Hi Welcome to the Eev blog an Electronics Engineering Video blog of interest to anyone involved in electronics design. I'm your host Dave Jones Hi Often when you're playing around with your microcontroller circuits, you build up your little board like this and you just want to program it and it usually doesn't draw all that much power. And you know that uh, the 5V USB port that powers your little programmer. Your In Circuit programmer can easily power your circuit and it's just handy to be able to power the board and program it at the same time via the In Circuit programming adapter.

This one's the AVR ISP Mark 2 programmer now the Uh Pit Kit programmer for the Pck microcontrollers that has the nice function of being able to generate an output voltage that comes on one of the In Circuit programming pins and it can power your circuit under test which is really handy when you you just want to run the thing on your bench. You don't want to have to have a separate power supply to do it, but unfortunately the Avrisp Mark I doesn't have that capability and I found that really annoying. So I thought I'd do a very quick mod and show you uh how I converted this to generate an output voltage on here and power your board under test. And here's my modified AVR ISP Mark I main unit and I've made two very simple mods to it.

One is an Onoff switch which all it does is it applies Um either 3.3 Vol or 5 Vols through to pin two of the Uh six pin in circuit serial programing header. Now normally pin two Uh is actually an input. Well, it still is an input, but it uh, it reads normally on an unmodified unmodified unit. it reads the input voltage from your circuit.

So which Powers the input buffer on the thing so it knows what uh signal level you're using. If you're using a 5V circuit uses it. It knows to use a 5vt input buer 3.3 or even lower than that, it can adjust to the level. and the beauty about that is that means because it's pretty much an input, we can simply wire a voltage directly to Uh pin 2 on the board in here which can actually power your circuit.

Now my circuit my my little board here only needed Uh 5 Vols it worked from 5 Vols but I thought I' go to the effort just to put in 3.3 Vols in case. uh, any future boards I design I'll be able to power those as well. So let's take a look inside. Now it's actually not a bad little design.

this. It's got four little Clips here on the back and you just get a flat later screwdriver in there and push them aside and it just leevers off and opens like that. and there's the main board in there. and they've also got two clips down the bottom here which allows you to take out the main board.

you can just uh apply. just pull it back like that and the board will lift out when you pull back on those connectors and bingo it just pops out. It allows you to do mods really simply. I like it.

Now my modification is really simple. The Onoff switch here just goes through is solded to pin two on the ISP connector down there and then the input to it goes to the 3.3 Volt or 5V selection switch there. um, single pole, double throw and it just uh and I've just added an LM 317 voltage regulator on there with some decoupling caps and some a couple of resistors to set the voltage and that's pretty much it and that's powered directly from the 5V uh input from the USB Now, unfortunately atmail don't actually uh, give you the circuit for this Mark 2 uh program and I couldn't find any circuits online so if you know of any uh please, uh, post the link so that everyone can take a look at it. So I didn't really bother reverse engineering the circuit to figure out how it works I just knew that the Um input because it's uh, it's a voltage buffer um based input that I could just wire the voltage straight through the output and it wouldn't do any harm because it's designed to read that input voltage.

So whether or not it comes from my voltage regulator here or the voltage comes from the connector on your board, it really makes no difference. So I was 100% confident that um I was safe in doing that now. Uh, if you just want to wire 5 Vols through to the output and not worry about adding the regulator with 3.3 you just tap off the 5vol pin which is that furthest one with the large Trace coming off it. Just wire that directly through the pin too and that's all you have to do to mod this thing.

Um, but I'd recommend putting on an onoff switch of course. Uh, but you can. The mod could be that simple if you just want to power 5V circuits. and of course, the USB is capable of supplying Uh 500 milliamps at 5 volts typically or 2 and A2 Watts I'm not sure how much uh Power this board takes I haven't actually measured it, but it's not that important.

It wouldn't take all that much at all really. So there is, uh, there should be ample Uh current to power your circuit under test. M You know, 350 400 milliamps or something like that should be readily available, but we'll check that later and see how that goes. And here's an up close shot of my mod.

and I've put the LM 31 seven backwards here because, uh, it made sense to do that because the input pin here is on this side of the device and I just bent that over and connected it directly to a capacitor. It's quite hard to see. Sorry about that, but a a capacitor that's directly connected through to that large Trace which is the 5V input from the USB connector so that goes directly into the Lm317. Uh, this one over here is the adjust pin.

as you can see I've got the resistor and the tab of the device is actually connected through to the center pin here which is the output. so I could have solded that wire directly onto the center pin, but I just thought it was uh, nicer just to put it onto the Uh tab at the back there and this is one of the feedback resistors and the other one which is used to set the voltage. I've got two in series there which will go into the circuit but they uh, these are the ones that set the output voltage and this here is um, just an input 100n uh input coupling cap for the 5 volt rail so it goes 5 Vols and it's soldered down onto the tab of the Um the the shield of the USB connector which is I've checked it is actually connected through to the ground. Point Um and this here is the another Um 100n output capacitor.

oh sorry, it's not 100n, it's like um 470n or something like that and uh I just picked the Uh highest value I found in my junk box that fitted and it's connected between the output and ground as well and that's all there is to it. It's pretty simple and the output just goes off to the 3.3 or the 5V selector switch and then switches through to the Onoff switch which goes through to pin 2. Simple and it all seems to fit nicely. This um, the actual voltage regulator isn't uh, glued down or anything like that.

it's just sitting there so it's not the most robust solution in terms of vibration. So um, yeah, I'm not sure long it will last, but I didn't bother putting it down it. it'll be fairly rugged, but it can actually move there. but it's not a big deal for just a simple development device like this.

And one thing you've got to be careful of there is just to drill the switch on the other side of this device as you can see I goofed it I didn't have my head screwed on and I drilled the hole on the wrong side of the device which without catering for the height of the device when it plugs in like that so it fouled on thep top of the regulator and that wouldn't have been very good to short out to that to the switch. So I just uh, shifted that one over and put a label over it. No problems fixed and let's have a quick look at the Lm317 data sheet, shall we? This is the National Semiconductor one, which just so happens to be exactly the same as the one I'm using. And let's look down here at the typical application circuit and this is basically what we have here.

The Lm317: We've got the input cap 100 in just for good measure because I'm not sure if there was actually one on the board I didn't fully trace it out, but I wasn't actually sure of the input configuration on there. so I added one just to be sure. I've got an output cap. Uh, basically one microfarad is a little bit.

um Overkill that's a recommended value. You can use it. You can use 100n as well if you want and it's going to be perfectly, uh, stable. I used I think it was a 470 in I'm not sure.

uh and I've got a um, it says it. use a 240 Ohm here. that's just a nominal value I actually used a 220 Ohm there and a variable value which we have to calculate for our nominal output voltage of 3.3 Vols. So let's do that.

Here is the typical formula for a standard LM 317. It doesn't matter which uh device you've actually got, but so we know what our vout is. our V out is 3.3 Vols and we actually want to calculate R2 here so we know what R1 is is. We're using 220 Ohms for that, so you have to rearrange the formula here.

So if you do that, R2 = V out on 1.25 which is the voltage, the internal voltage reference inside the device, and then you subtract one and then you want to multiply that by R1. So if you punch those numbers in, you will get uh a value for R2 for 3.3 Vol. So if we want 3.3 Vols over 1.25 -1 * R1 is 220 Ohms. That gives us an R2 of about 360 ohms.

And because we chose an E12 value here, we'll choose an E12 value here. To make it nice, we'll actually choose a 330 ohm plus uh, we'll make it a 33 Ohm resistor to give us 363, which is more than near enough for our purposes. Now I Know what you're thinking. What about this term here? Why did I leave this I Adjust time R2 out of the equation? Well, this is typically done because it's usually quite an insignificant term unless you've got quite high value resistors here.

Um, for you for your feedback, but because we've got quite low values, you'll see that it doesn't add up at all at much too much at all, so you can simply ignore it in most practical cases. So the I adjust is actually the current required for the adjust pin here, which uh flows through R2. Hence, the Um output will be the adjust current times R2 which is Ohms law. the voltage drop across that resistor plus your value.

So that's why they they add that term plus I adjust time R2 Now if you look at the data sheet for the device the other page, you have to look down here. Let's see if we can get it here. adjustment Pin current. There it is and it's uh, typical value is 50 micro.

but if you're doing these sort of calculations, you always take the worst case figure unless you know exactly what you're doing. So we'll take the worst case figure of 100 microamps and we'll multiply the 100 100 microamps times our 363 Ohms. That's only equal to about 36 m Vol So instead of our nominal 3.3 or whatever value uh, it works out to with those exact value resistors, it'll be Instead of being 3.3 it'll be 3.33 6. And it's not really a big deal for our purposes, but it's something to be aware of if you are designing LM 317 circuits.

Now, just in case you're wondering why, they typically choose a value of 240 Ohms here. For for the Uh sense resistor here, it's pretty much because because of the Um error current from the adjust terminal, it it effectively sets like a minimum load current required. Uh, for this device, which they typically specify in the data sheet here. If you take a look, a lot of the specs will actually show, um, like a minimum of like 10 milliamps.

These specs will only be appropriate for a for an out of 10 milliamps up here. So if you go down and you take a look at a couple of these, they will actually have uh, minimum values of 10 milliamps. There it is. Load load regulation.

you got 10 milliamps ey out up to IMAX. So really, um, having a low value for your sense resistor down here pretty much, um, ensures that you're pretty much going to meet that minimum load current requirement. You can actually increase these values, but then the error term becomes significant and you're not going to meet a lot of the Uh data sheet specs, so you really should keep that to about 240. Well, you know, certainly under like uh, 500 ohms or something like that.

To really, uh, keep the thing within spec I Typically use a 220 ohm. Now the other trap with the LM 317 is that it is not what's called a low Dropout voltage regulator. It's a just a standard linear regulator with a High Dropout voltage. That means the Dropout voltage is the voltage differential between the input, so the input must be X volts higher than the output.

Otherwise, it drops out of voltage regulation. which I Won't go into the details of how and why, but it basically means if you've got Uh 5 volts out of your regulator. for example, what you need. Typically for an Alen 317 as a rule of thumb, at full current, it's taken as a 2v minimum drop and we'll take a look at that in a second so you would need a minimum of 7 volts input.

otherwise your leg regulator will drop out of Regulation. Now here is if you go through the data sheet um, it'll it'll be the same for all Uh voltage Regulators Like this, all linear voltage Regulators will have a Dropout voltage specification. You can get special low Dropout types which go down to several hundred Mill volts or something like that. They can actually be quite low.

But here's our parametric graph of the input to Output differential voltage on the Y AIS here from 1 Vol up to 3 Vols and that's the Dropout voltage okay on the Y axis versus temperature because this Uh effect does change with temperature and as you can see curiously, it has Uh at different and these are Uh load lines are for different output load currents. This one up here, which curves like that is for the maximum 1.5 amps output and then 1 amp and 500 m 200 milliamps and then the lowest Um load line they've got is 20 milliamps here. So let's take a look at the 20 milliamp one. and let's take basic room temperature here for our 20 milliamp uh minimum load current graph.

So at 25 C which is room temperature, we've got a Dropout voltage of about 1.5 Vols. So if we want 33 Vols um output, that means we need an input voltage minimum of 3.3 3 Vol + 1.5 Vol which is 4.8 Vol. And um, technically our USB should give us that. although USB is 5 Vols plusus uh, uh, 5% so it could actually be um, you know it could actually be under that.

It could technically be 4.75 but we're not too concerned with that. We will actually uh, test it later, but that's at the minimum voltage. And let's say at our Ey out of 500 Millian which is the absolute maximum that the USB is going to give us um at 25 uh, room temperature. Cuz typically this development board uh here in Australia is only going to be pretty much used at room temperatur.

So I don't really have to worry about the extreme ends of the Uh of the curve. Really, Only if you're going up to some sort of in industrial uh temperature, do you sort of have to, uh, worry about the Uh differences in the Uh temperature differential of the load line. So at 500 milliamp, 25 C, we're looking at about um 1.75 Vols or 1.8 Vols or thereabouts Dropout voltage. and of course, uh, 3.3 volts our output voltage plus that nominal 1.8 volts at 500 milliamps.

Uh, that's going to give us 5.1 So we require a 5.1 volt input voltage, which our 5V USB is, uh, well, 5 Vols nominal. So we're pretty close to the margin if we're drawing. In fact, we're technically over if we're drawing an eye out of 500 milliamps. But because I you know this circuit isn't going to go into production, it's only a development board.

a development tool then. and I'm not going to be drawing 500 milliamps out of the thing I'm only going to be drawing a couple of hundred. So I might go down to the 200 milliamp line here. it's only about um, you know, 1.7 volts or something like that.

So really, you know I think it's going to do the job, but hey, let's test it now. I Thought we'd just check the uh performance of this thing to see if it meets the data sheet specs because if you remember I said that we're going to be pretty close to the limit of the Dropout voltage of this LM 317 because ideally I would have used a low Dropout regulator in this application, but I didn't have one left in my jump box I Just decided to use an Lm317 and let's see if it matches the uh Dropout voltage graph, shall we? Now to do that. What we need is the input here powered from my bench Supply uh over there. so my variable bench Supply So that's the 5V input USB I can adjust that and you've seen this before.

I've got my handy dandy constant current uh load so we can adjust the current to match the various curves on the graph and see what we get. and I've got 2 m as always. Uh, you need at least two uh on your bench to do serious work like this. So I've got one.

the Metr hit extra here is measuring the input voltage to the voltage regulator and the Fluke 887 is measuring the output voltage. Now, when you're probing the voltage like this, it has to be right on the input and output terminal. You got to probe it right on there. Otherwise, you could get voltage dropped due to current through your wires or Um or tracers on your board or something like that.

So you have to probe it directly at the input and have the ground directly on the ground pin as well. And I won't uh, go into it. but I've got it. uh all hooked up down there and there it is and uh, it's a bit of a mess, but uh, we should be able to measure the load current graphs.

Let's try it. The first one we're going to try and measure is this 20 milliamp one down here because that's guaranteed to work. We should get just over 1.5 uh volts there at Uh 25 degre C ambient uh temperature. And that's actually not the ambient temperature.

that's actually the Junction temperature. And if you're doing this, um, seriously, then you actually want to get uh a probe onto the uh case of the device. but even that is not the true temperature of the junction. You will have to use the Uh thermal data from the data sheet which shows The Junction to case uh thermal resistance you.

You've got to take that into account to calculate the junction. Anyway, for the purposes of today's experiment, we just want to get a near enough you so we don't have to worry about that. that sort of stuff because uh, we're not going to draw too much current from this so the regulator is not going to heat up much. But what we want is to measure the 20 milliamp load graph there.

at about 25 uh de C we expect just over 1.5 volts Dropout voltage. Uh now, uh, the Dropout voltage is defined as the Delta V out or the change in V out or the drop in V out actually of 100 m. Vol So what we want to do here is adjust our input voltage down until our output voltage drops by 100 m volts. Now as you can see, it's 3.36 output voltage.

and if I adjust the input voltage here, well that's a bit High Oops. Be careful with that. Um, you can see that it it pretty much is regulated. There's no problem at all.

so the regulator's working just fine and we want to adjust that down until that gets down to 3. 3.26 There we go. it's dropping. It could be a little bit tricky.

sorry. 3. Uh yeah. 3.26 So there it is I'm going to say that's near enough.

So uh, 4.8 Vols All we've got to do is plug that into the calculator. 4.83 Let's put all the digits in. we don't have to Uh, minus 3, uh 2. Let's say 26 there and that is 1.54 vol Dropout voltage Bingo It matches the graph there: 1 uh yeah, 1.54 It's just over that point there on the graph so that matches up perfectly.

Oh, by the way, I've set the current to 20 milliamps. Uh, down here the load current and let's repeat that for the 200 milliamp load graph. So I've got adjusted the output current to 200 milliamps there and we expect roughly Uh, let's have a look here: 1.5 or 1.65 1.7 odd Vols or thereabouts. So let's see if we get that, shall we? Let's drop our input voltage until we get 3.26 There we go.

Pretty much spot on. So it's Uh 4958 minus 3264 equal 1.69 or 1.7 Vol So it's spot on. It matches those graphs very precisely. Beauty Now, if this was a production design, we would want to characterize the Uh maximum output current we could take from this regulator under a worst case input scenario.

In this case, the worst case would be uh, 5% below the nominal 5 Vols uh USB input voltage or 4.75 volts. So you would adjust your input to 4.75 volt and then you would adjust your current up from zero until your regulator output voltage here. Uh, got to 5% below your nominal 3.3 Vol output voltage. That's 165 M volts below or 3.13 volts.

So if we adjust our current here, okay, we adjust. So we're looking for 3.13 Roughly, we've got Oh I could tweak the input down a bit, but really, we're we're just fluffing around the edges now. Um, so really, we're looking at what have we got here. Very touchy on the old uh power supply voltage control there, but we're looking at 3.13 volts there.

So really, you're looking at um, you know, at really 160 odd milliamps? There you go. 160 or thereabouts would be the worst case. Uh, A? Well, it would be the worst case output current we could draw from our voltage regulator while we're still within spec at room temperature, and then if you wanted to, uh, characterize it over different temperature ranges, if it wasn't just going to be used in a basic lab environment like this, then you'd have to go to a lot more trouble. And that's what happens when you, um when you're on the margin like this using an Lm317 voltage regul regulator, a standard linear regulator instead of a low Dropout one where we would have had plenty of margin.

Um, in instead of having you know the 1.5 to 2 Vols uh Dropout voltage which we get with the LM 317 if you're use an Ldr might have2 Vols uh Dropout voltage and we'd have a ton of margin and the output current would work right up to you know, 500 milliamps. Not a problem and everything would be sweet, but we're just trying to um, uh, be really, uh, cheap and simple and nasty here and use an Lm317. And if you're interested to know what the quiescent current is for, just the AVR uh, ISP MK I Programmer on its own, that's the Um input current for the USB the 5vt USB It's around about 100 milliamps and it's not doing anything. All it's doing is uh, it's got an LED down there which is flashing and that's about it.

and um, that's the quesence. So really, uh, that means that uh, we'd actually have out of our regulator a maximum of 400 milliamps to power our circuit under test and I Really like this. You can see the fast update rate of the bar graph here flashing in time with the LED. So it's taking those pulses of current as the LED switches off and on.

You can, actually, uh, get an indication where it drops to about 90 milliamps there, or a little bit below, maybe about 85 milliamps up to like 110. but the average is about 100 and you can see the advantage that the Metr hit Extra had on its expanded scale. Uh, bar graph here because the uh, the fluke isn't. Well, it's not actually displaying the average as nice as the Uh Metrahit Extra did.

It's jumping around like a jack rabbit and the bar graph as you can see is really tiny. You can't really, uh, make out, you know it's jumping. Just one segment where there. Whereas this was actually significantly jumping between two valid points which you could actually see and that's because the uh the fluke is a Um has a six.

well, in this case, a 600 count uh bar graph and we're down around 100. and based on the number of bar graph segments you can see, it's It's not nearly as useful as the Metro hit extra in this case, but that's going to vary between meters and between ranges. so it just so happens. the Metro hit extra was far superior in this case.

But of course that's where you put on your min max mode here. So I've put on min max like this and it's you might hear an occasional beep there. which means it's recorded a new value so you don't have to watch the display, just put Max on there and Bingo! We can just look at the Uh maximum is uh well because it's a negative. Current maximum was 92 odd and the positive was 112.

Pretty much what we got on the, but we could see that live on the bar graph here on the Metr hit extra and I've done exactly the same minmax thing here on the Metr hit extra and it's saying it's uh, 87 milliamps. Uh, and we're looking at Um 100. That's the uh sorry, that's cap. minimum is 114 and 87.

So as you can see, it really correlates to the actual bar graph display there. Live. So I don't um I'd have to I'd have to review the specs again, but looks like the Metro hit extra is actually capturing those Peaks faster than the fluke 87 And of course we're not done yet. We still need to check the uh output noise of the regulator to see make sure it's not oscillating for a start.

and uh, well. we know the noise performance is long, it's not oscillating and it's not dropping out. The noise performance is going to be good. So I've got it set to 20 MTS per division here.

and it's not. You know, a problem at all. It's uh, it's quite nice. It's uh, it hasn't dropped out.

it's well within regulation and everything's fine. But watch what happens when I turn down the input voltage here and it starts to drop out. Boom. There we go.

It's starting to even though you haven't quite seen it on there. There you go, you can start. Yeah, there you go. it's starting to drop.

So that's what happens when a regulator uh, drops out of Regulation there. Um, different. Regulators perform in different ways. Some of them will actually uh, perform quite nicely when they drop out so they won't oscillate like this and do other weird stuff.

But we're still talking about not a huge amount, um there. But if we keep dropping that uh input voltage, say to 4.75 or we Chang the output current. it's going to change with the output current. as well.

Um, but as you can see, it does actually um uh, doesn't drop out uh as smoothly as you'd like. but that's the LM 317. There are much better Regulators out there. and if you're keen on the actual details of the Uh circuit, there's not much to it.

Here's the Dave CAD drawing of it. It's basically just a 5V USB uh input any 3.3 volt voltage regul regulator you like I used an LM 317 cuz I didn't have anything else at the time. it's eh, not that great. I'd recommend you use a low Dropout Um and Ldo voltage regulator 3.3 Vol just a a selection switch to choose between them and an Onoff switch and hook that up to pin two of the ISP connector and that's it.

Bingo You've got yourself a modified AVR ISP Mark I programmer. So after all that, am I happy with this? Well, not not 100% No, it's um, there's quite a limit. Quite a few limitations really. un limited to a maximum output current of uh, basically 150 milliamps which is going to be okay for a lot of Uh circuits I do.

but really? um I think it was a it was a bit of a gamble to put in an L lm317 I know I was going to be limited I was hoping for a little bit uh, better but and then than the data sheet uh performance figures but they are pretty much spot on to the to the figures there so it's not that great. I think I might actually replace it um with a low Dropout voltage regulator and just a fixed voltage. a fixed 3.3 volt one so you don't even need uh the adjustment adjustment resistors on there. All you need is the regulator and input and output cap.

Problem solved. Huge amount of margin and well there you go. That was just uh implementing a um simple LM 317 on a design I Got a bit carried away here I was just going to only show you the mod and that was it. It was going to be a quick two-minute blog, but I thought I'd just show you some stuff on uh, some basic uh performance measurements on an LM 317 I Hope you liked it.

See you.