Hi I Got this little love voltage regulator block designed for use in model airplanes. Got in a previous Sir Mayor bag segment. took a quick look at it. am I goofed that up at the time? So I thought I'd have a more detailed look at it.

da Keith Sent this in and him and his Model Arm airplane buddies wanted to know if this thing was any good. It's from a company called Castle Creations and it's the BEC Pro BEC as a battery eliminator circuit. It's designed to power these solenoids and the receivers and other things in a model airplane from the high voltage model airplane battery itself, so you don't need a secondary battery and things like that, so you wouldn't know This thing's cheap. It's like thirty-eight dollars or something, which is apparently cheap for one of these things and he wanted to know if it was any good or not.

We took a quick look at it in a previous mailbag, but I thought we'd actually check the output power actually power the thing up and see if it's any good, because it's supposed to have a content maximum continuous output current anywhere from fever from eight amps up to 15 amps here depending on the input voltage, and it has an adjustable output voltage as well anywhere from four point eight volts to twelve point five volts. And if you take a look at the figures here, that this thing claims the input voltage for any where it's specified anywhere from 16 to 48 volts depending on what type of battery you've got In this case, that you know it tells you like an eight cell battery 10 cell and a 12 cell. That's what the S4 12 in series stands for and so can go anywhere up to 50 volts. And of course it's going to have a varying output power capability.

So here is the output current and a key said that he set this one to 6.5 volts. so I don't know how to adjust it. We'll just keep it set to six point five volts for the purposes of today's experiment and these are the maximum continuous output powers. Just multiply the 6 point 5 volts times the output current here at different input voltages and it can be anywhere from 50 to white so to 97.5 watts in this tiny with this tiny little heatsink here and as I mentioned in the previous mailbag, you know that sounds a bit suss because the heatsink like this is only going to have a certain amount of power dissipation capability roughly I Believe about it for this type and size one about six degrees C per watt.

So for each what dissipated, it's going to increase by six degrees. We'll go into details in a minute on that and the these are This is a DC to DC converter of course and it's going to have an efficiency of you know at best 90 percent at one of these particular things. They would have designed that for one particular output current. because you can design DC to DC converters with 90% plus efficiency if you choose the correct type of magnetics ie.

the inductor here, that's what this big device is. You know the MOSFET, the output capacitance and the switching frequency, and you can actually design it for quite reasonable efficiency. So let's say it's ninety percent if you're getting near a hundred Watts Here, let's just say it's a hundred watts at ninety percent efficiency. This thing has to dissipate ten watts.

so that's going to cause a temperature increase in our active semiconductors that are dissipating. The power here are namely, we've got our two MOSFETs here. these are d 4 o' nines. You might think this is the MOSFET, but it's not dead Giveaway here is not connected an A4 anode and the other one would be the cathode here.

that's the diet. And of course the inductor here will be dissipating power as well. But because that's not an active a semi conductor with a maximum Junction temperature where the magic smoke indicate escape, that one's much more Hardy out than the other. So we're really concerned about these devices dissipating power and because the heat sinks on the other side of the board and they've just got vias underneath these going down to the other side, actually transferring the heat from one side to the other.

We're going to get losses in this. So you've got your Junction to case losses and then you've got your case to PCB. Then you get your PCB. from top side to the bottom side those losses in the vias.

Then you've got your contact thermal resistance with the heatsink and then got the thermal resistance of the heatsink as well. And I've done the videos on thermal design and things like that so I'll link those in if you want to watch. Check them out. But yeah, we're really concerned about these heating up with such a small heatsink.

with a large power dissipation like this 50 to 100 watts, you cannot design it to have 90% efficiency at each of these input voltages. When you design a DC to DC converter like this to operate a over a wide voltage range and a wide output power range. Due to various factors in the design, you just cannot get a general purpose wide input voltage and a wide output current range to have a you know Universal high efficiency across all the input voltages and all the output powers. So if we have a look at the data sheet here for the actual device used in this which is the LTC 38:24 which is a high voltage, our step-down controller goes up to 60 volts is exactly what you would use here.

It's a pretty decent arc controller and take a look at the efficiency and power loss verses load current graph on the right there you can see that the top one with V in equals 12 volts. Look at it approaches that 90% efficiency that I was talking about there. but at higher input voltages V in equals 40 volts. It's a different curve entirely.

And we're looking at say you know 40 volts dependent like add a couple of apps for example, but this you know is based on which particular transistor you pick and and inductor and everything else right? Because this doesn't have the switching MOSFET built in, but you know you can see it drops down to you know it could be well under 80% You know 75 percent efficiency would be fairly typical of something like this at a higher voltage unless you specifically designed it for high voltage use only at one specific input voltage. And that's a disadvantage of a design like this is that you know the efficiency is going to vary quite drastically over that input voltage range and output current range as well. So it's always going to be a compromised. which is why the specs for this thing actually show as the input voltage Rises you saw on the specs on the back of the thing that the Mac some power dissipation in it drops the efficiency of a typical DC to DC converter and it's going to have something like that.

There's going to be a peak voltage and current where it is you know at its most efficient, but at lower than that its efficiency is going to drop off. At higher than that the efficiency is going to drop off. and when the efficiency drops off, you dissipate even more power in your heat sink here. and well, that ruins your day.

So let's see if this thing is actually can meet its claims here of a maximum continuous output current. Let's go I Just mentioned briefly though, the capacitors because these will be a big failure point in these things. If these are wet electrolytic capacitors, then they can dry up and you know over time and of course the output. the ESI increases and their life expectancy goes down So pretty terrible.

We've got 105 degrees C rated Elektra over here. This is on the input and our to output caps. These actually don't look like wet electrolytic dielectric material. They actually look like a polymer capacitor or what's called a solid electrolytic capacitor.

ie. there's no liquid inside them, but they. that doesn't mean they still can't They still don't have a maximum lifespan and the I'm not 100% sure that they're solid polymer types, but you would have to look up the specific data sheet. I Didn't have anything luck with that particular part number and ever.

but it's got no vent on. There's no like score marks in the top for venting and that if you see those then that's a dead giveaway that it's at a wet electrolytic capacitor with your traditional electrolytic capacitor that's going to have a shorter life span at higher temperatures than us and then the solid types. and it's a myth that all of these surface mount types are solid. That is not necessarily the case, and just because it doesn't have a score mark event in there on the top, then that still doesn't mean guarantee that it's a solid type.

This could still be a wet electrolytic type anyway. just thought I'd point out the difference there. Anyway, it's given the benefit those are properly SPECT solid capacitors. and by the way, that nice shine on there is a conformal coating, so that's very nice.

That's something for something that's used in like an outdoor RC plane. Anyway, let's hook this puppy up and see if we can measure it, measure its performance. and they've basically got the wiring on this a bit back to front. Look at the nice beefy 16 gauge wires they've got on the input here, but the input is going to be lower current then your output.

The output thing as you saw this thing has a maximum of 15 s and then you got. Well, we've got two wires in parallel here. Okay, so it may do the business, but you know we've got these these tiny little piddly connectors which are going to do the current. You can of course put them in parallel like that, but you know, 15 amps.

These are going to get maybe 5 amps each. So you know, like chops. So really, you know you can't do the maximum 15 amps with just These two. Anyway, let's cut these off and have a look at the wiring inside here to see if it's adequate.

Oh yeah, that's that's. pretty decent. That's pretty decent. There you go, no worries, you put two of those in parallel will be hunky-dory So to test a regulator like this, we need a power supply.

of course. I'm using my Ro Goldy P83 2. Now, this is only maximum capable of 30 volts, 3 amps per channel. but because it's got two channels or actually three here, we can put them in series.

So that's why I've got this wire here just looping the positive and negative like that and then we can get a maximum of up to 60 volts so we can go right up to that 48 volt input capability easily. With this, and the good thing about this is that it has power output as well. But of course any decent power supply is at least going to display voltage and output voltage and output current. but this will display our power as well.

So we know our power input to the module itself and we know the power output and the other bit of gear we need is an electronic load. You can actually use a resistive load if you've got those in Cobble them together. but a a nice precision electronic load like this be Precision 8601. Can't beat it.

And we've got and we can set the output current. So what? I'm going to do is actually set the output the constant output current to in this case, eight amps and it'll tell us the output power and in this case it is actually telling us the output voltage. But we need to do one more thing with this setup because these output leads aren't really actually quite wimpy. There's at like high currents like we're talking eight to fifteen amps here and we're gonna get serious voltage drop across these leads.

So we're actually going to get power dissipated into that and our figure on here will not be accurate for our output power. So I'm conveniently we've got the second one here and we're going to actually wire it into the back of this BK Precision unit for the remote voltage our sense terminal. So it's going to sense the voltage directly on the output here. So at the moment because there's no load that the output voltage at the terminals here is going to be exactly what the output voltage here.

So there's no current flowing at the moment through these cables. so there's going to be no voltage drop, but hey, there's it's gonna be very significant. Alright, let's test that. I've set both channels here to 24 volts.

so that's going to give us a 48 volt output and if we have a look at the data we had before 48 volts, its maximum rating is 8 amps are continuous. So I've set the 8 amp up here on the electronic load up here. We'll switch it on in a second and it should. What we claims to have a maximum power dissipation capability based on the output current here around about 50 - what's let's just call it 50 watts or thereabouts.

So at 48 volts input, you know at a hundred percent efficiency. and with 50 watts output, we need to supply. This power supply needs to be capable of supplying 48 volts at 1 amp. To give that 50 watts, it's easily going to do that.

It's got three amp output capability. So I set the maximum output current here to three years because look, we don't need to protect this thing. I Don't care. You know it's not the design phase.

This is supposed to be a finished product that's supposed to work. So I could set the limit. You know if you were testing a design your own decid in this you know just to be safe you would set the output current to 1.1 amps per channel or something like that. But you do have to be careful setting a low safe in quote marks current output limit because you can get a certain like power on surge currents and that may up and then this.

The power supply that you're powering your converter with may go into current limited motor drops of voltage and then that causes you know a problem with you converter and well that can ruin Your day. So anyway I just got to set to 3 amps. Let's go. So what I'm just going to show you now is if I don't hook the voltage sense lead up to the rear voltage sense terminal.

This you will know you'll see how much voltage drop we get across these cables at this eight amps continuous current. So let's turn our load on. We've got ya. Everything's right, let's switch it on and we're getting six point Six Point Six volts at the moment.

So let's switch it on. Bingo is dropped down to Six Point One Eight volts. So it's reading Six Point One eight volts directly on these terminals here and so there's going to be some loss in these leads. I can actually measure the other two wires hooked under here and see what we get.

and you can see that if I use my meter here to actually probe directly onto the output of the brick here. So I'm measuring just directly on the output terminals, you'll see it is still holding at Six Point Five Five volts. so it seems to be handling that 52. What's just fine and still giving our set output voltage, but you'll notice that it's only reading 6.1 up here.

That's because of the loss in the wires. So if we take these two wires and don't short them out, if we take these and plug them into the rear sense terminals and switch the voltage sense from the front terminals to the rear, we'll be able to get a more accurate reading. And this is actually important because you'll notice that our BK precision load up here is calculating and output power of Forty Eight Point Nine, Three Watts. And that is a very significant error because our meter here is showing that it's a Six Point Five Six volts here.

So if he goes six Point Five Six Volts times the eight amps which this is actually measuring. okay, then that's going to give us a value of 52 Point Four Eight Watts. That's the true 52 and a half watts is the true output power being delivered from this module. But we've got that error.

Very significant error up there on the BK precision. so that's a trap for young players. Make sure you use the external reference input I'll show you how it works. and here's the sense terminal on the back.

You can actually see positive and negative sense terminal. so just hook these wires directly to the output terminals of the power supply module that you're testing. so we'll turn our remote sense terminal on. Bingo! So let's try that again, Shall we? Eight Amps Constant Current Load on this thing.

Six Point Five Nine Volts with no load. Turn it on. Bingo. It drops down to Six Point Five Eight.

Exactly Five Five. Exactly what we saw on the multimeter before. So that's a big trap for young players. And Bingo, it's showing the 52 Point Four Six Watt.

So now we're getting an accurate measurement trap for young players. Voltage drop on the output leads. Now, just because this thing works, doesn't mean that it's any good. Okay, it's outputting the set.

Six Point Five Five Five Volts. No problems at the rated eight amps. Continuous output for a 48 volts input? No worries. But what's the efficiency of it? How hot does it get? It isn't gonna last five minutes.

Isn't gonna last five hours. Last 5,000 hours because it may shut down any second because the thermal overload. We don't know. So if we have a look at the input powers here, look: thirty Five Point Two Five Watts Thirty Five Point Two Five Watts.

Okay, so we add those together because we've got serious well. hang on folks. You'll see something certain to happen. The voltage is going up the input powers I Think and hang on.

Hang on. Yeah, that's starting to smell pretty toasty. I Think we've got a problem. Anyway, we add up these two powers here.

That's our input power up. Bingo. Well, it just died. I think Yep, Yep, it's cutting out.

Its cutting out. This thing cannot handle that with a free air heatsink and that's what I was about to get to. What we need to do now is measure the heat sink temperature here and see what we're doing. We can use a Director thermocouple probe, but hey, I've got a flurry 8 here so we'll give that a bill and as soon as it's powered up here, it goes hunky-dory Let's measure the center here.

Hopefully you can see that a hundred and sixty degrees. Yep, it's way too hot. I'm switching that off now, so that is ridiculous folks. A hundred and sixty odd something degrees here.

Hopefully you saw the temperature up here 161 degrees with free air, but that's what I would have expected was such a tiny heatsink. It's not going to work in free air like this, but hey, if you have a look at the data sheet for this thing, it actually tells you that it's rated for a certain particular airflow over the heatsink. And of course, the thing is, if that heatsink is getting to 160 degrees, then imagine what the junction temperature in the poor little MOSFETs in there I Get into. It's like these things will shut down burn up the magic smoke escape, so it obviously cannot do.

It's rated current with no airflow, but that doesn't surprise me at all. and we are actually measuring this out of its spec. We need an airflow over it Anyway, let's take a snapshot here. Let's go 34.8 what's double that? and then we can calculate our efficiency.

So what we've got here is sixty nine point Seven watts input power. Fifty two point Four watts output power. Which you saw here like this. So that gives us just divide Fifty Two Point Four by Sixty Nine Point Seven input output power over input power and we can get our efficiency here.

So you know around about let's call it 75 percent efficiencies. So you know that's fairly typical of what you'd expect of a DC to DC can wide range DC DC converter like this, but that means that heat sink that poor piece out little heat sink there. It has to dissipate a lot of power. How much power does it have to dissipate? Well, you subtract 52.4 watts from Sixty Nine Point Seven Watts and that Lisa was seventeen Point Three watts.

It's got to dissipate in that tiny little heatsink. We can have a look at the data sheet for the well, not this exact heat scene because I don't know the exact heat seem but I got one. The nearest I could find was this one. It's basically the third one down there on the list.

the Six Twenty Five - Forty Five. You can see the thermal performance graph here for the heat sink to ambient thermal resistance in degrees C per watt on the vertical axis there versus airflow on the horizontal axes there. And we're looking at around about our four hundred and forty Lfm here for the starter sheet spec of five miles per hour air flow. So if we actually look at the graph here, it's the third, it's the third one down.

And then we take four hundred and forty LFM and we then extrapolate that back across. We get a thermal resistance of the heat sink of about eight degrees C per watt. Eight degrees C per what is quite large. And we remember we said before that we had about seventeen point three watts dissipation in this thing due to the seventy five percent efficiency at this particular input voltage.

that Eagles multiplied those that equals a hundred and Thirty eight degrees C Temperature rise above ambient fam beings Like twenty degrees. We're looking at a hundred and fifty more than Hundred and Fifty-eight So pretty close to what we got there. Even with the No. even with the airflow over this thing, the nominal rated airflow notice we measured 161 degrees C So maybe the heats in we're got slightly better than that because we were getting no airflow.

But still, this is way above the maximum junction temperature of the semiconductors. And if we actually have a look at the MOSFET used here the Ad 409, then we can see that the maximum Junction temperature absolute maximum Junction is storage Canton Temperature 175 degrees C Of course we are gonna. You might say well that's under that. Well, no it's not because you have to look below that at the junction to ambient.

And what a junction the case. And then you've got a couple of degrees C per what there you've got to add it on. Then said you've got the videos and check out my other video which I've linked in week and I go through the detailed calculations and show you how to do that so it's certainly going to add up that this thing. There's no margin in this at all.

if it does work, it's going to be extremely borderline for this particular performance. And you can look at the actual diode used here as well and look at this maximum junction temperature right down the bottom there plus 150 degrees. C So once again, it's actually going to be hiding this if our heat seems already at 161 in not with any airflow over it. but you've got all these Junction the case and then the vias going through and then the thermal resistance of the insulating conductive pad in there, the thermal pad, and everything else.

It's just no. it's really borderline and because everyone's going to want me to do it. What happens when I blow some air over this thing? But let's find out. I've got this little art soon on fan here and the five miles per hour is 2.2 or there abouts meters per second.

So I'm going to get my anemometer here and I've said it. so it's around about yeah, you know, near enough to that. So now we can blow our rated or you know, reasonably close to it, our rated five miles per hour airflow over this thing so we can repair it up and measure it again and check it out. This thing's actually doing amazingly well now.

It's always amazing what air flow can do. We look at it. Ah, what's 70? 72? Something like that? Over 70 degrees? That's certainly now, certainly within a decent ballpark. So yeah, I'm pretty darn happy with that look.

They're boards on an angle there, so I'm probably blowing across the parts as well, so that's probably not the best. So what I'm going to do is for it because the whole idea was just to just to do the heatsink. So I'll flip it on that side where the board is down and see if that, see if it increases. I Probably would expect it to increase a little bit like that when there's no now no airflow on the bottom.

It's only under over the parts, which is certainly going to, but just over the heating. so let's have a look at that now. I've had it going there for a while and that's actually rather surprising. I'm getting under 70 now I Would have expected that to increase, but I guess more airflow.

Yep, turns angled differently now over the fins. So I guess it's it's more efficient and if we try and have a sneaky peek under like that, yeah, we've got the board on a bigger angle now. But yeah, we can. Like the case of the MOSFETs for example, you get in there where let's call it a hundred degrees C but that's probably still okay.

Yeah, I don't mind that at all. special. For the amount of power that we're delivering, then that we're dropping, That's yeah, it's gonna do the business. So at 48 volts, we're cooking with gas.

but does it work down at 16 volts input for a continuous output current of 15 amps? Well, I've changed my our power tech MP 3090 switch mode supply here it can only go up to 15 point 3 volts. But I'm gonna, you know, say that's got to be near enough. Anyway, Well, I've got the constant current output set to 15 amps. We've got our six point five seven volts Exactly the same as before.

Let's see what it does. Wah wah wah wah stroke down to one volt. It's not, it's dead. So let's say go to 12 amps here and see what it does for 12 amps.

Oh, there we go. And yep, she's working at 12 amps. We're getting the output power out seventy eight Point five, what's there? So yep, I'll leave that for a while and get some temperature. now.

hang on. It's starting to drop, starting to plummet. No, no, it doesn't like that. Doesn't like that at all.

Nope. Nope. it's dropped out of regulation. wham.

Gonski Anyway, as expected, it's much more efficient down at this 15 point two volts 5 amp. So looking at 76 what's input power sixty five point five watts output power about ADC percent efficiency there at the lower input voltage. and because of that, increased efficiency is running at a much cooler 56 degrees on the heatsink. There Once again though, in free air.

Noop, you really want that 122. Nope, Nope Nope Nope Nope. Well, you really want the airflow. So if you're using this thing for its intended purpose for a model aircraft, you want some sort of, you know, like the heatsink I Don't know.

Poking out the bottom I Don't know where this would be mounted on a typical aircraft. You don't actually want it like inside the fuselage with no airflow and just like trapped like that because it's not going to do the job you really need. the airflow over it makes a hell of difference. Anyway, the reason that we're seeing the discrepancy versus the datasheet.

Like the cool temperature we got on this operating versus the datasheet that I pulled out. This looks like it's about 4 degrees C per watt whereas our datasheet said about 15 degrees C per what. Obviously you know there's like just the design of the heatsink. It's a similar dimension.

everything else. but I Think the fins are actually larger on this one, so larger surface area I Think the datasheet one has smaller surface area fins on it. So yeah, that can probably double your thermal resistance of it. And then that makes a huge difference.

If you just looked at the datasheet value, then you would think no, this design is not viable. They're You know they haven't left adequate margin, but when you actually you know, do some tests on this. Granted, they're not massively controlled test, but they're you know I'm gonna give you a decent back on the envelope performance figures, Then you know this thing is actually a pretty reasonable design at the higher input voltages. But as you've seen here, I'm only I've been operating now 10 amps output current by the way.

so 65 watts output power to bet. Note it basically from what I see refer the testing required of course, but it basically does not meet that. 15 amps at an output continuous output current at 16 volts. I Can't even do it momentarily by the looks of it.

So yeah, I don't know what's going on there. further testing required, but at the higher input voltages? certainly? Um, it worked just fine. and there's reasonable margin on the design as long as you got the airflow. So there you go I Hope that's actually finally answered Kisa question.

This video is longer than what I thought it would. It's gone for about half an hour or so, but hopefully this is like a tutorial on sort of how to rough-and-ready tutorial. How do I characterize a DC DC converter? But like we, there's a ton more stuff we can do with this. We haven't measured it over the full envelope range and we haven't even checked the output ripple.

And you know, like tons of stuff we could. We go recharacterized this at every input voltage over every output current. Get characteristic curves for the thing and you can spend days and days and weeks and weeks actually characterizing just a one brick-like one regulator brick like this if you want to do it properly so we've just done. you know, as a couple of simple spot checks Today, higher input voltages seems to work just fine given sufficient airflow.

So anyway, I hope you found this useful and I will link in at the end of this video. The extensive Our Design tutorial runs for about 22 minutes as well. Worth watching on SMD Thermal Design how to actually dissipate power in SMD Packages like is used on this one and it takes into account the via thermal resistance. It goes through all the calculations and graphs and everything else so well worth watching anyway.

If you enjoyed it, please give it a big thumbs up. As always, links down below - yeah or the forum all that sort of Jess And don't forget if you want to support me, there's a Patreon link somewhere it down in the bottom corner right at the end of this video. Thanks to everyone who supports me on Patreon. Awesome! Catch you next time.

The copper will have a thermal resistance to, but it can get quite complicated so I'm just assuming there's no loss in there in that copper itself. The next one is the via. The heat actually actually has to remember that 10 watts of heat or whatever it is has to transfer through the via. It's going to have a specific thermal resistance and then it's got to get through that SIL pad that we put in there.

That SIL pad will have a thermal resistance to look up the datasheet for it. It'll tell you what it is typically and then we're going to have the thermal resistance of the bar here.