Hi Lets take a look at the High Voltage Differential probe In This case, the LeCroy AP031 Now, I Don't believe this is actually designed by LeCroy. They Just re-badge it from someone else; actually I don't remember who the actual manufacture is of this thing. Tool for safe operation of your oscilloscope. Measuring high volt not only high-voltage stuff, but measuring hi common-mode voltages.

Things like mains, power supplies and things like that, which you can't do really safely with your regular oscilloscope probe. and I've done a whole video on that, so click here if you haven't seen it. I How not to blow up your ass? Illus Cope. So it's a vital thing to understand that.

But one of the ways, of course, how not to blow up your oscilloscope when measuring stuff is to use one of these high voltage differential probes. It allows you to measure you so good. Just hook this onto basically any part of almost any circuit within reason. Being a high-voltage mains, power supply or whatever other high voltage systems three-phase stuff doesn't matter what it is, you can hook these two probes up to any part of your circuit.

By the way, I've just done that. I've lost my black one black a probe here. it's around here somewhere. Anyway, you can hook these up to any part of your circuit and you will be completely safe and then you just plug the other end the output.

So this is the import and just plug the output into your source code. so it converts the differential signal on here hence the name the differential probe and converted to a single-ended signal which goes into your oscilloscope. Now these generally don't have a particularly high our bandwidth I Think this one's about 20 or 30 mega hertz or at something like that. So it's you know, not designed for really high speed stuff.

but that's not the point. The point is that it has not only up I ten-to-one range IE like a X 10 your regular times team private divide your input voltage by 10 that you're familiar with. Interscope Also got X 100 as well or divide by a hundred so they correspond to a range of plus minus seventy volts on the 110th range and the one one-hundredth range has plus minus 7 hundred volts. And it's got a common mode voltage of plus minus seven hundred volts.

So a brief recap on what common mode voltage actually means here. Now this differential probe. It's just a differential amplifier just like Site and Object is a differential amplifier. but it and off of course has a ridiculously high gain.

This one has a relatively a small gain of either X 10 or at times 100 to compensate for the input attenuation here. So we can put up to a thousand volts Rms across here. and well that's no problem. But what we we're really concerned about is this common mode voltage and this is where you can come again sir.

And I've done that video. How to Not to blow up your Oscilloscope What that? Common Mode voltage means is it then reference to the output e. the ground of the output than common of the output here is connected to your ass or scope and that's then connected through the minds. If I did that as a circuit real, maybe that should be a mains earth ground.

but then it's the difference in voltage between your mind earth here and either of your input voltages. I've drawn that as a voltage source there between any of those imports and in the case of a high voltage differential product like this, that can be up to plus minus seven hundred volts between the output here I've drawn extra terminal. They're probably shouldn't have drawn that and the output and either of the input terminals. and that's important when you're measuring, say a mains power supply or something like that and it is very high relative to the earth on your oscilloscope.

Now there's a common misconception about this high voltage differential probes. Some people think that they are isolated probes ie. they have like a transformer in there to isolate the input from the output like this. and nothing could be further from the truth.

and we can prove that the only thing we need is a multimeter. Let's measure between say, the positive input and the output ground here and I'm not touching it so it's not my fingers in the way. Look for Meg Ok and that's a very consistent for making I mean we can swap that over South There's any like active circuitry stuff going on. you're expected to change, but it doesn't look.

It's exactly the same. So we've got for Meg between either of these inputs here and the output ground. and we can repeat the exact same thing for the negative input terminal and the output ground. You guessed it, it's that for Meg Again, it's very consistent.

It is not isolated, and we can demonstrate that again using our high-voltage make a year. So we cannot select 500 volts because we've got a seven hundred volts maximum here. so we don't want to do a thousand because, well, we could blow the arse out of our high voltage differential probe. They're not completely infallible.

they have ratings for a reason. so we're connecting between the output grand which goes to our scope and the input. so we'll be able to measure that same resistance we got before, but instead of this moldy made are only working at you know, a couple of volts. This thing will actually test it at 500 volts.

So here we go. let's test that and Bingo! we get the same for Make it is genuinely a resistance. This thing is not isolated. So how does this thing work if it's not isolated? Well, those resistance ratings might have given you a bit of a clue, but there's only one way to find out.

You know we say you're on the EV blog. Don't turn it on, take it apart. And needless to say, this puppy's gonna have a favorite of shielding inside there, so we might have to crack that open. In fact, we might have two yd solder some cans to get to the front end practically all over these shielded.

We've got two separate sections here, so our outputs from here. so I'm guessing that this is the input and this is the output can. and maybe this is just some my power supply stuff up here just keeping that separate. but don't look.

We got to at least a couple of tremors on. There's another couple of holes, so maybe they got some trauma capstone? The bottom of these are obviously yeah. 10 10 pots here for a rat, frequency compensation and other sorts of training and I got this I can off here. It was just very lightly.

You are soldered onto the sides here, but it looks like this one has a couple of sort of points down here. These are actually these tabs are actually from the kids. Thankfully, they're not all soldered, they just attacked it down. so hopefully I can enter probably another couple of points over here.

But anyway, we can start to see the structure of this thing. No, they're high voltage isolation slots here and here here and here like this. So here's our input is obviously two separate sighs. This is be like a symmetrical because it's a differential type pro, but I wonder what we're gonna find under.

He might be able to start guessing and we're in like Flynn We've got ourselves a high-voltage isolation shield here just so that nothing acts or our shorts over to the middle can of course. And that's important because we've got a lot of through-hole components here. Vertical Check it out. It's all a bit.

It's all about how you doing, but look. Numbers rubbed off the chips. mongrels. They don't want us to do what these are.

they're probably, you know there. they just our pants right there. just so that you know, not Jelly Bean, our parents. But they're not going to be anything hugely special now and take a look at with what we've got here.

Struggled across the isolation slot going right down here. These are all of our high voltage input components, so you can see we're going to solve some ceramic caps here. These will be for our frequency compensation as we'll see, and they've actually hate shrunk those maybe to provide I It's not, maybe some insulation, but they're just around them like that. so maybe some explosion protection or something like that.

What they do in most forever, you know, input moves are motivators and stuff like that, and that might be my guest. But like you know, there's some resistors in here that have been over so they're obvious. Maybe trying to isolate the leads on there, but then that's all exposed and well. anyway.

we've got some resistance here. It looks messy, but let me try a buzzard out. Now You'll actually notice that it's virtually completely symmetrical and that's what you'd expect from a differential front end like this. So we got and I Thought we had a component missing in there.

No, but let me buzz all this out. and I'll get your Dave Kehr Jordan and this is designed and all laid out by me. Mr. Wu Good on you.

Mr. Wu And its sapphire? Yes, I Believe that's the original equipment manufacturer. They're the original designers of this thing and it's rebadged under many different names. So if I take off that hates drinking, start to see some of this structure here.

Here's our import: yr Negative: A positive side of this circuit is going to be absolutely identical, of course, and that's in series with another resistor just end on in. I Love how they've even got the silkscreen symbol down in there. 42 in the in resistance like that. That's kinda neat.

So there's two, have one mega resistors in series and then there goes down to another one mega resistor which then goes over the high-voltage isolation slot there and then go into another hundred make resistor down into their one make to make three make for Meg Does that ring a bell? And they've just got some 6.8 half caps there? they all 6.8 Yep, but there are 1 kilovolt job so they're high voltage are ceramic caps of course and these don't connect anywhere on the underside. So if we go on the underside there, you can see that the capacitors don't actually connect. So it's not like you've got one capacitor across each resistor there, so it looks like we got to use it. Well actually one, two, three series caps across the entire string of those 41 make resistors.

And of course they could have used one big high voltage capacitor across there. But I it's not easy to get greater than one killer of all capacitor. so they use just use 31 that kilovolt ones in series and which I gives you your voltage. Likewise with the resistor, they could have used just one big ass for make a resistor on the import there, but hate you wouldn't have got the high-voltage isolation when you string components in series like this.

The voltage ratings of each resistor in each capacitor add up, so that is effectively 6.8 / 3 gives you capacitance and then the voltage rating is three kilovolts total and I Do like how they added this little metal shield here over this for and where the cables coming that just sails up the gap in the can. therefore the cable entry. Someone was thinking mr. was thinking good on you Mr.

Wu And you don't want to craft a canary switching converter? that's why there's two inductors there and they're L1 and L2. After putting a bit of speed on that chip, it turns out someone at the Tonys factories not very good at scrubbing off those numbers. So in true EV blog style would just speculate on what's going on here. Let's give it a go any way.

I want I Don't plan on really doing a full reverse engineering of this thing cuz it's a bit of a pain in the butt and feeling a bit lazy today anyway. so I just want to show you that it wasn't an isolated probar basically. and it was using input dropping resistor high-value drop resistors to limit the current and provide that attenuation ratio on the front end. So anyway, we're going to positive and negative input here.

We're going out for Meg over resistance here. for make worth of resistance here. we've got some compensation caps around here and around here. Then the two resistors in there I'm gonna be the lower end of the divider so we gotta for Megan put divider and then it's going to go over to these two resistors here where they've got three caps.

They actually look to be there all in parallel. so I don't know why they've got so many in parallel. Anyway, we'll have a look at those are values maybe do compensate for each frequency range which is common technique for like bypassing for example, because I all have a different impedance and a characteristic based on that, their value and their package size and everything else. but they're all identical.

Ceramic capacitor sounds a bit unusual having three in parallel. Anyway, for Megan poor sister, whatever value that is down there, haven't looked at that so that will give a fixed division ratio about the positive and negative input. This tantrum: I hear, that's probably I just used for an offset adjustment thing and then we've got us some trim at caps in there. Just a trim these two values in there and you probably saw the holes in the metal case wherever that is.

Think it's over on the other bench. But yeah, you can get to these after the fact after you sold in the middle can to work trim those and then it looks like you've got the output of this going over to hear another output going over here that's going into this middle can package here. haven't actually looked at that one yet. but based on the number of pins there and a common technique used in Scope Friends which I'm sure I've seen before when I've done Skopians and things like that, that's probably going to be a man matched j fit .

very common to use AJ fit our differential unite, roll-your-own j fit input and so I reckon that's what's going on there. That's probably a quarter of purp and this I see up here. What would that be? Okay, here's our output is probably coming yet. looks like it's coming from there.

So yeah, yeah, it's coming year. There's the output coming over there. So yep, that's going to be our I'm output driver. so that's converting the differential output.

Sure. Anyway, I'm not sure the arrangement of the RPM there, that's I don't know. Anyway, I'm I reckon that's its differential all the way. probably through to there and then differential.

single-ended cable driver. so that's just like a high-spec You know how I slew rate, low noise are payable something like that to drive the airport and Bob's your uncle. We could another couple of tremors in here I don't know what they'd be doing. Maybe some offset stuff.

things like that. Another couple were trainees down in there too. Two of those back-to-back but not sure. It's maybe use some sort of I don't know.

Is this some sort of current mirror or something going on? There are not entirely sure anyway. and that's going to be the basic arrangement. and I tell you what? I think they've tried to actually scrub the number of that can? That's kind of what it looks like. Marks don't really show up or as ever, 3968 is it.

So let's go Dave can reverse-engineer In addition, please excuse the crew da of this model I didn't have time to build at the sky order painted. This is a very, very rough schematic of what's going on here. In fact, I'm not actually entirely sure what's going on here. This chip which I thought at first I Yeah, that's gonna be an RPM quad-carbon but I don't actually think it is.

Anyway, let's have a look at this thing. as we've seen, we've got our 41 make resistors in series and therefore 6 P-8 caption series for some frequency compensation on that. and they're all their thousand volt capsule and they're going into what the 25k here and the tremor which is just used for an offset balance adjustment there. So you know we've got a pretty good division ratio here.

This is why it can survive and measure all the high voltages. It's basically differential input and then I believe it's going into a jewel are matched. j fit here because this is very typical friends. Although I couldn't really find any info on that part number that I saw down here.

It said it was like an in general Jeff it but it didn't say it was jewel or anything else. So from some obscure company that's probably not even around anymore. Anyway, I believe that is a jewel J15 put that makes sense. We've got some dire clamping here for input protection.

By the way, these three caps here hundred and twenty past eighty, two parts and 82 path and then probably a much smaller art trimmer. cap value here. Just a trim. the frequency compensation on this thing I'm after at all assembled in the metal cans.

everything else is on. You get in there and hold your tongue at the right angle and tweak These tweak your balance adjustments in this couple of other parts in here which I haven't reversed engineered and showing. its going to unusual that they've got the three in parallel. Not entirely sure why they're actually doing that.

I'm there all the time. same type of that ceramic, so maybe they're just trying to get a tighter tolerance or something by putting the three in parallel like that. That's all I can think. also that the tremor has a better, more accurate, more controllable adjustment range.

perhaps? Anyway, going into 2j fits and then I'm after that. I'm not sure what that puppy there is doing, but it by just looking at the peanut and everything else, it doesn't seem to be an option. So what? I think it is. and because it just didn't make sense in this sort of arrangement here to have an Op-amp directly on the output here.

it's not that one. Forget there for a second. a differential J frontin like this usually need some extra transistors here and usually formed a forming a current source down here. so that's the usual arrangement.

So I think my guess is that that he's actually one of those that transistor array chips and I couldn't make heads or tails of the number. the number, the partial number that I got off. That certainly didn't make sense in terms of a quad our camp part number. So I believe that some form of our transistor away, right? They were quite common back in the day.

You can still get them in various forms to your eye and other companies make him and stuff like that and lots of obscure providers back in the day match transistor things that the reason that they're good is because they matched on the same die if you're trying to do a discrete front end like this with all separate stew, separate j fits and then separate bipolar. I usually bipolar transistors that one BR j fit so these will be B JT's in here so I hope like I'm missing a whole section of circuitry there. I mean this is not just two resistors and accounts that you know there's gonna be no like. Usually these have like a 40 like five or six transistors usually in IE current in a differential our configuration with the emitters tied together and stuff like that.

anyway. I reckon that's what that puppy is doing there so that makes sense. And then the output. The differential output is then going over to this puppy which had the number rubbed off so I don't know, I can't even get a partial on that one.

I'm so that it be converting the differential into what the single-ended driver to drive you coax over here. and that's about all. she wrote this couple of other transistors in there I haven't bothered to reverse-engineer all that. There's a couple of other trimmer adjustments in here which are probably award and this one might be output I gained.

for example in this one, he might be I current source adjustment so I reckon that's what's going on there. So if anyone actually has a complete schematic of this thing and I'd really like to see it. Anyway, this wasn't really a hundred percent reverse-engineering our video. it was just to show you what's inside.

One of these are differential probes and then they're not actually isolated. They The reason that they're safe is because they use these resistors in series like this and you're protected. So if you're holding onto the output of the ground and output like this and you've got hundreds of volts 500 volts floating around in here relative to the ground that you're holding onto, your pretty damn safe because you've got forming in either lead like this and you got multiple capacitors in series as well. You know, like you, if you just had the one capping, they could get a single fire then it might render it a bit unsafe.

It can go through the 25k to ground and your output. United This is your ground reference output here so that could you know. If you're holding onto that are doing whatever then you know it could be a problem. But these are very safe probes just by nature.

of the 41 make resistors in series and the high-voltage isolation slots and the raid caps and everything else. So Bob's your uncle how it works. It's just a basic Rj fit differential amplifier with a big attenuator on the front end and just for kicks things inside the Dc-to-dc converter and then rub the number off yet again. Why? Protected crummy, discreet Dc-to-dc converter.

Gimme a break and we just have a quick squiz what rails we got here. Of course nothing is marked on here. There's no voltage rails, no ground. smart, But of course.

Foreground: you almost always take the ground plane on while the plane on their the main play, almost certainly going to be ground. And then, as I think I mentioned before, we got the two inductors there and there. I think this is only going to be giving out. There we go at nine and a half faults and the other one should be- similar? Yep, negative.

Nine and a half. So there you have it, that's inside the Lacroix Apo 31 / Sem Fire 9001 different high voltage differential probes, so I hope you found that interesting. Just how these things work with the front end. I mean the exact reverse engineering details and a huge big deal, but the fact that uses those are two big resistive dividers with a differential fit front end.

That's what we wanted to know. Rather interesting little bass. This one is ancient arm I don't know the date on this one. then rubbed off all the bloody codes on there anyway.

this design must date back a long time. I got this one back when are still working it cell I think so. Five-and-a-half-year Zv blog full-time for years about him. You know it's probably a good 10 years ago.

I got this end. it was an old model are back then. it might be a more modern variant these days because it is pretty how you doing it's all through hole design I'd be surprised if they haven't done a new modernized version of it or something like that, but it's probably still identical, but you probably wouldn't. Manufactures still manufacture that one today I Would modernize.

It was some matte surface mount stuff, at least just a you know, just to make it neater and tidier. And I get your production costs down, but the design would still be as valid as it was back then. and well, you know, a sum like the fit front in might have gone obsolete or something like that. The transistor rain might have gone obsolete.

you might you know, modernize those or something like that. But the basic, our technique, and everything else would remain very much similar. and there's other ones on the market as well. You can get them for like 300 bucks on ebay these days.

Not this particular model, but there's another one. Actually, I see. I think I can get my hands on one of those. It'll be interesting to do a a teardown of the cheap three-hundred-dollar ebay one and see Anyway, hope you enjoyed it.

Catch you next time! Hi I just read a post on the EV blog for with somebody was asking about the oscilloscope probes and how they can be potentially dangerous if you hook this ground lead-up to the wrong . in your circuit you can bloody circuiting blog scope bang and it really is a big trap for young players. Have mentioned it before but they wanted to know exactly under what circumstances that could happen.