Hi in a previous video where I showed off this very cool DPS 303 power Supply module. It's complete lab power supply in a tiny little module like this. It's all fully self-contained and I'll link in that video down below if you hadn't seen it and it's a real cheap and easy way, it's like sub $25 delivered for this module and you can turn it into a bench power supply. Just voltage in and voltage out.

Fantastic! And here's the specs for the things. Just as a recap: 6 240 volts input It is a buck only So you have to have an input voltage range at least the dropout voltage greater than the output voltage range. You want the maximum voltage range of 1 to 0 to 32. So for example, if you only needed like a 12 volt a 0 to 12 volt lab power supply, then you could get away with saying like a 1314 say 15 volt input something like that.

So output current 0 to 3 amps, 96 watts which I'm dubious about based on the you know the efficiency of the heat sink in this thing. but that's one part of this video. I Want to characterize this now because so many people said that they want me to characterize this and see how efficient it is over various ranges. So that's what we're going to do anyway.

Very nice. 10 millivolt resolution, 1 milliamp current set resolution, and adequate point 5 percent you know, plus one or two digits voltage and current accuracy. So what do we need to characterize this? Well, you can do it with your multimeters. this? is where I've always said having 4 multimeters is very handy in your lab because you can characterize not just modules like this.

You can characterize your own design. DC to DC converters you can measure. Yeah, you have two multimeters for the input measuring the voltage and the current ie. the power going in to your converter and then a 2 meters measuring the voltage and current on the output ie.

the output power. When you've got input power / output power, you can get your efficiency so you can plot efficiency curves over various voltages and whatnot. So very handy to have for supplies. but we're not going to use multimeters today.

We're going to use the Roy Goldy p83 to power supply. This is like 0.05 percent class instruments, so it'll allow us to know the input voltage we're feeding in and the input current as well. Near enough. I'm not going to bother worrying about like the drop across the leads here.

it's like it's going to be near enough. Um, this thing does have an external sense so maybe I probably should take some wires around in the back I don't know anyway. and we've also got the Rygel DL 302 one, DC Electronic load and electronic load. I've done videos on do-it-yourself electronic loads.

You can build these really cheap you don't have to spend. you know, 500 bucks for a commercial one like this. but once again, very accurate. Point: Oh, five percent class.

measuring the output voltage and the output current as well. And this one. I Do have the sense output so I'm feeding that directly back to these terminals. I'm not going to worry about the once again, the drop across these output cables here.

it's you know it's gonna be near enough because we're not going up to particularly high currents anyway. I've enabled the external sense there. and if you want to know what the dropout voltage is with no load anyway, I'm feeding 32 volts in and I'm getting a maximum of 31 volts out of this thing, so that's not bad. 1 volt dropout with no current load by the way in the previous video are completely overlooked.

the fact that you could set the voltage and the current just by pressing these keys I Thought well I just went in to set and you know it was a little bit Dicky to go in there and select your digit and all that sort of stuff. It was a little bit tedious, but you don't actually have to do that. You can. If you want to set your voltage, simply press V there and bingo, you've got your control and it.

it does seem to have maybe like it goes in 10 Mille volt increments there. It does maybe seem to have a little bit of velocity control, so I can turn that relatively fast. not as fast as ideally like, but it's still quite usable, so very easy to set your voltage and current like that. So the user interface is really my good and it permanently displays your set voltage and current at the top, which I absolutely love.

All power supply should display set voltage and set current if possible, and also the input voltage as well, just so that you're aware if the input voltages is drooping or anything like that. Anyway, let's get on to the characterization. Now, there's two ways to I can get an efficiency characteristic I curve of a power supply like this one is to fully automate it. and these two instruments probably ideal for this task because they're both our Ethernet LXi enabled.

So if I hooked them up to network the PC you could write a script that outputted your necessary voltage from here. Of course you can read, you'd be able to read back your input current as well, so you to be able to read your input power going into your module. And of course, you can control the electronic load, set it up for any constant current load you like, and then sweep it over there, and you can run various sweeping runs and get all the data points at at basically one milliamp resolution or whatever the resolution of this thing is. But unfortunately, even though you can fully automate these instruments here, this one down here is not.

So you're going to have to actually set, you know, turn the knob several of from the same company you manufacture this. they do offer While you know, Wi-Fi enable ones, Bluetooth enabled ones and all serial enabled ones. So you can actually get digital control of these things so you can actually for virtually hardly any more Cost just like five bucks or something. Esther? I don't know.

Five ten bucks extra. You can get the wireless or serial enabled version of these modules. I believe in the bigger one. I don't think it comes.

not sure if it comes in this smaller one. so yes. well. I could semi automate this thing by like I could sweep the load for example for a particular output current and then I could get a characteristic curve of the load for each particular output voltage and get a whole set of characteristic curves and the efficiency at each point for each particular output voltage in each as over the and sweep from zero to three amp output current and all that sort of stuff.

But hey, I'll just get a cup of ones you know at like low output voltage like five volts and then twelve volts and then 24 volts. Maybe you know those three characteristic curves might be OK for a zero to three amp current sweeping. I don't know, You know 100 milliampere incrementally like that. Good enough when you're taking these measurements, manually writing them down on paper, you know it takes time.

So I'll just I'll just get you know, fairly crude plot. It's going to be good enough. It's not like the characteristic plots are gonna sit like you know, curve up like this and then suddenly go boom and then ugly. You know, like they're they're going to have a characteristic shape.

You know, something like that. Typically there's going to be a peak efficiency, so you know I Expect this the efficiency that this thing are based on. just some, you know, temporary feeling, the heatsink on the back and stuff. It's going to be like greater than 90% at its peak.

So it'll be interesting to see how efficient this is. over the full range for a couple of input voltages and a couple of output voltages. So I just wanted to show you what the noise looks like here when you go from no low which we've got at the moment. basically 20 millivolts peak to peak there, and we actually switch on the full 3 amps at a 24 volt output.

There you go, it jumps up to like 110 millivolts or thereabout, so you know it ain't pretty. So if you're building a supply out of these things, you probably want to do some. at least a modicum of output filtering would be nice. Ah, Bugger! I Decided to go to town I just took apart my leads there and now I've got the sense and the load leads going directly into the Phoenix Contact connectors.

and I've got much thicker gauge wire. Really overkill. go into the input there so we don't have to worry about the drop across there. The power showin' here should be the input power and absolutely because we're doing the sense in the output.

Power showin here should be absolutely spot-on So yeah, no worries, it was a bit how you doing before. Interestingly, the quiescent current here with no load changes when you turn the output off and on even though there's no load. so it's not, you know. Point Five Watts drops down to point Three nine with it with the output off.

so we should actually include the on figure in the efficiency calculations. I Think All right. So what do I think we're going to see on the results for this thing is that it's going to be optimized because it's like design. It's got bugger-all heat sinking on it and it's designed for like 70 watts output cable power capability or whatever it is, it needs to be like really high efficiency, even like ninety percent is not going to cut it.

so I'm They're probably going to be shooting for like ninety five percent, or you know, something of the mid-90s But you can't design the problem with universal supplies like this. You can't design them to be universally efficient over the entire input voltage range and output voltage range and output current range as well. It's just not possible so you've got to pick probably the worst case position like this speaker where you could design your thermals around the worst case condition of like the maximum output voltage driving the maximum output current ie. that's 70 Watts the power delivered.

Worst case, they can designed around that. so I expect to where it's its maximum power delivery capability is where it will be most efficient at 90 to 95% efficient or something like that and then at lower output voltages for the same a given input I'd expected to be lower efficiency. So say you've got you know, 30 volts input and then you like for 20 volts out you'll get an efficiency curve and then if you do 5 volts out it'll be lower efficiency again. So that's the sort of results we expect.

That's just you know how things are with DC to DC converters like this, they're always a compromise and they always basically as buck converters. Like this basically have the same characteristic trade-offs right? So I finished a sweep at 5 volts. so I got basically the input power and the output power in from 9.1 amps up to 3 amps basically and I just know that right? I was just going to do the sweep again for a 12 volt output here and you'll notice that the quiescent current. Well, the quiescent current remains the same.

if you switch the output off like that, it's not in the exact same note point three, nine, zero What? So that's that's the electronics in there. As I said, that's not going to change, but you switch it on with absolutely no load at all and it's no point Six where as we'll get in 0.5 before. So that's all part of the efficiency of this thing. And after laborious ly measuring and plotting or entering and plotting the data, this is what we get.

Tada. We have our efficiency curve here for the DPS 3w 3 power supply for a 15 volt input and we have a separate data for 30 volt input and 40 volt input as well. I didn't measure like a low input voltage. just like what's the point no one's gonna have like a you know, a 7 volt input volt.

enough 0 to 5 volt power supply. But anyway, there's only so much you can measure. So this is our efficiency from 30 to 50% I Just sort of expanded the scale when you plot things from 0, it kind of like just crams all the data up in the top and you don't see as much detail and this is quite interesting. Look at this.

This is the let's go in this one here. The red one is the 12 volt output. so 15 volts in 12 volts out and look at that efficiency. That was pretty much where I thought it must be.

It's actually up like 96 percent where I thought it kind of like had to be 4 right up to 3. M So I got 0 to 3 amps here by the way I didn't I measured naught point 1 amps down lowering because I knew there'd be that big taper down at the lower end and not much happening up the lower at the higher end. So I did not point to and I did not point seven-five here which is an oddball one. And if you have a look in here, the chart type I think I've done a video on this somewhere is actually a scatter chart.

It must be to then get the linear axes down the bottom here. It's kind of weird I think Excel does the same thing I'm using LibreOffice here. so yeah, otherwise they're just linearly spaces the samples so you have to use a scatter plot anyway. doesn't matter.

Small little light tech tip there. So anyway, that's a very, very efficient and as always down at low current it tapers off, but it's still a respectable you know, eighty two percent and four five volts out with the same fifteen volts in, it's still a very respectable ninety percent efficiency Peaks it there. it's a little bit under, but you know that's hunky-dory But look at this. for 1.5 volts out for fifteen volts in, it's not great-looking Peaks it around sixty eight percent here and like a right up at the higher currents.

it's pretty miserable, even though it's delivering a lower amount of power right? it's still the three amps, but it's only at one point five volts right? So four and a half watts as opposed to up here, which is, you know, 15 watts for the five volt output 1 for example. That's because like I said before, the the converters, you have to optimize them for a certain input voltage and a like walk input power and output power range and you just have to live with whatever efficiency you get outside of that. And that's usually not a problem in product design because if you're designing a switch mode converter for your product for example, then you know how much power your products going to take. as long as it doesn't drastically change power modes from one to the other.

Of course, you know how much it's going to take. So you optimize your DC to DC converter design efficiency for that particular current, you size your inductors, your size you, you know, your transistors and whatnot. and for that particular configuration. but on a universal power supply like this, there's just really not much avoiding this until you.

unless you go to great complexity and great expense. you know, spared no expense. You're gonna just have to put up with a poor result like this. Now here's the interesting bit.

What we really care about is how much power does this thing dissipate So what? I've had to do right? I've deliberately kept this one a bit clean, but here's what. I prepared earlier. Here's the exact same graph. again: A But I've included a second y-axis on the side.

Here, this secondary y-axis and this is the power. the power dissipation in the module. So that is our from zero to four watts. and that's what these dashed lines here, the colors represents the exact same.

So red is twelve volts out, red is 12 volts aren't here. and this is you'll find these in some data sheets as well, though. Do power dissipation loss. So this is how much it's dissipated.

And here's the interesting bit before they say the twelve volt output right, which is delivering the most amount of power. Three amps times twelve volts. So 36 watts output power right? Which is quite a lot for such a little module. right? It's only dissipating at worst to point to what's there.

And of course, the heat sink in there. It's sized up. It might be. Maybe it's you know, five degrees C per watt or something like that, so it might raise.

You know, a ten degrees, which is kind of what you feel if you stick your fingers on the back of those things. So guesstimate. You know, Ballpark it. It's probably a five degrees C heat sink.

About five degrees C per watt heat sink, right? And it's not. It's not too dissimilar for the five volt output. The blue line there, which is what you'd expect because they're a similar sort of efficiency up here. They're both plus ninety percent.

But look at the loss in the 5 volt one. Now, the 5 volt one is only delivering 1.5 volts at 3 amps. so it's only four and a half watts as opposed to 36 watts. So it's delivering much less output power.

But look at the power dissipation. This green line going up here. it is now. didn't my? Even though it's delivering right less than an order of magnitude, it's that little module has to dissipate.

3.8 What's So there you go. They're the trade-offs when your efficiency drops like that. because it's not an optimized to what can't be an optimized design. Almost by you know the nature of DC to DC converters trying to deliver a full range of alpha power.

You just can't get it. So it actually dissipates more power and has and heats up more delivering a lesser load. And it's exactly what you'd expect. There's nothing surprising here.

This is basic DC DC converter stuff. and by the way, that power loss includes to the quiescent loss of not 0.39 watts. So even though it's not necessarily being dissipated in the the output transistor, the output die at the output inductor, and that sort of stuff, the power devices that deliver your output load, it's still part of the module so the modules going to heat up. So I think it's you know it's a good thing to in actually include that in your total module power loss because that you could label that module power loss for example.

So there you go. Interesting, huh? Now we've done exactly the same thing again for a 30 volt input this time instead of a 15 volt input. So double it's not quite up to its maximum. Its maximum is 40 and as you can see here, efficiency.

once again, we've got different output voltages. We've got 24 volts. this time we've got 12 that's in the green, 12 volts in the red and the blue is 5 volts. And once again, I didn't do 1.5 here.

but if you did it would be you know, horrible. Like down in the 60% Probably just like last time. So very similar. but Wigan is similar.

Result like that. Once again, that 95 odd percent right up there. Does it hit 96? You know, like it's pretty darn good. Exactly where you'd expect it to be for this type of, you know, small heatsink module that we've got here.

It's fairly respectable at 5 volts out, but unfortunately the power losses are bigger this time with the 30 volt input. That's just the nature of the beast, and they all track fairly well. even though the 5 volt one output here is are quite lower efficiency at under 80% here. right up at the 3 amps, they all track very well, so we're getting about four and a half watts maximum dissipation at the full 3 amp output current.

But once again, that and the heatsink on it is like it doesn't get hot. It doesn't get too hot to touch, it doesn't. It does an adequate job of doing this. And why not? Let's do 40 volts input.

only two output voltages the maximum, which is our 32 volts us the red one there. Once again, we're talking. you know that same 95% Obviously, they've optimized it for that output power level, and once again, it drops. You know, it tapers down.

It's pretty horrible at the low currents here for the five volts out, but looking at the power dissipation here, there's obviously a couple of little oddball values in there. They're probably just typos. this should be, you know, fairly flat. There shouldn't be little spikes and things like that.

The efficiency is pretty horrible right down here at a low currents, but you know that's the problem with designing these wide range converters. As I said, but once again, you're at low output current, so the efficiency doesn't matter so you're not dissipating much power. Once again, we're under 4 watts there for those. sort of for those two particular five volts and 32 fold output from a 40 volt input.

So there you have it. That's some characterization of the DPS three double O three power module, and it's a really neat little unit. Um, noisy is all buggery, but you know, add some filter in or something like that if you're concerned about that sort of thing. but it basically does importat our claims fold a full voltage and current with 96 watts output capability.

and it does it with basically than that tiny little passive heat scene because it really is. You know, quite efficient, and if you want to really want to push it to its extreme limits we haven't characterized every extreme limit here, then you might have to have a bit of forced airflow, but that's a witty little module for like, what? under 20, 25 bucks delivered or something like that. Absolutely crazy. Um, yeah, maybe I do like a reverse engineering of that and see what's doing in there.

We know it's the what is it, the Xl7 double O 5 converter chip or something like that. but it's going to be using the external MOSFET and you know, external mosfet because it's not built in to do that sort of jazz. But anyway, that is a really neat little module. neat designed, and it basically isn't lying in the spec.

So I rather like it. So there you have it. I Hope you enjoyed that. If you did, please give it a big thumbs up.

and at the end of this video, I'll link in yes, somewhere up there somewhere. Yep, all over the place. Some other power supply videos measuring I've done one specifically I'm measuring the DC to DC converter efficiency and all sorts of stuff LinkedIn watch him if you haven't seen him. I've got tons of power supply type videos anyway.

I've enjoyed it. Catch you next time.