Hi, it's simulation time again. if you remember a previous video I did on a switching tracking pre-regulator for a linear voltage regulator. In this case, it was a Uh switching. Boost converter can be any standard type of switchin boost converter followed by a any type of linear regulator.

In this case, it was an LT 380 we were looking at and uh, we Bas that circuit um on a P Channel mosfet which is M1 here and that came from a recommendation in the Lt380 data sheet and it had a few little uh issues with it. but as it turns out, uh John Barnes has sent me an alternative circuit to that. so I thought we'd take a look at it. uh, but if you haven't seen the uh previous video, click here and you'll be able to watch that first as a background and in the previous video with this P Channel Mosfet Here we had a few issues of choosing the the correct uh type of Uh mosfet and uh, some Dropout issues and we had to uh if you wanted to increase the voltage, you have to put a a diode in series with it or an LED or something like that to boost the voltage up and it it did work.

Um, but it wasn't the best circuit. so John has suggested a uh just as simple circuit but it uses a P Channel transistor here and uh I thought we'd uh check it out and see how it Compares and if we take a look at it, it's pretty much identical almost to uh the mosfet uh circuit We've got our Uh voltage input over here, which might be a battery I've set it to 2.7 volts which might be the low end of a single cell lithium ion uh battery for example. and we're using an LT1 1935 boost converter here. But really, it's just a standard boost converter.

Can be practically any type which operates uh, identically. there's you know, hundreds of them, so there's nothing unusual. There doesn't have to be a linear technology type, can be any kind, and we're using that is the tracking pre-regulator for the Lt380. Again, once again, nothing special.

with the Lt380, it can be any linear regulator a 7805, whatever. Um, and we want this tracking pre-regulator voltage up here to uh be uh, in this case, around about 2v higher, always 2 volts higher than the output. and if we have a look at the circuit, uh, as it turns out, R4 will be the resistor which controls basically controls the Uh offset uh voltage here and it should be fairly independent of the base current and uh, really, we should have no Uh issues at all with the Uh particular type of transistor used and our temperature coefficient for this circuit will basically be dependent only upon the Uh temperature coefficient of the base emitter Junction here. So let's uh, try it and see what we get and you can see that John's original circuit uses C1 here.

Um, that's obviously included for some sort of stability issue I'm not convinced that we, uh, entirely need that, so we'll try it with and without that, but uh, we can adjust a few parameters here and we're going to use the parameter sweeping feature which I've done a video on before. So if you haven't seen that, click here and you'll be able to watch that uh video first. so you'll know exactly how we're doing this. And of course, we're using LT uh Spice Again, it's a free circuit, uh simulation tool and it's uh, uh, probably one of the, uh, industry standard, um, simulation tools.

Now it works really well. I Highly recommend it. So let's uh, run this thing and see what we get. So what I've got here is I've got V2 this voltage source here which sets the output voltage and I won't go into details of how it does that.

you got to watch previous videos, but I set it to 1 volt here and which means we'll get one volt on the output. So we're looking for 3 volts on our tracking pre-regulator input here, so let's run it. Let's go into simulation. Edit: our simulation command: We're doing trans Transit analysis.

Uh, our stop time is 15 milliseconds so let's just run that. pretty basic stuff and you can see that there's some uh, startup stuff there around about3 milliseconds or thereabouts to start up the regulator, but that that looks like it's working. A treat that's at about 2.9 volts or thereabouts So let's have a look here. There's our output voltage.

Of course, it's not stable at the start there, but once it gets there, it does stabilize out and there you go. Uh, we're getting basically uh one VT out and and if we compare those side by side, it's pretty close to Uh 2 Vols offset like that. So that's uh, working. a treat.

No problems at all. So I've changed R4 to 20K here and let's run that now and give it a look. There it is. we're talking uh, what are we talking here? We're talking four volts or there about.

So that's 3 volts above our 1V output voltage there. So uh, changing that R4 value from 10K to 20K doubling that value has increased our voltage from 2 volts to 3 volts. Now, if you're wondering why this uh simulation takes more time on the 20K 1 as it did on the 10K 1 and we get this, uh, transient stuff on the input. It's because the DC to DC converter is now working because we've only got a 3.3 volt input voltage here, let's stop that.

So I've only got 3.3 volt input voltage and before we'll getting a 3vt tracking pre-regulator which means this is no longer operating as a boost converter. So when this voltage goes down, uh, when our output set voltage is only 1 volt and our configuration here is giving a 2 volts above that trkking pre-regulator that's below the 3.3 volt input voltage. So if we up this, uh, so if we, uh, increase this value again to say 5 Vols then you'll see that it will, um, instant. So our input voltage is now 5 volts and our tracking pre-regulator should be 4 volts.

so it's below the V in input voltage and let's run that again. and we should find that there's none of that transient stuff at the start bang in this simulation is instant. That is because that that DC to DC converter is no longer operating as a boost and it's got less to uh, analyze. it's going straight through and bingo we get our 4 volts, um, offset voltage or thereabouts.

So what I've done now is I've gone and changed the output voltage to 5 volts. the input voltage is 3.3 Vols. So uh, we're always going to get uh, this thing to work in that boost operation. Our track in pre-regulator should be 7 Vols cuz we've got our 10K value there and we run it and you can see that our output voltage in the green.

there is 2 volts above our output voltage of 5 Vol So it's tracking just fine and you can see it stabilizes in you know, 0.15 milliseconds or thereabout so we don't need to run all that simulation time cuz that actually that'll take some time so we can knock that down to you know, let's say knock it down to 0.5 milliseconds or thereabout so we can, uh, now run that again and bingo it. It just doesn't take us long. So now we can, uh, do some parameter sweeps of uh, let's try our base resistor R1 here and see what effect that has on our tracking pre-regulator voltage. But before we do that, let's just have a look at C1 here and this is the waveform.

Okay, 5 volts out 7 Vols Tracking Prg and we've got the these Wiggles here that's with C1 in the circuit. let's get rid of C1 So let's go down here, delete C1 and let's run that again and see what we get Much cleaner. Start up there without C1 So I'm going to leave C1 out and uh, let's parameter sweep R1 and to do that is real easy. We go in here instead of 22k, put in the curly brackets which indicates to the simulator that it's a parameter.

we'll call it RP just a label and then we can go up to our spice directive. Up here, we can add a spice directive, do step, command parameter, and then the label we gave it which was RP and then the value we want to sweep over. Well, let's go from uh say uh oh, I don't know. Let's go 1,000 ohms up to 20K in 1,000 ohm steps.

so that'll give us 20 waveforms so we'll sweep through. Let's put our parameter uh, our um spice directive on the circuit there and run it so it'll sweep through. This R1 value will go from 1K to 20K in 1K increments. So we should get 20 different waveforms.

Let's run it and oh, let's there we go and it's stopping and bang. That's 1K So that green one was 1K This blue one will be 2 2K red one will be 3K Looks like it's making absolutely no difference at all. Now let's do a parameter sweep on our output voltage. So V out here I've uh put I've called it V out actually and we're parameter sweeping there.

it is our do step command V out from 1 volt to 10 volt in 1vt increments should be interesting. Let's run this and see what we get. Let's have a look at our output voltage bang. So there's our green line: 1 Vol blue volts 2 Vol blue lines 2 volts Reds three four five bang.

Look at that and you'll see it takes longer to um, uh, stabilize at the start. of course the higher voltage you go, but it's drawing a very interesting parametric graph there for V out I Like it and if we have a look at our track in pre-regulator voltage that is track in you'll know notice it's tracking 2 volts above each time. Very nice and we can just get that on its own of course. And there it is.

So we've got 3 volts and then it'll be 4 volts 5 Vols 6, 7, 8, 9, 10, 11, and 12 volts. It's working a treat. What I'm doing now is I've changed the transistor from a 32 and 3906 to a BC 327 and uh, once again, it's working just fine. Not a problems at, not a problem at all.

So by now you're probably uh thinking this circuit is rather interesting and exactly how does it work? because this base resistor here has no effectively no effect and doesn't set this pre-regulator voltage. And we can show that if we set it to you know, 0.01 Ohms for example and then run this thing. there's our output voltage and there's our tracking pre-regulator voltage: 7 volts, 5 volts. It doesn't matter what that base resistor value is, it can be anywhere from a dead short up to you know, hundreds of K and it's going to work just fine.

So there's nothing that base current is not setting the voltage. What is? Well, if you look down at the feedback pin down here. If we look at the voltage across R2 down here, let's go in. it is 1.25 volts.

There it is. and that's no coincidence because is the band Gap voltage reference inside the DC to DC converter and they're all the same. They're all going to be around that value because that's the physical construction of the band Gap voltage reference in there. So that's effectively a constant current Source through R2 there R2 And because we got a constant voltage across that, 1.25 volts across 10K that sets a constant current and we can actually look at that if we run it.

Let's look at the current through R2 Here we go there it is. It's 125 microamps. No surprise. 1.25 Vols on 10K is 125 microamps and you'll notice that if we look at R4 the current through R4 is exactly the same.

There's the two of them laid on top of each other because the base current does not contribute at all. So the current. So the Uh pre-regulator voltage is equal to the constant current set through R2 which gives a constant voltage drop across R4 here. So let's shut that down.

expand it. so the voltage pre-regulator voltage here is equal to the constant current through R4 and the voltage drop across that plus the voltage drop of the base emitter Junction here and that's it. That's what sets the value of the pre-regulator voltage. so it depends on the constant current set up here here, current through there, the voltage drop, and the basy M.

So this circuit is going to be pretty darn good and the only issue there going to be with it is the temperature dependent dependence of the Bassy meter Junction here. so we can actually try that. I've set it back to a 23906 Uh transistor I've added the Uh Spice directive doep, uh temp. So we're going to stem uh step the circuit temperature from 0 to 50 C in 5 deg C increments so our base resistors set back to 10K not that it matters and let's run that and have a look at our tracking pre-regulator voltage with temperature and here you go.

it's you know, there's not much in that in that at all. that's a for 5 Celsius jumps there I mean we're we're talking. You know we're talking nothing really. There's very, very little in there.

So really, where we're only talking about, you know, 100 MTS or so there over that sort of uh temperature range 50 de C And that's not surprising because the typical uh Basy Met Junction uh Tempco is going to be about 2 MTS per degre. C So there you go. I Uh, quite like that circuit, so thank you very much. John um I think I'm going to, uh, use that one.

It's better than the Uh mosfet uh circuit recommended in the LT 38 data sheet. seems to work a treat and uh, if you, uh, want to discuss this, jump on over to the Eev blog forum and if you like this type of video, please give it a big thumbs up. Catch you next time.