Hi just a quick video on the new Tektronix uh 2 series scope. Uh, sorry, I haven't finished a full review. that'll actually uh, but take some time. leave it in the comments down below.

if you want the full review, that's my uh plan. I just yeah, things didn't pan out uh, unfortunately. but I've got the tear down video of course and I was going to throw this on the second channel as just a quick video over on Eev Blogger 2. As I often do, you should be subscribed over there if you're not past 100 000 subscribers is where I put lots of secondary content.

But hey, everyone on Twitter the majority said that they wanted to see this on the main channel. So here it is: main channel video. So this is not a review of the new Tektronix 2 series. It's just looking at waveform update rate for this thing and we're going to compare it with some other brands and I'm going to show you how to measure waveform.

Updating: Let's go. Okay, let's measure the waveform update rate for this. Now I already know it's supposedly not that great according to Andy Ted in one of my second channel videos, but let's actually measure it because it's not on the data sheet unlike the Series three, which is, uh, here is the spec sheet for it: 200 and up to 280 000 waveforms per second. which is not too shabby.

So what I've done is I've set the auxiliary output here to trigger and I'm feeding in a one megahertz, uh, square wave like this. I've got nothing else turned on so it's doing absolutely a bare bones and we can actually measure the output here. It is here. So this is the trigger output.

So this is our waveform update rate in uh, waveforms per second and you can see we're getting 1.9 k waveforms per second, but you'll notice that it's very bursty like that, right? Is that a word bursty? Um, so in here we're getting all of our sample updates and I can freeze that and we can actually have a look and you know we're getting like a dozen or more like updates each. Each one of those is a trigger and you can't see it. Uh, because I don't have enough memory depth, but trust me, it's capturing that uh, properly. So anyway, we can redo that and I'll show you that here that that is actually a very small pulse in there so it's not very consistent.

It jumps around all the time and this is uh, consistent with like other like low end scopes on the market um that don't have something like the keysight mega zoom for asic of course, which uh has phenomenal update rate and we might actually do a comparison in a minute to actually, uh, show you. So we get like uh, 20 or something uh, waveform updates in there and then we have to wait, what? Uh, 20, You know, 30 milliseconds something like that before we actually get another one. But there's other periods over here where it's much bigger. But as you can see, it's uh, quite a random.

And if I do anything on the scope over here, you'll notice that if I try and move a waveform, it completely stops. Okay, there's no more updating there. and if I pinch and zoom that, boom, It is not updating. and of course you can see it on the screen over here.

It is just not updating. and well, you know that's fair enough because you're interacting with it. and if I drag the trigger level here. Yeah, Trigger level? Same thing.

no updating, and the absolute best waveform updating rate I'm able to get on this thing is 18.8 k waveforms per second. And that's at like a real fast time base like 10 nanoseconds per division, 250 points memory. Um, and I'm also in normal trigger mode on this scope, so I cannot get better than that. And if I slow down the time base, you'll notice this is still the same until I hit a hundred nanoseconds per division here.

which gives us seven and a half k waveforms per second. Then it jumps to 3.8 kilohertz here. And then if I go down lower than that, I've got to change my time base down 1.9 kilohertz. At 400 nanoseconds at one microsecond per division, we're now down to 760 Hertz.

So yeah, it's It's not that great. but uh, we are getting increased memory here. Oh, one incredibly annoying thing. Check this out.

If I want to set my manual memory depth okay, I can go into horizontal mode. Manual record length: 25k. Okay, you think if I go up it would go to 50k or something? Nope, goes to 26 27 and then it adjusts the micro time base. Over here, my horizontal scale is now 1.24 microseconds per division.

Why? Why I don't want that rubbish? And if I go in here like this and I want to update the k points per second, then let's just say I wanted like you know, at low like a thousand points, it's actually K points. I've got to go No for to change the scale over here and enter. Okay, so we're now 1k point. But the annoying thing is is that it doesn't matter what I do.

I can't seem to have a fixed record length of 1k points. It doesn't matter whether or not I have the sample rate changes affect horizontal scale or record length. Watch this if I change my time base. Okay, it's changing my K points, right? I've told it no.

I only want one K point. So I've got a this is manual mode. Okay, why why this is just ridiculous. Check this out.

This is crazy. I'm at one microsecond per division. Okay, and I want to change my record length. I want to.

Actually, I, I, I don't want 25. I only want 1k. Thank you very much, right? And I put 1k in and it's changed me back to 40 nanoseconds per division. Because I've got this sample rate changes effect horizontal scale.

I can change this to record length. I can change this back to 1 microsecond per division. I can go back in here and I can go 1 k points like that and it's taking me back to 496 square division. I can't set a manual bloody memory depth on this thing and then change horizontal ranges.

It's ridiculous now. Just remember when we measured that frequency there to calculate the effective number of waveforms per second. That doesn't include the dead time. That's just what it's capable of if there was no dead time.

But thankfully, the Tektronix actually gives us a display of the wave for effective waveforms per second. Here it is. It gives us the number of acquisitions, which is an acquisition is a complete waveform acquisition. So this is the number of waveforms per second and you can see it counting up.

Three thousand, Four thousand, Five thousand. So that's probably about two thousand odd uh waveforms per second. and this is at the maximum time by. well, this is ten nanoseconds and you'll notice that it resets every time I change the horizontal time base so we can actually get a stop.

Watch out and at the different uh time bases. So this is uh, 40 nanoseconds per division? Sorry, it's off the screen, you can't see it. There's 100 nanoseconds per division? and as we saw before, if we go to one micro second per division, there we there we go. Ready.

Let's start again. There you go. So you can probably use your own stopwatch and count those yourself. The number of waveforms per second.

It's not many, we're talking about a thousand and that took like 10 seconds or something. Okay, so I can get a stopwatch here and let me, uh, try this at 10 nanoseconds per division and go. And then we can actually get a figure for the real that, the maximum, What number of waveforms per second? So let's give that say 30 seconds and I'm going to say 30 seconds is about now. 43 000 acquisitions in 30 seconds.

So get the confuser out. that works out to about 1400 waveforms per second. That's the maximum this thing can well actually does in practice with this sort of like de facto reference standard 10 megahertz input with uh, 10 nanoseconds per division at 250 points down there. So yeah, it's not great is it? Anyway, we've got our 18k waveform updates per second.

so we've got our maximum. Okay, so I'll leave it about there. Well, actually, I'll dial it one back like this and we'll be able to see as we do various things on the screen. Now if I turn on the other uh channels, let's go to three.

Whoa. look at that. Extra channels really slows it. So yeah, even if I turn channel 4 on here, it it does that.

So let's turn channel 4 off and Bingo. Yeah, we do. If we have a single channel, we do actually get not faster waveform updates per second, but less of a dead period. A dead time actually between bursts of acquisitions.

Anyway, let's turn our cursors on, shall we? So we'll go up here. Cursor's on. Yeah, that adds a little bit of dead time. Turn our cursors back off and we'll do some measuring as well.

Let's measure, you know, Rms and peak-to-peak and stuff like that. And uh, hasn't slowed it down. No, it hasn't slowed it down a huge amount. but there is some extra dead time there.

Then if we turn on, say Fft because that's always a good example of like some really advanced stuff. And there we go. It slows down for a bit while it's actually thinking about it and doing something internally to change into Fft mode and set it up and everything. But once it, you know once it actually turns on the Fft, it does.

Actually, I get back to well, not as good as not having all these things on. It's actually still not hugely bad. Now of course, the update rate really doesn't have anything to do with, like how responsive the scope is in actual use in actual use. It's actually, um, you know it's It's a fairly responsive scope even with Fft and cursors and measurements and everything else on.

So it's actually it's not too bad. But yeah, for a new architecture that, um, well, what? It's been 15 years since the last tech arc. And like, low end tech architecture I believe around about that, sort of, uh, period. for the Mso 2000 series.

Um, yeah. this new Lexington architecture with the Xilinx Zinc processor. It's just. it's quite frankly, uh, rather disappointing.

Um, I just didn't really expect it to be that poor. I expected a new scope to have. you know, a decently fast update rate. It's a shame because the three series, even though I don't have one um, and I've never measured it on, it apparently has up to 280 000 waveform updates per second.

And this new one, um, yeah, it's cheaper, but this low-end Keysight 1000 series has 200 000 per second. So, and interestingly, even with uh, Fft and cursors and everything else happening here, we do actually get more updates in an actual burst like this. so you know then we'll get in. You know, like 20 odd before or something like that.

So yeah, that's curious. but you can see that we're 50 milliseconds per division. so we're getting like 150 milliseconds or something. You know, dead time where it's not acquiring anything at all.

Okay, just to show you the difference here, I'll compare the new 2 series with my Keysight 1000 series. I believe this is now 200 000 waveforms updates per second. but the one I've got here is only 50 000 because the Mega Zoom four in here is capable of one million uh, per second. But you? they don't enable that unless you spend more and more money as only uh manufacturers do.

It's not just tech, Um, anyway. Uh, so this one will definitely do a million per second. but I'll just show you this little uh puppy here which is, um, you know, the low end of the keysight uh range and we'll compare the difference. So I still got the Tech 2 series out here.

And there you go. That's what we're getting. That's 100 milliseconds per division. We're getting a little burst there and so we can go in and we can measure our burst.

The frequency is now up here, so that was the well. You probably just saw it. I can freeze that and that's the 18 kilohertz that we get there. Okay, so we go back.

100 milliseconds per division. So let me now change this over and I won't touch anything else here. So auxiliary out. I've actually set my function generator output on here to trigger output.

That's a software feature in here. Not all scopes, by the way, have the ability to output a trigger signal, so if your one can't then it's more difficult to like measure the waveform update rate. You have to do it in roundabout ways. Um, but anyway, let's now plug that in and boom.

Look what's happened. Look at this. There's hardly any dead time. There's just this little boop boop.

This is all this is all. Look at this. here it is. and there it is.

50 050 k samples per second. There it is, right. So 50 000 waveform updates per second and you can see that there's just a little bit. we're triggering on the one, but there's a little bit of like jittering in there because the A6 is doing.

You know, whatever. But look, that is the difference, right? So it's almost always sampling right. Look, just all the time. There's no dead appearance, and sure, if I, uh, you know, tweak the vertical and stuff like that, it's still.

You know that there is some extra dead time in there to do. uh, you know, stuff like that, change the vertical divisions and change the horizontal and stuff like, right? I get little bits of dead time in there, but it's basically it's pretty much sampling, especially when you're not touching it. It's sampling all the time. and I've turned on my measurements down here and I can turn my cursors on as well.

And I can fiddle fiddle with the cursors. And even when I'm fiddling with the curses right moving the cursors around on that waveform, there's like there's really no difference in the waveform there. Let's see if there's a little bit of difference in the jitter? No. See, there's no.

I'm I'm moving that cursor across there doing that. It's only if I start. well. no, I actually can't even see that because there's so little.

Oh yeah, yeah, we got. Hey, we got one. There's so little dead time in this thing that, um, yeah, you just. you really don't notice it unless you're you know, really changing ranges.

very fast, but even then it's like orders of magnitude better than the new architecture Tektronix. And this is a 12 13. I think it's now. I think the Mega Zoom 4 Asic in this is now 13 years old.

Come on. uh. Keysight, When are you gonna? I'm doing Agilent? Come on Agilent, When are you gonna? Um, release the Mega Zoom five? But yeah, um. there you have it.

So this will only slow down if we go at slower time basis because it's got to feel that it's going to take time to fill the memory up. And the keysight uses automatic, uh memory. so you know we don't know exactly how much it's using. But anyway, I'm at 100 microseconds per division.

Sorry, I can't show you this and show you the detail on the frequency there. But uh, trust me. And if I go up to the fastest speed, that's uh, two nanoseconds per division, it's still the same 52 kilohertz signal. So if I slow it down 200 nanoseconds, 500, right? it starts getting a bit slower and we can.

Oh, do that. There you go. So two microseconds that's 22.7 Kilohertz and let's go to 50 microseconds. We'll have to single shot capture that 1.8 Kilohertz.

So you know the slower it gets because we've got uh, a slower time base on here means it's got to fill up the memory and it's got to take that amount of time. It just you know it's all. Scopes are going to slow down and the million waveform updates uh per second on these keysight scopes means nothing if you're using it at a slow time base and it's got to fill all that memory it's you know. can I beat the laws of Physics Captain Just to show you that in the real world, I've got uh, tech's own demo board actually? Uh, the Mdo demo board.

I've got the rare anomaly uh, set in, so I'm not sure of the exact details of that. You can probably go look up, uh, info somewhere, but anyway, there you go. I've got a hold off in there of 2.2 microseconds to try and to get the display stable. And hopefully you can see all that well.

Well, yep. Oh yep. Yep. Yep.

There should be like a runty pulse. An anomaly? Well, Yep. see it. See it.

This is the fast update rate of the keysight anyway, so you can see. Yep, Yep. saw another one. Yep.

got one there right? So you can see that Now I do the same signal on the tech. There you go. exactly the same settings. 2.2 microseconds are hold off time.

Uh, so that you know so that it doesn't Otherwise you get jittery, trigger and stuff like that. so you've got to hold off on complex waveforms like this. But yeah, I'm waiting for the cows to come home. Um, I'm not.

I'm not seeing anything, not seeing any jitteries, not seeing any runt pulses. Nothing. Not a sausage. and that's due to the, uh, small number of uh, waveform updates.

But why are you? No, I think I got one. I think I think I might have captured one there. But yeah. wow.

You gotta wait a while. Um, and yeah, it's a purely a uh, percentage chance of it happening based on oh well. did I see something there? I was too busy looking at the camcorder. I thought I saw it at the corner of my eye.

But yeah, it's a basically a percentage chance of capturing something? Uh, well. maybe. maybe. But yeah, you can see the real difference there.

And I swapped it over to the frequent anomaly and you can see it's picking up. We saw it. Yeah, see. So this is picking up the frequent anomaly there.

You know it's once every couple of seconds, maybe five seconds. this one's going up there. We go there, we go right, and we'll do the same on keysight. Swap it over, haven't changed anything, and we get it all the time.

It's like it's constantly There it is constantly there. So then you can go, aha, I got you your little mongrel and then you can set up a run, pulse, uh, trigger, or some other triggering. um to actually capture. Uh, that sort of thing.

Yeah, so that really illustrates, uh, the difference that uh, waveform updates uh, can make. And yeah, I'm just disappointed in the new tech architecture. I'm you know? I? jeez. yeah, I expect it better.

and even if you compare it with this uh, Siglund Sds 1104 Xe, um, and you know, nothing special like low end scope and look, you can readily readily see it. No problems whatsoever because it's got a better update capture rate than the new Tektronix Lexington Architecture Whizbang 2 series disappointing, and the old firmware 50k waveform version of the 1000x series. You see that frequent anomalies, just it's showing up all the time whereas we were waiting. You know, up to like five seconds in one point.

You know, three, five seconds or something on the two series tech to even see anything. and this is like multiple times per second. Boom. It's just so easy to detect here.

Look, it's nuts. And this old Rygold Ds2000 series don't know how old it is now, but geez, can't remember last time I did a video on this. um and there you go heaps. But here's the thing right once you include once you increase your memory points, okay, it's going to go down.

so we're on fixed 14k point memory there, which is heaps for just general use and screen. So really, when you're operating your scope. pro tip is that you know you should choose the lowest amount of memory for the fastest update rate possible for a given time base. And that'll help you detect infrequent stuff like this, which is going to rely on sheer statistics of whether or not it's actually picked up in the acquisition or not.

Um, of course, if you've got a scope that has zero dead time is just acquired all the time and it's going to present that information to the Uh screen. it actually decimates it because you can't even on the key side over here, right? You know, Million waveform updates per second? You're not actually updating. refreshing the screen a million times a second. That's just ridiculous.

So what it does is actually uh, takes those samples and effectively, uh, like, combines them into like a displayed version of that waveform periodically so I don't know how often it actually you know it does that. But then it's able to pick up stuff like that and actually display it. So we'll change that to 140 k points here and you're still seeing it. Okay, let's go to 1.4 meg points and there you go.

less frequent because the update rate is going to drop based on the amount of memory it's got to. uh, process 14 meg points? Did we get one there? I don't know. You're going to have a hard time seeing it now, so don't Always Oh well. There we go.

Don't always think it's the best thing to actually set your scope to the maximum memory depth? It's not always good Anyway, I hope you enjoyed that little comparison of waveform update uh, rates and how to actually and measure them. And it's yeah, just a tad disappointing from the new architecture. Um, Tech 2 series. Like I don't know why? Is it a hardware limitation? I mean I can't understand how it would be a marketing position thing to actually deliberately that I don't know.

Tech will have to, um, tell us because the next one up in the series the three series as I said has a maximum of supposedly maximum of 280. 000 per second. So yeah, that's like that's like really good and modern. but the new two series for that aspect it.

uh yeah, it gets a thumbs down. It's disappointing, but as I said, it's actually a reasonably responsive scope and when you turn things on, it doesn't really slow it down all that much. So as I said, waveform update rates doesn't actually correlate into a slower, a slower user interface experience. It just means more dead time because it depends on how you prioritize the thing in the design of the user interface in the acquisition engine.

Whether or not you're going to prioritize the user interface, which you should, you don't want to slow operating scope but then also you want the fast update but that's why something like the keysight does it all in and asics so that it you know it doesn't matter what you turn on Ffts or any of the other cursors and measurements and all the other processing functions, it it doesn't matter because the main processor is not handling that sort of stuff, but you saw the dead time on the tech here. It was quite substantial, but you know it is still a fairly responsive scope. Just got to live with the dead time you.