Hi In a recent video, I casually mentioned that one of these switchable Time 1 * 10 oscilloscope probes has a drastically reduced bandwidth in X 1 mode compared to times 10. We're talking like order of magnitude or more lower bandwidth. So like a 100 MHz bandwidth uh probe or a 200 MHz bandwidth probe does not have that same bandwidth on times One, it is much much lower. and a lot of people were surprised by that.

and I actually didn't find that. uh, rather surprising because it's not a well-known thing that's and it's not something that's not mentioned in a lot of places at all. People just go blindly around thinking that my 100 MHz probe is 100 MHz Regardless, but it's not so few people ask me uh, could I explain why that time 1 uh bandwidth is lower than time 10 As it turns out, it's rather interesting. so let's discuss that.

I Don't think this has actually been uh discussed anywhere, not as a topic on its own. Anyway, there's lots of tutorials out there on how uh, time 10 uh probes work and compensation and all that sort of stuff. But um, why is the bandwidth of a times One probe lower I Don't think that info is readly available out there. So let's investigate now.

I'll just say that this, uh, certainly is not going to be a tutorial on how to use Uh probes and why and all that sort of stuff that needs to be an entirely separate video and there's other ones out there. So we're just going to focus on why the X 1 is slower than the X 10 or has lower bandwidth. So as an example, we're going to look at this: some: RP 3300 passive oscilloscope probe Fairly typical, it's a high bandwidth one 350 MHz comes with the Ry gold DS 2000 series Scopes and this is what I showed last time in the video that callus people to ask about this. Now let's have it's a very uh, very typical uh Time 1 * 10 uh probe.

There's no compensation up the top here. the compensation is done down the bottom, but it can be done in uh, either place, but on the higher bandwidth ones typically done down the bottom. Now let's take a look at the specs. shall we look at this bandwidth Time 1 DC to 8 megaherz 8 MHz Are you kidding me? It's useless God and * 10 Look DC to 350 MHz So they sell these as 350 MHz rated Pros but you switch it to X 1, You got this crap.

8 MHz probe You bought this wonderful 200 MHz bandwidth scope and you switch it to X 1 and it's just garbage. Why? Well, we're going to find out. and because of that, of course the rise time. Big difference 900 P seconds 900 Puff Look at that for uh, uh, time 10 and uh, 40 NS for Time 1 Huge difference, but that's not surprising considering the um, the relative bandwidth difference there.

and of course the input uh resistance is going to be different between the two. One Meg everyone knows one Meg stand and input impedance cuz it just goes through to your. It's basically just a bit of coax when you put it on times one there basically just a bit of coax. Well, we'll see goes all the way through to your scope.

but in times 10 mode of course it's 10 Meg Ohms input impedance because of the extra 9 Meg resistor in here which a lot of people are familiar with and of course the input capacitance that will come into play as well. Time 10 it's only 16 paa farads very small equivalent to the tip capacitance which we'll take a look at. Time 1 is 100 paa farads. Huge amount of capacitance there so that will come into it.

Now the first thing you should say is well show us all right. We've got our signal generator here. We've got our scope I've got it set to 1V uh Peak to Peak sine wave at 1 mahz so we'll be able to adjust the frequency and there it is 1V Peak to Peak You probably can't see that down there unless you're watching in HD but that is 1vt Peak to Peak Not a problem right at 1 Mahz, it's you know the probe really isn't uh, introducing any attenuation at all bandwidth limit doesn't come into it because as we saw oh, by the way, I'm on times one here with the scope probe I can switch it to x 10 of course and there we go. we get 1/10th the Ude.

so we're on times 1. So let's actually adjust the frequency here up and see if we do get that rated 8 MHz bandwidth. So cuz we're 1 VT Peak to Peak um is specified typically at minus 3db bandwidth which should be uh 0.707 So point if it's 1 VT Peak to Peak should be 707 volts. let's say 7 Vols Peak to Peak.

So here we go. we'll wind the frequency up 2 MHz three. Aha, you can see the amplitude dropping. Look at that.

let's see what we get at 8. Mahz. there we go. we're getting 800 molt so it's actually bandwidth is better than what it uh says on the data sheet and 707 let's go down.

probably can't read there. There we go, that's practically 707 10.1 MHz bandwidth of this Probe on times one mode useless. and of course if we keep uh, increasing the frequency on that, I'll go up a digit. Here we go and we're up at 30 me 40 MHz 50 60.

See the this is a 350 MHz rated probe and it's ABS there's 150 MHz useless I mean you know, yeah we're still getting something down there, but you know I mean the attenuation is is absolutely horrible and we'll switch it to x 10. Here we go and we'll do the same thing. we got to turn our vertical up there and we're at one. We're at 1 MHz at the moment and let's keep going There We go.

It's dropping a little bit, but I think that's probably. oh no, it's going back up a bit bit. So that's um, probably the amplitude stability of this ryol uh function gen. But we're up at 160 MHz now and that well, 160 MHz is the maximum frequency of this uh Sig gen.

But there you go. It basically did not drop in amplitude. You know there's a little bit of change there, but the amplitude stability of this I Don't know the spec, but and basically that probe is just fine. It is not attenuating at all because it has a Rer bandwidth of 350 MHz on time 10 mode.

And of course that bandwidth limit is going to kill any high frequency content like this Square wave. For example, here's a 10 MHz Square wave on X 10 mode. so a bandwidth of her Pros 350 MHz and of course it looks like a square wave and we see the overshoot and the actual signal Fidelity in there. But if we switch it to x one that look at that bandwidth limit of that times one probe, it's useless.

Now you're probably asking yourself at this point, something smells a bit funny here because isn't like scope probes supposed to be like good quality coax and uh, and when you put it on titin One mode, it's just a bit of coax going in, right? Oh yeah, it should be. So let's get a bit of coax. There we go and let's whack it in there and see what we get. Here we go look at that a bit of overshoot, there bit of ringing because it's not uh, terminated properly.

But let's increase this in frequency. We're at one, we're at 1 MHz there. So let's wind it up. and as you can see, see: 40 MHz 50 MHz No problem.

look at. Well, it's routing off there. but let's not look. but there is no attenuation of that signal by just a bit of coax.

So why is your humble oscilloscope probe got a lousy 10 MHz barely 10 MHz bandwidth on times 1 mode when it's supposed to be just switched directly through Hm? Let's go to the Whiteboard and Tada Here we have the circuit of a typical time 1 * 10 switchable probe. You may have seen these before. Yes, this is a new whiteboard mounted on the wall. Isn't it neat? And yeah, you may have seen this circuit all over the place.

Very typical. There's lots of Articles and tutorials on how x 10 compensated probes work. We won't necessarily go into that here. What we want to find out is why that pesky x 1 probe has that really low bandwidth.

So let's start out at the tip. over here. here's the probe tip. There's a little bit of probe tip capacitance.

They don't often show this in the circuit, they just show the typical Nmeg resistor and the compensation cap. but we'll show it because it's actually rather valid and it'll come from the data sheet as well. Uh, these are all going to be typical values. Of course they will kind of vary, but uh, let's just go for typical.

Now on a X 10 probe, you have a series emeg resistor in here. So it voltage divider with the 1 Meg input on the oscilloscope over here it forms at DC forms that uh uh, time 10 divider. They call it time 10. Yeah, I Know, it's really weird.

it's divide by 10 Now of course there has to be a compensation cap across that 9M resistor to, um, take into account the AC bandwidth of the probe. So you got to compensate that for your input because the input of the oscilloscope here also has about 15 paa farads of input capacitance. Very typical across the three. Scopes I've got here.

Um, even from the 500 MHz Um, Agilant, right down to the Uh Ds1052e, they're all around 50 around 15 paa farads plus minus a few peaka farads input capacitance. so that is designed to compensate for that. Ignoring the coax at the moment. Just assume it's an ideal Um Coax ideal transmission line and that's all there is to it.

Now there is is actually two different schemes and I've shown you before. there are these types which have the compensation not in the probe tip here. So this capacitor is actually fixed. It's not variable like I've shown here.

the variable capacitor is over here in a little RC network over here. Typically a going to be a series resistor in here with an adjustable cap 0 to 50 paarat Something like that. Exact values are going to change depending on how the probe is actually manufactured and typically your high bandwidth ones are. You won't get the adjustment in the top here.

you'll typically get it at the base down here, but you can also get the uh uh, lower bandwidth ones which have the adjustment in the top here. So they actually have the adjustable capacitor in there, the trimmer cap. And they don't have any network down here, so the coax just goes directly into the scope so you can ignore that those components don't exist. All you got is the 1, Meg and 15 pad input capacitance of your scope.

Now let's take a look at a times one prob and how that works here and you've noticed I've added some sneaky little calculations in here. Now how times one probe Works Incredibly simple. All it does is got a switch in there and shorts out the 9 Mega resistor and the compensation capacitor. That's it.

So these no longer exist. So you've got a direct connection from the tip directly through to your oscilloscope input. And of course, if you don't, uh, have that Network There, you've got nothing. So all you've got is your tip capacitance.

You've got the capacitance of your coax, which let's just say it's 100 paa farads for a meter. For argument sake, it's going to vary depending on the type. But let's just say it's 100 paa farads And once again, that will be you know, equivalent to another capacitor in there, like that distributed, eh, whatever capacitance of the coax and the capacitance of your Um scope input and your 1meg input resistor. So you would think that okay, you've got all this capacitance on here and it's going to have an impedance at frequency.

Now you should know your formula for capacitive reactants or the effective AC resistance of the capacitance at a frequency. It's one over 2 Pi F C. So let's take an example of 10 MHz which is roughly the bandwidth of the X 1 probe we measured. And let's plug in these values of 15 paa farads which we got here and here and that works out to about 1K.

So at 10 MHz you've effectively got no longer got a 1 Meg ohm Um, oscilloscope input Here you've got 1 Meg in parallel with 1K here in parallel with 1K Here and then the capacitance of the coax 100 picofarads. That's about 159 ohms at 10 MHz So very low. Your 1 Meg input impedance scope is now like, you know, 100 ohms or thereabouts, incredibly low input. But still.

Let's say you've got no Um loading effects on your circuit. you're driving it with a low impedance load. Then your signal should go Faithfully directly from the input through to the scope without any attenuation. So why are we getting uh, the attenuation on that? That um, drastic attenuation about 10 MHz bandwidth on the times 1 Pro Whereas a bit of coax, which is essentially exactly the same thing.

it's just. well. let's say it's longer. You know it's got 115 Then effectively, what's the difference between a times 1 probe and a bit of coax? There shouldn't be any difference, but there is.

We measured it. Why? H Think we can solve this one with a multimeter. Now, before we get to the multim meter, let's just have a look at the input specs again and see how they relate to what we've seen on the Whiteboard. Let's look at the input capacitance here in times One mode 100 paa farads.

Now, that's effectively the capacitance of your coax cable only because the Tin One switch, of course, shorts out the input compensation Network and all you're left is with the coax. But it can actually also include um, the compensation network if you've got it at the end of the cable like here, um at the scope end rather than Um this one here, which it just has the Cox going directly into the Uh scope. and that is like in that case, they'll have the compensation Network up here. Now there's actually not much um difference between having the compensation Network in here and and down at the end of the probe down here you can.

The performance can be uh, the same effectively. uh depends on how you tweak it and uh, actually design it. but um, most of your high performance X 10 uh probes will have uh, your compensation Network down this end here. Now if we have a look at the uh times 10 mode here, it's 16 pigar and order of magnitude less effectively.

The reason for that? that is the effective tip capacitance. and I say the word effective because it not only includes the uh small amount of capacitance between your tip here and your input compensation Network In here you starting your 9 Meg resistor down in there, but it's also the effective reflected capacitance from your coax as well. So it's that 100 paars. but because we've got a Time 10 divider in there, that capacitance gets divided down and then becomes effectively 10 times less at your tip.

So that's why that one's 10 times less. This one's a little bit high at 16 paa farads usually uh, a little bit less than that us, but usually you'll see an order of magnitude difference between those and also, uh, this is why uh, your fixed x 10 probes um have it can have a better performance than these Compromise Time 1 * 10 Pros because your compensation Network can be right at the tip, that Nmeg resistor can be right up there. and there's bugger all tip capacit. whereas this one's got just a little bit extra and a bit of extra transmission line in there as well to, uh, deal with.

So yeah, your fixed Times 10 probes are always going to be um, uh, higher performance or can potentially be higher performance than these switchable compromise blah probes. And that's why a X 10 probe is highly valued because um, it has a much lower input capacitance than your times one. so it loads your circuit down much less at high frequencies. That capacitive reactor acve reactance formula we looked at.

Remember that now what we're going to do. We've got our regular coax here. It's just like half a meter of Cox and we're going to measure the center conductor and wellow. No surprises what we're going to get here, folks.

zero Ohms, there you go. It's a direct short because the coax just goes directly through H Yeah, it works as a transmission line and everything else, which we won't go into today, but what happens if we measure this score? Probe on times one, by the way. So let's well, let's put it on X 10 first. Okay, and we should get our n Meg that we looked at on the board.

There There it is. spot on N Meg No worries, but we switch it to times one and there should be nothing else in there but the switch. right? That's it. You'd expect to get exactly the same as the coax.

Do we? There we go. 330 odd ohms between the tip and down here. What is doing that? Is there a series resistor in there? Well, to show you that, there's nothing, uh, no. funny business going on in this tip here.

I've taken apart a * 10? Uh well. switchable * 1 * 10 Pro Yes, it is, um, uh. different to the Ryal. It's not a Ryal brand, it's a Velan brand.

It's just a cheap ass old one I had lying around and there's the time 1 * 10 switch. So time 10 is down in that position down there. so we're up in one position and that should short out the Uh Network there. So the compensation.

Network So let's measure between that probe point and here. what do we get? A dead short So there's no funny business going on inside that probe that Well what we'll call the probe. it must be in the coax or the end of the coax. And here's the part of that chopped off Vellin probe and let's measure it.

There we go and the fine wire. I'll show. I'll get a very good closeup of this in a minute. but basically I'll have to hold my finger on there I'm not touching the other one.

There you go. 230 odd ohms. this oscilloscope coax and trust me folks, there is nothing. There's no series resist or other uh circuitry inside there.

I Could chop it open to prove it, but uh I don't know. Couldn't be bothered. Oscilloscope probes do not use regular coax like a regular coax cable. These are what's called a lossy transmission line.

They're designed to have that high resistance and check it out, that's what's inside an oscilloscope coax cable. It's a single strand there. No. I haven't actually.

um, you know when I stripped it? No I didn't Uh U accidentally cut off all the other strands. It is a single tiny strand like that and often it is. You might be able to see that it's actually not entirely straight. It's actually got little kinks and bends in there and that's actually designed in in some of the, uh, more higher quality ones.

Remember, this is only a 60 MHz vellin. Um Heap crap And that's what these probes are optimized for. They're optimized for their times 10 mode the high bandwidth mode. which is why they going to use this lossy Uh transmission line which has typically um, most probes are going to have a couple hundred ohms.

It will vary by plusus a couple hundred. Um, you know the actual DC resistance isn't in actually, you know it's it's important, but it's not the main um, you know driving. Factor transmission line design Theory I won't go into it, it, It is actually quite complex I may have to do a separate video on this, actually simulating um, all this stuff and exactly how it works. but that's these probes are optimized for or the transmission line coax.

The lossy transmission line is optimized for the high bandwidth X 10 mode. So the X One mode is actually just a bit of a cludge added on. And that's why we're crippled with the the Uh low bandwidth on the Times 1 mode because they're trying to incorporate both Uh features into a Times 10 probe. So stripped a bit more of that out and you can see somewhere along here the crinkled nature of that wire.

There it is. see, it's sort of crinkled and uh, some of your more high performance ones are much more uh, crinkled than that. and uh, presumably they do that to help with the flexibility uh of the cable and of course, uh, but you'd have to, uh, take all that into account about the final resistance cuz this is actually not just regular wire. Of course, it is resistance wire which has a certain specific High Resistance uh designed to uh match the compensation uh networks and that's how the whole probe is able to get its high performance.

I mean this is only a 60 MHz one, so this isn't a particularly high performance one at all. but you can see the white foam uh insulation in the coil there and that had have a very low dialectric loss very specifically Uh chosen for the task. and there's a black outer protective jacket there and then the Uh braid shield and then the outer jacket on top of that. So there's a lot of Art and Science which goes into the design of these probes and then the Uh matching or the coax used in these, and then the matching uh compensation networks to give you a flat uh usable response.

Up to you know, in some cases, uh, well, over 500 MHz Which is a really amazing technology. So these aren't just regular Co Taes. So after that little investigation, what are we left with? we're left with the knowledge that this bit of coax ain't a bit of coax, it's a lossy transmission line. It has DC resistance in there, and that's going to which we measured at around about 330 ohms.

so that now, aha, we have an attenuator. Whenever you got a series resistance and a resistance at the other end, you've got an attenuator. So let's look at the entire circuit we've got now for our times. One probe okay, these don't exist anymore.

Our resistor and cap don't exist so that just goes straight through wide straight through like that. and this 1K really doesn't come into the equation at all. so Wa may as well not even exist. So we'll leave that out and we've got 330 Ohms now in series with let's take out our compensation Network Here let's assume we have a probe that doesn't have the compensation Network at this end and then our 15 Paa Farad scope input has now become a 1K resistor as well.

So we effectively have 330 Ohms in series with 1K Ignore the 1 Meg it's couple ERS magnitude bigger than the onek, so it doesn't matter. A rough rule of thumb is something's an order magnitude bigger 10 times bigger. You don't worry about it, you just take it out of the equation for back of the envelope type calculations. So if you do the math with that 330 ohm and the 1K over here, which is the impedance at the rollof minus 3db rolloff frequency of 10 MHz we measured it's around about that minus 3db point or that 707 I Think it's 75 or something like that near enough.

So that's what determines your upper frequency limit this series resistance. Um, and your capacitive reactance of the input is scope. But of course it gets a little bit more complicated than that because that 330 ohms. Yes, it's at DC and yes, it's going to be quite similar at Um AC as well.

But uh, this is because it's a distributed resistance. It's actually all the way through here. So I've shown this is all different resistors in there and this is how you can actually simulate the thing using the lumped model like this. And then you've got the individual capacitor is in here and there will also be inductance as well in any Uh model of a trans of a lossy transmission line like this.

but I haven't shown any of the inductors so of course that's 330 ohms at uh, you know, let's just say it's going to be constant at AC and DC the input is going to be different at DC at DC we're going to have our one Meg So that's why down at the low frequencies with the X 10 probe this 33. the resistance of this um, coax doesn't matter at all cuz you know one. Meg Do your voltage divider math there and this doesn't matter a rat's ass. So the higher frequency you go, the greater attenuation you got and and you know you can Muck around with all little RS and C's and stuff in the lumped model.

But basically you're going to have that upper frequency. so you're going to get that frequency roll off that looks like that at a certain frequency and it's going to be 3db down at about 10 megahertz there. If that's your 1vt input, it's going to be 0.707 As we measured on the scope, it's going to have a rollof because you've got effectively a capacity here which changes with frequency and let's say a fixed series resistor element. So why do they even bother having these switchable Time 1 * 10 probes.

If the time 1 mode bandwidth is useless, Well, it's just convenience. Um, essentially, they design these things for the high bandwidth on the X 10 mode. Uh, but there. But that by nature of its very design with the lossy transmission line.

um, it just cripples uh, the Times One mode and makes it pretty much useless. Well, it's not useless, it's just low bandwidth. So the rule is, if you want high bandwidth and you don't want to attenuate your Uh signal, which the times 10 Uh probe does, that's one of its disadvantages. One of the advantages of the Times One Mode: you're not attenuating your signal by uh, 10 times, then use a regular coax.

And of course, you can get very high bandwidth probing with just a regular coax. If you you know there's various Uh termination resistors on the input, and uh, stuff like that, these can be very high bandwidth probes, incredibly low impedance of course because you've got the direct Uh capacitance of the coax there, and high frequencies. But if you want um, high speed with um, no Uh attenuation, then just a regular coax can be uh, useful. And I might do a video on that too.

And there's other videos out there that, uh, explain that. So why do they go to all the effort to have a lossy transmission line in here? Uh, when you know you could just use an ordinary coax? Well it, this is the way. This is the technique that they use to get ensure the widest bandwidth possible in a Time 10 probe and also the flattest bandwidth possible over that mode. so it doesn't at the end, at the end of the bandwidth, doesn't ring up like that or doesn't go down like that or anything like that does no funny business.

It's a lot of Art and Science goes into this and it was originally developed by Tectronics way way back. and uh, they were the ones who pioneered the lossy transmission line technique. and they're very cleverly and very carefully designed to get like the you know your five six 700 megahertz bandwidth passive probes which uh, you know you take for granted these days, but they're You know, you're paying a lot for that. And there's a huge difference between a cheap ass 60 MHz probe and a 600 MHz passive probe.

and that's often why they cost so much. They're very difficult to actually get right and trim all and get it all right. In terms of the compensation Network in here that we looked at matched to the Um actual physical properties of the coax. So a lot of Art and Art and Science goes into just designing the physical properties of that cable, getting that, getting that little wiggle inside exactly right, getting the lossy uh parameters all right so that they match and you don't get any funny business overshoot undershoot at the end of your bandwidth and all that sort of stuff that requires a whole separate video though.

Well, that was a lot of talk. Just to get to the simple conclusion that a oscilloscope probe coax actually has a lossy transmission line resistance in it I Could have just told you that at the start and saved you what is it? 15 20 minutes Worth or something. But ah, I hope you found that interesting. and that is why these X One probes have a lower bandwidth than a Time 10 probe order of magnitude or greater.

It's something a lot of people don't know. so I hope that was useful. and if you want to discuss it, jumping over to the Eev blog forum and if you like the video, please give it a big thumbs up. Catch you next time.