. It reads nothing. It works exactly like a regular diode. And the other one here, look at that. No worries. It looks and tests exactly like a regular diode. But you might have noticed something there on that reading. Look at this. 0. 142 volts. That's a low voltage drop. So, let's get the classic 1 in41. Eh, we got half a volt drop. And the one in 4148 diode once again half a volt drop. The shocky diodes they're huge advantage is that they're a much lower voltage drop than your regular P and junction diodes. And that's why people use shocky diodes in switch mode power supplies when you care about efficiency. You don't just want to throw away uh your power willy-nilly in the losses in your diode. So that's why shocky diodes are really useful for like high frequency uh switch mode power supplies with low losses. That's where you want to use them. But

Dave, if it tests like a regular diode on a diode tester, why is it like feeding back the voltage back out? Look, I've got that's a reversed biased shock key diode. So to figure out why it's doing this. So instead of our voltmeter here, we're going to turn it into a current meter by moving our jack over here. Oh, insertion error. Warning, Will Robinson. Put it over to micro amps over here. And we're going to have a look at what reverse bias current flows through that diode at 5 volts cuz 5 volts is going to matter. And look, there's a nonzero reverse current. It's only 2. 4 micro as right micro amps. That's not much, but it's enough to upset the apple cart in the example circuit that we just had. That leakage current is the huge downside of shocky diodes. Shocky diodes, as we'll see in a minute when we go to the data sheets, are about three orders of magnitude worse leakage than a regular even jellybean PN junction diode. you pay a price for having that shocky barrier construction with the extra metal layer in there. But going back to our original question here, why are we reading 5 volts on here when the diode is reverse biased? And if you use a, you know, one in 41, it shows zero. Why are we getting something here? It's

because multimeters, digital or analog, have an input impedance here. And you might, you should know that it's around about 10 mega. Could be 10, could be 11. It depends on the meter uh design internally, but let's just say it's a 10 megga ohms input impedance. So effectively our multimeter is a 10 megga ohm resistor. And you saw that we had some leakage current, some reverse leakage current in this diode of like what is it 2. 5 micro amps or something. And if you get your confuser out here, I think it was 2. 2 micro amps times just Ohm's law 10 meg resistor. What do you get? 22 volts, but we've only got a 5 vol power supply. So, we're going to read a maximum of 5 volts here. If it was 22, we'd be able to measure 22. In fact, we can wind up the voltage. So, if we wind the voltage up there, you can see it's going to track, but once we get above that like 22 volts or thereabouts, like we we're up to like I think this is a 40 volt diode. So, I'll just go to 40 volts maximum. There you can see it's now only reading 30 volts because the leakage of this diode ch a changes with voltage as we'll see in a minute in the data sheet in the characteristic curves. Um and we've got a 10 megga ohm resistor there. And sure enough if I put my fingers across there I'm going to lower the effective resistance of that and it's going to go up to the maximum 40. Well not quite have to wet my fingers maybe. And once again, that is not going to happen with a regular PN junction diode. But once again, if I put my fingers across there, I'm adding leakage by effectively putting a resistor across there. And you can think of that reverse leakage current as an equivalent resistor across the diode. It's a very high value. It's like, you know, mega tens of megga, but it's

there. So, let's have a brief look at what's going on here. I'll link in this website from TTI. It's just a very nice uh page here with basic resources. And if you want to go into the uh physics of this, which I definitely don't want to do, I'll link in this Perdue University uh shock key diode thing down here. And it's basically a we've got instead of a PN junction, we've got a metal and N junction here. So, they're physically constructed quite different to a PN junction diode. It's not a PN junction plus metal. It's just the end junction plus a metal junction like this. So technically it's easier to manufacture. You don't have to do the extra doping and things um to do that. But you can get into the physics of it. You can go right down the rabbit hole which yeah we do not want to do but bedtime reading. Go for it. So the symbol is quite different to our regular diet which is just the straight line. We have these little curly bits on the end. Sometimes they add the little bit going down and sometimes it's just a h like that. Um eh whatever. But that could be e if you just if you don't add the bit that goes down there, then you could easily confuse that with a zener, which is an angled one. So don't confuse it with zenna. I've done a video on zeno diodes. I'll link that in as well. So I've got some end type silicon and just a metal junction. So it really is that easy. And you can physically see inside a barrier junction diode here. They just have a bit of metal and there's the end type silicon there. Easy peasy. And by choosing a different type of metal here, you can actually adjust the effective barrier size in there and you can change the characteristics of your shocky diode. So you might use platinum, titanium, nickel, aluminium, tungsten or you know all different types of metals. You can actually select them or even alloys I guess if you want to go into that deep. You can adjust the properties of your shocky diode. You can't do that with PN junction diodes. It's just a P. Well, you can adjust the doping and things like that, but it's a very different um physical construction and being able to adjust the properties of your barrier in there. So, anyway, that's the physics of it. So, here's a classic diode characteristic curve. We've got uh current of forward current versus forward voltage here in this quadrant over here. And then we've got reverse voltage in this direction and reverse current in this direction. So a normal diode is shaped like this. And so its junction voltage, its forward bias voltage is quite high. You saw it, it's like half a volt, but when you increase the current, it goes higher, like 7 volts, a volt. So even more at really high currents. So the voltage drop ain't that good. And if you try and use these regular P and junction diodes in, you know, switch mode power supplies and things um where efficiency matters, you can be like losing a lot of um heat, a lot of power in your diode. So you'll use a shocky diode here, which has a much lower forward voltage drop. But when you reverse bias, the voltage on a shocky diode, this is its curve down here. It actually has like a little lip there that goes down like that. But eh, don't worry about that. So the reverse current is going to be much higher. It's this is not to scale. It's kind of, as I said, it's like three orders of magnitude more reverse bias current, leakage current on a shocky than a regular PN junction diode, even a Jellybean one. So you'll have much greater leakage current at um even very low reverse bias voltages like we saw there. 5 volts, that's nothing, burger. um you yeah we will get in micro amps of leakage current and this changes greatly with temperature as well as we'll see in

a minute. So let's look at the data sheet of a shock key diode versus say a 1N4148 uh regular signal diode. They tell you right up the front, very small conduction losses, extremely fast switching, low forward voltage drop, high frequency operation. That are some of the main advantages of Shocky. They're higher frequency switching. They're lower voltage drop. They're even lower noise if that matters to you. And they have no reverse recovery uh charge and all sorts of things. They really are excellent diodes, which is why they're incredibly widely used in all sorts of power applications. But there's two big downsides to shocky diodes. One is that reverse leakage current we've been looking at. The other is that they're generally not as high a voltage. You're only talking like 100 200 volts maximum reverse uh voltage on a uh shocky diode. You can get like really specialized ones that go a bit higher, but PN junction ones can go very high like you know 500 volts or even a 1000 volts or something like that. But um yeah, the reverse leakage is a killer. So this is the one uh used in the uh example. And right off

the bat, we've got our electrical characteristics here. We've got our maximum average reverse current at rated DC blocking voltage. So at 40 volts or whatever this diode is, look, it's milliamps. It's milliamps. We got 1 MILLIAMP. AND AT 100° Celsius, it's 10 milliamps. Not in the micro amps anymore. Milliamps. Unbelievable. That's just incredibly high leakage. Crazy. But our 1N4148, our jelly bean PN junction diode. Oh, reverse current nanoas 25 nanoas typically at 20 volts which is quite a high voltage. Um so yeah but you know it'll go up to you can get micro amps at you know when you're talking about 150 uh degrees Celsius for example but still like you know it's as I said like can be three orders of magnitude lower at regular uh temperatures and voltages. So for the 4148 PN junction, we've got the reverse leakage current here. And you can like it's micro amps see that at 25, they've got three different characteristic curves for three different uh temperatures. At 25°, we're only talking like m. 01 micro amps. We're talking 10 nano amps here, right? At like 20 volts. That's just like right there. That's crazy. But here's the shocky diode once again at 20 volts here at 25 degree C. We're in the milliamps now for our reverse leakage current. So we're talking about like four micro amps there which is basically in around the region that we were measuring there um at the lower volume. You know we were measuring like 2. 2. So yeah it's basically bang on to the data sheet here. We were just using the SMD version of this. Um, so very

similar characteristic uh curve. So that's why we're measuring a couple of uh micro amps. But when you go back and you compare that to the P and junction, you're talking nano amps. Three orders of magnitude. That's a thousand times. If you don't know what order of magnitude is, I've done a video on order of magnitude. Link it in. So to answer the original question, let's go back to the diagram here. We've got what was missing was the 10 megga resistor inside the Fluke 77 multimeter here. Around about 10 meg. And we've got our reverse biased uh shocky diode here because it's a shocky diode. It's got three orders of magnitude more leakage. So there's going to be a reverse leakage current through there. What is that value? We can calculate it with Ohms law. We got 4. 73 volts divided by 10 megga ohms. Get your confuser out. And that value is 473 nano as 473 micro amps. doesn't sound like much, but when you've got a high input impedance multimeter like this, it's going to matter. And that's why you're reading the voltage there. And that's why with a shocky diode of 10 nano, run it again. So 10 meg time nano as you should be able to do that in your head. It's 0. 1 volts. So that's why we were reading like we weren't reading precisely zero. We were reading close to zero was 0. 05 or 0. 1 or something. you know it was in that order because we'll get in nano amps of leakage current. So this particular question is interesting because it's a measurement thing where the old school trap of your multimeter has an input impedance. But if you had another multimeter like a bench multimeter, they can have gig ohms essentially almost infinite not quite but gig ohms of input impedance on the lower voltage ranges. It's simple ohms law. You've got current flowing through that circuit from your 5V rail here. And that current will increase based on the voltage. So if we go back to our curve here, we can see here that the leakage current does increase with voltage like this. It depends on the shocky uh diode could be better or worse than this, but it's going to increase. And the P and junction diodes, they'll increase as well. um you know a similar sort of order differential increase with uh increased reverse voltage here but we're still three orders of magnitude lower than a shocky and there's some other differences with uh shocky diodes compared to PN junction uh PNS will have avalanche breakdown as opposed to the breakdown of a shocky which is more sort of like smooth and gentle and things like that and we could uh maybe if you want a follow-up video of you know stuff like that let us know in the comments. But there you go. I hope you found that as interesting a question as I did that it goes into the differences between different types of diodes. They can make a huge difference depending on the application. So shocky diodes in this sort of uh application you want as low voltage drop uh there as possible. And also if you're using it as a reverse uh polarity protection like a clamping reverse clamp voltage clamping protection for example having this shocky diode clamp at like you know 2 volts or something that means that you're protecting other PN junctions transistors integrated circuits in your circuit that uh you clamp them below the typical silicon uh threshold of like 0. 6 volts. Whereas if you use a regular PN junction diode as a reverse clamping diode, it might be 6. 7 volts even more as it clamps the you know really high current through there and then that 6 volts can turn on or higher voltage can turn on PN junctions in your circuit that you're trying to protect and well yeah you can release the magic smoke and come a guta. So a shocky diode is yeah the perfect thing to use in like a forward voltage drop protection application like this. um that just stops reverse current going in this direction but not leakage current. So it would be an excellent uh reverse clamping uh diode in here to actually protect this point in your circuit uh from any uh reverse voltages. So um yeah better than a P and junction diode. It's just yeah this particular case it's just a measurement thing. But in this particular case, blind view was probably troubleshooting whatever uh circuit this is and just happened to be probing around and saw a 5 volts on this point and went, I haven't got the USB 5V USB supply plugged in. Why am I measuring 5 volts here if this diode is supposed to block it? Should be reading zero. It's because of the pesky leakage and the internal resistance of the multimeter. That's really interesting. And of course, the classic uh trap with using multimeters in circuit for measuring voltage is that you've got any point in your circuit, if you're measuring this point here in your circuit, just be aware that you're putting that 10 megga resistor across there. And if you're measuring voltages in high impedance circuits, that can really ruin your day. You can get false readings because your multimeter has a 10 megga ohm input resistance or thereabouts on almost every voltage

range AC or DC. So to show you the difference what the input impedance of your multimeter makes and how you can actually come in this particular case uh with a digital multimeter and it's high input impedance. This is one of your more rare examples where lower input impedance helps. So, we've got 10 megga ohms input impedance here, but let's get an old school analog multimeter. In this case, we've got a Bobby Dazzler. It's the triplet 630 NA. I've done a tear down video of this. And you can see ohms per volt there. So, it's got 10,000 ohms per volt. So, we're on the 12vt range. 12 volts times uh 10,000 is 120K, not 10 meg. There you go. Using another ometer, we can actually measure that 120k input impedance as opposed to 10meg. So with our reverse bias shocky diode here with 10meg, we're measuring practically the 5 volts. And if we plug in our analog meter in parallel, so we got 10 meg in parallel with 120k, which is basically 120k. What? There you go. It's dropped to 33 volts. And if we actually lower the range down to three when we can barely see it, 3 volts there. Look. 08 volts. And if we go to 0. 6 6V range, then we're even lower input impedance on this multimeter and we're measuring zero exactly as we should. And if I go to the 6V range, which I can because I've got a voltage doubler here. And if we forward bias our diode, there we go. We're measuring our 5 volts or 4. 91 volts with a bit of uh forward bias uh voltage drop there. See, input impedance, it's everything. Or another way you could have done this is use your low impedance range down here to actually measure. You're going to get like a 2k um auto range, but it's so low it's not going to measure it. But it will measure the 5 volts in the other direction. And we're getting a bit more voltage drop there. 2 volts because it's a uh lower value load. It's about 1k 2k something like that. So there you go. Hope you found that video as interesting as I did. If you did, please give it a big thumbs up. As always, discuss down below. Catch you next time. —