Hi I'm gonna show you something a little bit unusual and a bit mind-blowing if you haven't seen this before, so check it out. I've got a little at 40 in pin and microcontroller here. it's a little Msp430, G2, 3 2 1. just that, you know, basically lay pin micro flashing 2 LEDs back and forth.

It's just programmed to do that until the I/o pins. Nothing unusual here at all. We're pairing it from a 3.3 volt rail, hence what? I've got one of these USB power bricks here you've seen recently and they've got a bit of an unusual arrangement on the pins here. pin 1 is actually the VCC pin or the positive power pins I've got that.

He has pin 1 of the chips going up to my positive rail there and then pin 14 is actually the ground pin so you'll see that going down to there that's going down the ground. So I'm power my chip from a 3.3 volt rail and then of course to have the micro actually workers to start up a work. I've got the reset which is pin 10 down here so it's hooked up. so of course if you disconnect the reset pin, it's not gonna actually do anything.

and you can sort of, you know, make it do weird and wonderful things if you try and float the reset pin so you've got to have the reset pin connected there and the micro starts up. It's programmed just to flash the lids back and forth and I've got two eye dropper resistors. So using tool the I/o pins, nothing unusual, right? Well look what happens if I remove the power pin. What do you think's gonna happen? It should stop flashing, right? Look at that.

It still works without the power pin. Why is it so so? I Call your attention to the title of all my shows. Why is it? So, Why is it So now I can assure you that there is no trickery involved here. I Am NOT Doing the breadboard is intrigued underneath or anything weird like that.

I disconnect my power little power brick here and it stops working. But you remove the power pin and the chip still works I guarantee you no trickery. Stop the video here. try and figure out what's actually going on and how is this chip being powered like I can disconnect it.

There it is, start it back up. it starts. But the power P and the VCC the one and only VCC power pen is not connected well. Now of course for any circuit to work, let alone a complex up micro controller like this, we have to have a power loop somewhere.

It has to be getting the power right. There's you know, there's no battery in there. There's no capacitive element that might you know, super cap or anything to keep it charged. If you remove the power, it should stop working just like that.

But it obviously works so clearly there has to be some sort of you know power. get into this pin. Can you figure out where it's coming from? Well, it's pretty obvious, isn't it? There's another 3.3 volt input over here on the reset pin, but it's not a power pin. it's a reset pin.

Sure enough, if we pull that wah, it stops. There you go. but still. Why is it so so Clearly something very unusual is going on here and this microcontroller is being powered through the reset pin instead of the power pin.

Now look, I've got no power pin connected there there. I Get pin 1 is not connected at all. but I've got the reset pin hooked up to my decade resistance box here and I've got a set to zero. Ohm, so it's just like the link I had there before.

Now look what happens when I actually increase the resistors? Let's go to 1 K 2 K 3 K You can see it getting a bit dimmer there we go, but the micro is still working. But you can see that these LEDs are being dimmed by the Raziel effectively a resistor in series with this reset pin here. but it's still working and 10 up. And once you get to 10 K you know it.

really. It's probably yeah. I think it's dead. It's dead on 10.

K So let's play around with this a bit more and it experiment. I've put a pull-up resistor on the reset pin I've actually put a 2.2 Meg resistor on there. just happen to have that handy. Really incredibly high value, right? and I've put my power pin back.

Sure enough, if I remove the power pin. okay, it's not going to work as there's too much voltage dropped across that to point to make resistor. Clearly, we're not going to power the micro and went by or whatever mechanisms happen in here. We're not gonna be able to power the lids.

but look what happens if we plug in any other pin. Any other I/o pin look Bingo! It magically starts working again so we can actually Anton try the next I/o pin here with this I/o pin here. This one. No, we're using that one.

The how about this I/o pin down here? We can power this micro through any virtually any pin on here. Amazing. Is your mind blowing? Oh, why is it? So what is going on here? Is this some sort of like evil voodoo? I Mean, is this something really bizarre about this Msp430 chip? Well, the answer is no. So this can actually be a trap for young players when you're actually building up your breadboard or PCB prototype or whatever and you accidentally forget to apply power D or VCC pin your micro or other types of chips as well.

Seven for series logic. All that sort of jazz. They can do exactly the same thing. They can still be powered via their other pins, so your circuit can appear to work, and then you can have some intermittent fault where, well, a combination of the inputs might cause it not to work, but still, you haven't seen this before.

If you don't know the mechanism behind this, it's like magic. What's going on else is possible. Now, if you're unfamiliar with the phenomenon that's going on here, then the datasheet, unfortunately is not going to tell you diddly squat and you can trust me, you can go search through all like. this is like a 58 page datasheet.

It's not hundreds, but it's you know, So fairly comprehensive data. Sure, you can even go down here the schematic for each individual pin and you can see that well. I Don't know what look? can you see any mechanism in there that allows you to power the chip through the I/o pins? None that whatsoever. And you can keep going through here until the cows come home.

And even if you go over to the Msp430 user's guide here 644 pages. Trust me, if you look through all of these 644 pages, you won't find any hint at all about what's going on here. So what do we have to do? Well, you either have to know this all you have to luck upon a manufacturer's datasheet who actually supplies some info on this. You can search for the word Diode and there's this little bizarre thing here that says Diode current at any device pin plus minus 2 milliamps the absolute maximum rating and that's the only hint to what's going on here.

Let's go over to another random microcontroller you might be familiar with the 80 mega 8 AVR microcontroller. Aha, Here's the magic. All Io pins have protection diodes to both VCC and ground as indicated in this diagram. here is the I/o pin and these are your input protection diodes and you'll notice that they're actually reverse bias, so they're never actually on unless you overdrive the inputs.

This arrow up here, for example, is showing that it's going up to the VCC pin, the positive rail, and of course this one's going down to the ground pin of the chip. These diodes are actually inside the chip and sometimes they have a series resistor in there as well. and this is how they protect the input pins on most almost. you know.

Let's say almost all CMOS devices will have these input protection diodes. That's how if you touch the pin, you know you zap it with the high voltage it's going to. the diode is going to conduct because it's above the rail or below it and it's going to clamp that energy and dissipated in the diodes. So that's how they actually protect modern devices be they complex microcontrollers that we're looking at here or standard seven for you know, HC Double O Series Logic Go look at the data sheet for that, you'll find exactly the same input protection.

So let's take a look at, say our seven for HC double O quad Manship Now they all are going to have ESD protection like this. You can actually see it written there, but very few data sheets are actually going to show you diagrammatically the actual diodes In this. There's nothing in here from Nxp. Go over to this on semi datasheet.

There's nothing in there. Ever. No diodes. The diodes incorporated datasheet.

Nothing here. TI Nothing here at all. We go to Fairchild. Nothing in their datasheet.

What we have to go to is a St Micro Datasheet and Bingo they have it there. It is the input and output equivalent circuit. Look, even. they've got the diodes on the output and the input protection diodes exactly the same.

In this case, they have a series protection resistor Exactly the same as inside. You know, microcontrollers and almost all CMOS chips like this. That's what they use for protection and that's what's at play here. Bloody input protection diodes.

Real trap for young players. So back to this. NXP one for a second. Even though they may not shot, they actually do tell you.

look right up the front of the datasheet. inputs include clamp diodes, bingo and you actually notice that they actually give you a little application hint here as well. This enables the use of current limiting resistors ie. a series resistor on the input pin to the chip to interface the inputs to voltages in excess of BCC because they'll be automatically clamped by the ESD input protection diodes.

So you can actually use these diodes here. and you can actually have an input resistor on here. and you can use these and actually overdrive the inputs safely. as long as you don't exceed the maximum diode current.

And that's why we saw back in the TI Msp430 microcontroller at the absolute maximum dire current rating. plus minus two milliamps. Long as you don't exceed that, you can actually overdrive the inputs using a series input protection resistor. Just a neat little application there, just in case you need it.

So what we've actually got here is a diode protection input on each and every pin. For example, the reset pin we originally played with here. It's just like any other pin. It's got the diode protection so we can power the chip through there and it goes into physically the VCC pin inside.

So yeah, we get a voltage drop, but the chip is still going to work. so let's actually measure that. We've got ground here and our power supply up here, of course, is three point, Three, Seven Volts. There it is, and we'll measure our floating pin here.

Our floating VCC pin. Nothing's connected to it, but we're powering it through one of the random I/o pins. here. Let's measure the power pin There it is two point Six Five volts because we're getting about a knot point six volts diode drop here and the internal silicon is still seen.

That two Point Six Six volts which is more than enough for the microcontroller to still keep working. So every single pin has a diode going to the VCC pin. So it's going to look like this so we can power this chip through virtually any one of the other pins. And that's what we showed here.

So it's actually no different to taking a diode and putting it in series with the power rail. It's exactly the same thing, except that diode is built in as protection on each and every pin. But wait, the mind blow doesn't stop there. Watch this.

Not only can we remove the power pin, we can remove the ground pin. Whoa. This is heavy so you might be thinking hard. Dave I Know all about these ESD protection diodes and it's clearly the one going down to the negative rail down here.

That's doing the trick just like it did on the positive one here. But where? Where is it? How we've only got one ground pin here. we've disconnected it. It's not like we've got any other pin here tied to our member would.

still what the to point to make resistor going up to? plus Three Point Three Volts: What's going on here? Where's that? Where's the Diode? Where's the ground? So this is a little bit of a red herring here. It pulls the video and see if you can figure out how this chip is still working. So were you able to figure it out? Well, yes, it's a diode, but it's not the internal diode inside the chip this time, not the ESD protection diode. We've got two diodes on the outside here.

Look at this. they caught LEDs if I remove one of them. Oh, it stopped working Wow Now this is a rather quirky example where because we're alternating to digital outputs like this ie. when one output is that high, the other is low.

It's the software is not setting the output to high impedance. So to switch the LED off, it actually sets this output low. high, low, high, right? So it's actually driving the output and when it dries the output low, we actually have a ground on this pin. but this actually goes through our circuit.

So we are actually applying ground on this pin via the diode dropped there and the resistor. But it's still just enough to make the chip work. I Mean it's completely out of spec and everything else. and let's measure it and we might see something unusual here.

So let's actually go in and take a look at our ground pin here. Okay, so I'm on the real circuit ground down here. but let's measure it and see what we get. 1 point 4 volts and wow, it just stopped.

It flashed for a second at 1.4 and then jumped up to 1 point. Nine volts You saw it, it was. the chip was on such you know, a low margin of operation well outside its I recommend a 1.8 volts minimum voltage, but it was still able to work but just put in the 10 Meg of how. our 10 Meg input impedance of our meter here was enough to actually cause the thing that stops working.

And of course it won't start up again until we physically go in there and connect the ground. So we go in there, tap at the end and it starts working. but as soon as you load that down with changed it just a smidge boom and I can actually stop that working just by touching that pin. There There we go.

Pingo So on edge, but we can actually use that negative pin input protection diode to power the chip as well and then it won't be influenced by, you know, external borderline factors like that. Okay, so let's start up our chip. it's working and let's just connect one of the other inputs here to ground. If I can get the damn thing in the breadboard there, it is.

Okay. So we connected one of the other input pins to ground. and now if we actually measure it, the ground pin you'll find Bingo! It's actually raised by 0.23 volts. So if not not 0.6 volts, you're actually getting about 0.2 volts across that input protection diode there.

So because we've grounded one of these inputs here, it's able to actually effectively ground via a diode, drop the ground pin of the chip. But why are we getting nor point two volts across this diode on the ground pin. when we actually ground this input and the current is actually flowing through the ground pin and we get no point. Six volts across the one on the positive rail? Once again, I'll let you pause the video, see if you can figure it out before.

I Tell you. Well, here's the answer. It all has to do with how much current is flowing through the LED and where the current is flowing from. Let's take our original example, where we've disconnected the 3.3 volt power pin here and we're powering it through some other I/o pin like this reset pin.

but as we saw, it can be any other pin. Well, we've got our 3.3 volt rail. Our current is flowing through here. so it's flowing through the internal protection diode and it's also so it's powering the chip as so.

I Guess you could say you know this is the chip power consumption because it's a microcontroller. You know it's very small. You know it's reasonably small. It might depends on the frequency it's running at, and you know all that sort of that stuff to do with processor power consumption, but it's also flowing so it's flowing down.

They're pairing the micro, but it's also flowing out the I/o pin and down through the Led. We've got a 1k drop a resistor in there. so 3.3 - 0.6 volts, you know, with - 1.8 volts. maybe for the lead, you know we're talking about like one milliamp through the Led.

1 milliamp a reasonable amount of current. So we're actually going to get that. You know, typical naught point 6 volts across that diode there. Now let's look at the other example where we're using the ground pin protection diode.

This is it here. and we've disconnected our the ground pin on our chip and we're using one of the other I/o pins. It doesn't matter which one it is. Once again, so we've grounded this pin here.

Okay, so this is our protection diode. This is the load of the processor that I talked about it before, which is frequency dependent. everything else. and we've connected up our 3.3 volt rail directly onto our proper power pin here.

Well, now the current is split. The current for the LED actually flows from directly from the pin out here and down through the LED and likewise, it'll go through the other LED and that bypasses the protection diodes are the only current flowing through the protection diode. now is the processor current and it's obviously lower than what the LED was taken because we're getting only naught point 2 volts a drop drop across there. Now because if you remember your A Diode basics, you don't get a constant naught point 6 volts across a diode.

you do after a certain current, but there's a curve. But the diode curve ensures that at very low currents, you actually get low voltage drops like the naught point 2 volts were seen and OPL to measure these currents. So in the case of the missing negative ground pin, let's actually disconnect it. It's still working.

but let me use one of the I/o pens. There we go. it's turned off your sweet. Notice it'll switch back on.

So let's connect it down to the ground and you'll notice that the current is only about you know. Look about Point Eight Micro Amps. Its bugger-all. so that microcontroller is obviously like, you know, going to sleep between those you know alternate things: earthy, extremely low-power It's not taking anything at all because the current for the LEDs is not flowing through that ground pin.

So we're measuring the current out of there. It's incredibly low. But now, let's go back to the original example where we're actually using the positive one. So let's disconnect the positive rail there.

Let's hook it up to any of the I/o pins. Let's say this one here and let's see what we get. Now that one, Point Four, Five Milliamps. that's our diode current because all of the diode current is a plus.

The processor current That you know. Point Eight, my grams is flowing through that positive protection diode we saw before, So it's still the chip is still within operational voltage range. It only needs 1.8 volts to work. It's still getting a three point one volts.

so that's why the thing continues to work. in this case, powered through that negative pin and protection diode and just for kicks. I'm going to show you a bit of a more practical scenario where you can come a gut site. you can make an absolute fool yourself.

Big truck for young players here. What I've got is I've added a second chip here I've added a Yahtzee D 4060. It's a 4000 series. see my site.

Like this little part. I've always liked it ever since I was a kid. It's got basically a built in oscillator which you can use with Eva and RC oscillator. or you can use with an external crystal oscillator and it's basically a binary ripple counter output.

So I've just put some resistors and caps in here just to get you know a frequency that you know we can actually see and I'm just tapping three of the outputs here and feeding those into input pins to the microcontroller over here so you can see those are three red wires going over there and I've just hooked on three. LEDs onto these ones as well. So this is our Q for Q, five and Q6 and you can see that counting up in binary there and the VCC pin to the micro. we've disconnected it.

There it is. It is not hooked up at all. I Just forgot to show the pull-up resistor on here. so I'm going to pull up resistor.

It's a hundred K There's no way we can power the chip through and power these LEDs through that hundred K resistor. So the power has to be provided through the three input pins to the microcontrollers. these red wires going over and you'll notice that when it's counting up. Okay, where our chip is working just fine because we're powering it through the input protection diodes, which are effectively a big all gate there.

So if this input is at logic high ie. the LED is on, so we're getting basically a 3.3 volts into there, then it's enough to power the chip. Same with this one. Same with this one.

So if any one of these LEDs is on, the our microcontroller will be powered. but you'll wait until they all switch off. What happens? Well, our micro switches off, there's no more power, it can't provide it through here. Our VCC pins disconnected and our circuit stops.

and then it restarts once one of these inputs goes hi magic. So this is actually a real practical scenario where a lot of even experienced engineers come a gut. So because there might be something, you might have built your breadboard circuit ROM ya PCB for example, you might have. You might have might have forgotten to solder one of the power pins, but will it be at a power pin or a ground pin or something? and you build up your circuit and it can appear to work just like it is now.

Okay, it's working, but then all that. occasionally it had just looked glitch and reset or do something weird or something like that. So this is actually a not that uncommon our scenario or Fault in a prototype part circuit where you forget to have the power pin. and if you forget the power pin, your circuit can still work as I've spent the last 20 minutes demonstrating.

So there you go. Just watch out for this. Remember the golden rule of troubleshooting thou shalt measure voltages that includes ground and power pins. because I've shown both scenarios where removing the ground pin and removing the VCC pin can still result in your circuit, your micro or it can be.

you know, whatever chip it is. Even this 4060 will have the same thing. We could power this chip and so forth. So there you go.

You just got to be careful of ESD protection diodes. they can be very useful. Showed a scenario. we can actually use them to your advantage, but in some cases they can actually really cause a big troubleshooting headache if you forget your power pin or your ground pin.

So there you go. Just watch out for it next time. I Hope you enjoy that this was a lot longer than what I expect I Thought it'd be quick and easy, but you know I did go through a lot of different art scenarios there I hope you found that are really interesting. If you did, please give it a big thumbs up.

I Don't have a wide enough zoom here, but there's a thumb there it is. Give it a big thumbs up. and as always comments down below: subscribe Yeah, all that sort of jazz. Catch you next time you.

.