Hi Some people wanted me to take a look at the new Sieglin SDG 2100 X-series true of function generator which starts at 499 U.s. dollars 1.2 gig sample per second, 120 Siz 120 megahertz model the $4.99 one starts at 40 megahertz I Wanted me that take a quick look at actually using this thing and maybe plug the output into a spectrum analyzer or something before I have to send it back and yet we can do that and we can compare it with the older model signal entire SDG 50 82 And also we can have a quick comparison with the Rai Gold EG 4162 as well which you've seen in previous videos. so let's take a quick look. This certainly won't be our comprehensive, but we'll just give it a go and we've got some spectrum analyzers here.

the Tektronix MBO 3000 a try, Gold, Esa Eight One Five but I think the but the Tektronix has a better. We're getting a lower noise floor and I've done a couple of quick tests and I think we'll use the tech in this regard today. Now off the three units here, I'll quickly recap the new Siglent SDG 2100 or 2000 X-series the lowest model unit, the 40 megahertz one starts at 499 dollars. It's a 1.2 gig samples per second 40 megahertz for the entry level 1 up to 120 Megahertz 16-bit converter in it as we saw in the teardown through 16-bit converter with 16 Meg of arbitrary memory.

the this older model siglent which they still are sell I'm not entirely sure why, but anyway, maybe they might discontinue. Anyway, we've got it as a reference point. the SDG 5000 series. This one starts six hundred and eighty dollars for the I think is at the forty years might might be the 40 megahertz up version once again, and that's got 512 kind of sample of memory, so not nearly as much.

And it's only got a 14 bit converter and the Rye Gold DG 4000 series. Once again, it's only got a 14 bit converter as well, so that's where the new X-series wins. It's got a 16 bit converter 16 Meg's sample memory. So the entry level model this the DG 406 to is 60 megahertz.

This is the hundred and sixty megahertz top-of-the-line one, but the 60 megahertz bottom of the range unit. It starts at nine hundred and twenty dollars. So much more expensive than the X-series down here, but at 60 megahertz bandwidth compared to what 40 for the base model unit. But it's only a fourteen bit converter and only 500 mix samples per second.

This is more than twice as much at one point. two key samples per second and this has only got 16 K points sample memory as well. So, but as you can see, it's a different form factor. Rygel actually have the DG 5000 series I think it is, which is a similar form factor to these that sequence down here.

so which is probably a better comparison. This DG 4000 is a bit different. This one has a lot more building waveforms out. 130 building waveforms.

Much more comprehensive. Our arbitrary waveform support built in than either of the sequence. So just show you the waveforms. On the RAI goal here might have very much more comprehensive built in waveforms than I Mean it's just got a ridiculous array, You got math functions and this isn't even the complete list.

You can actually scroll through these lists. Trig functions, Any trig functions, window functions like it's just yet crazy. If you want building waveforms, this is the puppy to get. And in addition to all those arbitrary waveforms, the Rygel has a harmonic function as well.

I don't I Haven't looked through the siglent, but I Don't believe it actually has harmonic functions. You can actually choose what type you want on or even harmonics and the number of orders of your harmonics and the phase as well. So if you in depth you know it's just a more flexible instrument. The Ryo goal in terms of waveform generation I mean you can do that with the seedlings because it's an herb generator, but you've got to do it yourself using whatever software to look up and actually download the waveform to it.

You can't do it from the front panel. If we go check out the waveforms on the siglent it does have. You know it's got all your basics. It does have building ARB functions as well, but not nearly as many as the signal.

It's got a decent amount. Let's have a look. we can. Actually, by the way, it's got true ARB function which will have to take a look at I think Anyway, the ARB type.

We can do the built-in waveforms okay, and it's a similar thing. I mean we've got common ones here. we've got some math ones. We've got some yeah it engines if you're doing, but I don't know? Well, they've got cardiac pulsing engine anyway.

Bizarre. Anyway, there are some windows and Triggs and that's you know. basically it's It's a reasonable selection, but not as many, not as comprehensive as the Rygel and in addition to the waveforms, we can do all your modulation of course, that's all you know. That's all pretty much on par.

We can do sweeps as well, which is all fine and dandy, so no problems there at all. and we've got burst stuff as well. so it's alright. but as I said, doesn't have that the harmonica feature if that's useful to you.

But of course the siglent is actually touchscreen and I can go in there and actually touch. and I can select these menu items as well with the touch screen. But I don't really see the point of the functionality of the touch screen on this thing. quite frankly.

I mean you know you can. You can select a waveform. you can kinda sort of go in there and kind to get your finger on those. but if you've got a big finger, my fingers not big.

it's quite small, but I'm having a hard time sort of. you know, getting on there. it's just easier to use the function button. So the touchscreen: I'm gonna say bit of a gimmick I Mean it's not like you can go in there for your herb one and actually like drawing a waveform with your finger or a little pointy thing or something.

That'd be kind of novel, but yeah, like it's I. Don't get the point of the touchscreen, really? I Don't think I'd use it and all these function generators are dual output here, so no problems whatsoever. User Interface: I Prefer the siglent one. it's much more intuitive.

The RAI Goal: As I said, the user interface on the Rygel is a pain in the ass. It really takes some getting used to some really weird stuff you know, tiny selection dots for your frequency and it's just yeah. don't get me started. but both the Rygel and the Ceylon somewhat annoyingly like you'd expect.

You know, like to have like an output menu here, but this is just like our put on or off which is you know which is fine but how do you change the output impedance? for example, Well, you can touch it there, but if we want to find that on the menu, you go into utility here and you're going to output setup and then that's where you do your load. You know I I Don't know I just don't look and it just popped back as well. So that's kind of a little bit annoying that it doesn't stay in the stay in the menu there. So 50 ohms BAM It just goes back.

Don't like it anyway. I'm just going to work the output of each of these signal gems onto the Tektronix Mdo spectrum analyzer and just have a look at the output. I'm just going to have a nominal wire 10 megahertz sine wave. You know, half a volt peak-to-peak and let's see what we get.

And there it is. There's our ten megahertz carrier. We've got a couple little outside components here. they're at ten point.

Okay, there we go. There are markers of automatically ten Point oh, two nine one, and nine Point Nine seven. They're at minus 92 where two DB minus two DBM down on the carrier. So we're talking about.

You know, minus 90 DB See there or our reference to the carrier there and our noise floor? they're looking. You know just about what a hundred and five Minus 105 our? DBM Maybe Let me disconnect the input and see what the noise floor of the scope is and there we go. That's a scope noise floor about, you know, - 115 or something like that. and that's the exact same signal there on the siglent.

SDG 508 - the 5000 series and as you can see, not nearly as good as the singlet. Look at those sideband components Wow They are massive, so remember that's a 14 bit converter compared to the 16 bit in the X-series But geez, Well, let's take a look at the right Go now. and there you go. That's the right Go! I'll just save that, take a screenshot and there you go on.

The sidebands here are a little bit higher there at Minus 89, so you know - 86 or something relative to the carrier and looks like we've got a little harmonica that over here. So yeah, it's a little bit worse. but yet, noise floor between them is not much doing so. Once again, back to the Siglent X-series the new one and as you can see, it's much much cleaner.

But have we got a little something in there? perhaps? Hmm. let's tighten that span up a bit. Where it? What were we before? 200 Kilohertz span? Now at 50 Kilohertz Span and these guys are they gone? Whoa. Uh-huh Didn't hold my tongue at the right angle.

doesn't matter, Nothing doing there. And there's the output of the X-series when we're looking at a 50 milli volt peak-to-peak signal here. So we're at Minus 20 DBM reference now and we're look that's clean as a whistle. We're back to our 500 Kilohertz Span now and there's the Siglent SDG 5000 series.

And once again, that ain't pretty. The X-series much much cleaner and there's the output of the 5000 Series Siglent with the span set to 20 kilohertz. so you can see those components a bit better. And there's the output of the ROI goal.

Or once again, that's you Know, it's not too far off the signal, but the Siglent X-series is. actually it is a cleaner output. look lower side bands and we'll just check the spurious performance of the RAI goal here, the 4000 Rai goal. And once again, we've got our 50 milli volt peak-to-peak carrier at 10 megahertz.

Here you can see some more components the highest one here is it mine? almost - at 90 DBM down there. But you know that's reasonably clean where our span is over 100 megahertz now. So let's check that on the Sigler and that's the Siglent 5000 series. And look at that, what's going on there? No idea what that business is.

But anyway, it's similar to the Rye goal and there is the new X-series With, you know, decent performance. Well, let's put the new signal and X-series up to 80 megahertz. There we go. That's its performance.

once again. 50 millivolts, our Peter peak - 22 DBM and check out the seedling 5,000 series. Once again, we've got some weirdness happening down here. Some decent Peaks are not nearly as clean as the new X-series Hmm.

and there's the RIE gold DG 4000 clean as a whistle. and as a reference, this is the output of my Marconi 2023 signal generator. Yeah, it's a bit higher performance although, but don't know what's going down here on here. Down at the low end though.

but yeah, look at that clean as a whistle and there's my Marconi. For those who haven't seen it, it's pretty Shmi And I'm gonna have a winch again about the Tektronix MTA 3000. It's a slower wet week look, it doesn't even detect the button presses when you've got a really low resolution bandwidth filter on the thing, so it just becomes unresponsive. It's God's really rather annoying, but hey, it does perform.

Oh, it popped up eventually. half the time it doesn't but yeah, only when it's yes, set to those low resolution settings. Menu off. damn it.

Menu off. I'll just show you why I found it. You can't actually upgrade the bandwidth with a license key on this thing. So if you buy the 40 megahertz model for a hundred and ninety four, four hundred and ninety-nine bucks, sorry, geez, hundred ninety-nine be a bargain $4.99 which is still a bargain.

Yes, you can, actually, uh, presumably buy the license key and upgrade that I don't know how much that cost though. All right, let's have a look at its square wave performance. I've just got a 1 megahertz R square wave one volt peak-to-peak on the output here and I've got some averaging set on the scope of course, and we get a little bit of ring in there, but that's exactly what you'd expect. Yes, I am our 50 ohm input terminated on the scope by the way, so that's no problems at all.

Let's switch that over. In fact, I'll leave it right there and I'll switch it over to the Siglent 5000. You can see the average in doing it's doing its business there and the Siglent 5000. there we go.

That one's A. You can argue it's a little bit cleaner, but the you can tell the higher our sample rate of the Siglent X-series on there. No problems whatsoever. So and so we not.

Let's go to the right, go Zack with the same signal and there we go. It's very similar. Got a little bit of something happening in there, but then T you know, like it doesn't matter, they're all fine and by the way, the fastest rise time you can get on this is 8.4 and nano seconds whereas fastest ride to rise time on the ride all 5 nanoseconds so just a little bit faster. Edge rate in Siglent claim to have this easy pulse technology and here's a explanation from the datasheet here.

It tells you that if the in pulse mode if the output frequency is not an integer multiple of the clock rate, then well, you're going to get a one clock sample jitter on the thing and well I set up the same conditions that they've got here and I cannot reproduce that on the try goal 4000 so the Rygel 4000 doesn't have the same problem. I'll plug the exact same signal here into the singlet X-series There it is: Siglent X-series I'm What I'm doing is I'm measuring the the rising edge over I'm triggering here, measuring looking at the rising edge here. So that's the Cichlid X-series and there's your eye goal there. No difference.

and that's using the exact same condition they have with the 1.0 1 megahertz, 50% duty cycle pulse waveform. But here's where the Siglent X-series has a massive advantage with this easy pulse thing. If you generate a 10 Hertz like very low frequency pulse waveform, it still allows the rising and falling edges to be that minimum 8.4 nanoseconds. You might think well, yeah, that's ordinary, but let's have a look at the Rai goal.

Okay, we're in pulse mode the same 10 Hertz but the like trailing edge and leading edge. Okay, Look, it's set it to a minimum of 195 microseconds. Look, if I try and put in 1 nanosecond, it'll tell me the minimum is 195 microseconds and so that's the limitation. the DDS technology used in the RAI goal, but the Siglent doesn't have that.

Let me show you on the scope. and I've got a 10 Hertz output here. But as you can see, they're not the same edge because they it's the classic two freerunning clocks issue. it's a trap for young players.

This one. To show you this: if you've got two freerunning oscillators like we have here, you don't know where it's where they're going to start, relative from one to the other. So you've ever got to reference them from the same master clock or right trigger them in some way. Or maybe maybe actually set a delay output or something like that.

So yeah, I Just want to show you that classic problem with two separate freerunning clocks. I Always love that. And eventually, if you sit there long enough and they're not both exactly ten megahertz reference. I've done a video on this measuring clock, our drift: You'll actually see them slowly drift relative to each other.

So if we wait long enough and then not both, exactly ten megahertz oscillators inside that, well, they don't actually have to be ten megahertz precisely, but they have to be exactly the same. Otherwise, if they're not exactly the same, you'll see them slowly start to drift. And I'm bugger if I can find a way to, actually you know, synchronize or trigger the start of the pulse signal here to the external input like there's an external In Out on the back, which out you can use for sweep mode I believe to start the sweep, but not in pulse mode. So anyway, it doesn't matter because we can do it with the delay.

We can just yet tweak the value in it, hold your tongue at the right angle until you get the signal. We can actually measure it on the scope and then just punch it in. Okay, I found what I would call a bug. Okay, we've got our buttons here which select our cursor position right.

Everything's hunky-dory so for one hour, just the delay like this, right? it works fine. But look what happens if I go over here and then I go down like this: 76, 70, 5, 4, 3, 2, 1 Look what happens. It jumps cursor position over and then goes to the next digit. Damn, that's annoying.

That is really freaking annoying. Why have they done that? It's got to be a bug. Surely, because it doesn't do it in the up. Direction It stays put only on the down direction.

Anyway, it turns out I can't fix this delay because we've got a maximum of a hundred milliseconds delay here so we can't actually get the waveform over I'm gonna have to invert it on the scope to actually be able to line up these waveforms so that's just a bit of bad luck. I Mean you know? look if I adjust a delay, look, you see, we can. That's that's that feature. they're jumping down by one.

Yeah, that bug I was telling you about. See, we can't quite get there. We can get with the different edge here, but yeah, we just can't delay it enough. But that's okay.

I can just go into channel 2 here and I can invert channel 2. No dramas there. we go. now.

we can line up our edge right now. Here's where the wheels fall off the Billy cart with the RAI go. And here's where the Siglent X-series saves the day. You can't see it, but I've actually got both waveforms in there.

Ok, so they're identical. We've got a 10 Hertz waveform, but the minimum are rise in full time if that. Rygel was that massive. 200 microseconds.

Just saw it. So if we go in there and actually have a look, look, look at the the yellow waveform. There is the Rygel Look at that. You can see the individual steps there.

but look at zoom all the way in and there's the blue. One of course is our siglent. Absolutely perfect. It's a Bobby Dazzler look at that.

But look at the horrible. Well, in this case, our full time. It'll be the same for the rise time on the Rygel. It's absolutely atrocious.

So that's the difference between the DDS output. You can actually see the individual samples there. Wow And you can actually see that drift I told you about. You see it slowly drifting, slowly drifting towards there.

It means the two clocks that the two freerunning oscillators inside this thing aren't exactly the same frequency and there's nothing wrong with that. That's just what you'd expect. But yeah. see, we can see the clock drift.

There Be pretty close. Hmm, You'll see exactly the same thing on the rising edge here. Look at that. so that is the massive difference and a big limitation of the RAI Go.

A lot of other DDS generators say that's a win for the Siglent X-series that easy pulse technology. Very very nice I Like it and even though it's 120 make function generator, well you try and do 120 megahertz, square wave and no siree. Bob 25 megahertz maximum and we can actually get 20 volts peak-to-peak output into a high impedance upload of course at 1 megahertz. but that's not going to be the same over the entire frequency range.

So let's actually do an experiment. 10 Megahertz Okay, we're still 20 volts peak-to-peak you know? let's go 50 ere. let's go I Don't know. 25 megahertz.

Bingo! We drop down to 10 volts peak-to-peak and let's go for the fool, let's go the Full Monty: 120 Megahertz Yeah, we can still even at 120 megahertz, we still do 10 volts peak-to-peak so that's not too shabby. And sure enough, at 10 megahertz we can do our 10 volts peak-to-peak There we go. Where do you know? 10.10 point, 2 volts peak-to-peak Near enough. But if we increase the frequency on that, let's go up to say 50 mega Hertz It's dropped a bit nine point, nine point, eight or there abouts.

but though this is to be expected. But when we just want to see what it's like and let's go the Full Monty 120 Megahertz it's still doing okay. Now that that's that's as performance. 120 megahertz as you can see, it's starting not to be a really good looking sine wave there.

But yeah, that's what you get as a comparison. Here's the Rygel at 120 megahertz I Granted, it's got now put rated output bandwidth over 160 megahertz, but it only allows 2.5 volts peak-to-peak maximum. But as you can see much cleaner sine wave, let's actually change that to a hundred and 60 megahertz. it's maximum output.

So there we go. It's dropped significantly in our The Chariot, but it's still a pretty good-looking sine wave. so if we plug the siglent back in, you can see that it's a yeah, it's some sign. Performance at it's rated bandwidth 120 megahertz.

it's not that great. So yeah, it's a lose there. And by the way, it does come with a Cal certificate. Just found that in the box and by the way, it's shorter and a bit more compact than the five out the older 5000 series.

So I rather like that and the only thing you get with it a CD I Haven't looked at it, presumably the manual and maybe the waveform editing software and account certificate and and the power cord in the box? nothing else so that's a bit disappointing. What can they throw in a couple of BNC cables or something? Come on. So I think that's about all I've got time for today I Have to actually send this unit back. It wasn't designed to be a full review.

I mean I haven't got time to test, like, install and test the arbitrary waveform software and stuff like that which is a big deal with an arbitrary waveform, it, and generator. If you're going to generate them, how well it integrates with the oscilloscope for example, or if it does at all, the sequence scopes to be able to capture or other scopes to be able to capture an input waveforms and then out put them on a arbitrary generator like this. That's one of the benefits of arbitrary generators. You can actually capture and then simulate that same waveform and then slightly modify it and stuff like that.

That's one of the true benefits. and this thing has more bells and whistles and you can poke a stick out. And I haven't even looked at half of them, haven't even tried the dual outputs and everything else. But yeah, and all that synchronization and modulation and sweep functionality and burst functionality.

But it's jeez. I'll tell you what, for 499 bucks for the entry-level 40 megahertz Won! Wow. This one's got winner winner chicken dinner written all over it. There's a few little quirks in it, but I don't mind the operation of it.

It's performances are pretty good, you know, 16-bit converter at 1.2 G examples per second. kinda does the business and well for the price you kind of complain. It's certainly seems to I kick the pants off the rogue or one anyway. bang per buck.

So there you go I Hope you found that I Yes, I First looked first impressions our review of this thing and if you want to discuss it, jump on over to the Eevblog forum. Links down below: leave comments all that sort of stuff and thanks to Charles at Trio Test and Measure for loaning me this one, it's the only one in the country at the moment and he wants it back I Just got a pestering email. Whether you're going to be finished with that, Alright Charles I'll take it back now. Thanks mate.

Catch you next time. Oh, and by the way, I'm the none other than the siglent CEO will actually be coming to the Eevblog lab on his tour of Australia So there you go. How many CEOs do that? That's in December early December sometime. So I guess if you've got any questions, um I think that I'll like I'll put up like a forum thread for it or something perhaps.

Anyway, you've got any questions for the cichlid CEO himself? then yeah, please leave them. I'll try and get him to answer them. you.