Teardown of the new Siglent 1000X Series Oscilloscope
How to remove warranty void security stickers: https://www.youtube.com/watch?v=colVXihqdGo
Datasheets:
http://www.latticesemi.com/Products/FPGAandCPLD/MachXO.aspx
Logic Level Translation: http://www.nxp.com/documents/selection_guide/75016799.pdf
Synchronous Pipeline Burst SRAM: http://www.netsol.co.kr/upload/f_prod/S7N80xx31M_R1.0(0).pdf
Micron DDR3 memory: http://www.micron.com/parts/dram/ddr3-sdram/mt41j64m16jt-15e
HMCAD1511 1Gsps ADC: https://www.hittite.com/content/documents/data_sheet/hmcad1511.pdf
AD8370 Variable Gain Amplifier: http://www.analog.com/media/en/technical-documentation/data-sheets/AD8370.pdf
LP3878 Low Noise Regulator: http://www.farnell.com/datasheets/1896335.pdf
ADF4360 Frequency Synthesiser: http://www.analog.com/media/en/technical-documentation/data-sheets/ADF4360-7.pdf
Blackfin DSP Processor: http://www.analog.com/en/products/processors-dsp/blackfin/adsp-bf526.html
Madman Muntz: https://en.wikipedia.org/wiki/Madman_Muntz
EEVblog Main Web Site: http://www.eevblog.com
The 2nd EEVblog Channel: http://www.youtube.com/EEVblog2
Support the EEVblog through Patreon!
http://www.patreon.com/eevblog
EEVblog Amazon Store (Dave gets a cut):
http://astore.amazon.com/eevblogstore-20
T-Shirts: http://teespring.com/stores/eevblog
Donations:
http://www.eevblog.com/donations/
Projects:
http://www.eevblog.com/projects/
Electronics Info Wiki:
http://www.eevblog.com/wiki/
๐Ÿ’— Likecoin โ€“ Coins for Likes: https://likecoin.pro/ @eevblog/dil9/hcq3

Hi Welcome to Teardown! Tuesday Yes, we're gonna take a look at this brand spanking new hot-off-the-press practically not available yet Siglent STS 1000 X-series oscilloscope. You know we say here on the Eevblog, don't turn it on again. take it apart. As with most little modern compact scopes like this, looks like we've got some screws.

Four screws down the bottom here and well, where are the matching ones on the top? Well, yeah, and the handle there. and look, this is interesting. They've got no threaded nuts on, no nuts on the B and C's That's nice and convenient to take it apart. Beauty once again thanks to Charles a trio tested measurement whose is kind of reluctantly agreed to let me tear down the only unit in the country.

So I don't break it, Is it just me? Order those little gold QC pass stickers not instill in a QA and QC pass stickers not instill a lot of confidence I'm just used to seeing those on the two-dollar ebay crap items you get I don't know, it's just gives me a bit of the heebie-jeebies. First of all though, what I'm gonna try and do is get this pesky warranty void if removed Sticker: a nice little siglent hologram thing there to know that it's genuine and try and get this off so that Charles will feel a bit better off. Maybe if he doesn't watch the video he won't know that we took it apart and well. hello threaded machine inserts.

but I think they might be going into plastic, but they've got some thread Locker on there too, so be interesting to see once we take it off. If there's actually a metal threaded insert in there, that'd be impressive I am in like Flynn Errol that is alright. So we're in and heavy shielding as we've come to expect on all of these type of scopes. Look, we've got another being see down there, unpopulated I wonder what that is? A second channel or something? maybe? Anyway, looks like we've got some sort of programming header there and there and there and an unpopulated header down in there so you ain't we can see the board down there.

looks like it's my. it. looks like it should be one big one big board under here, so we should be able to see everything on it. hopefully.

I don't have to like take the whole thing out, flip it over the other side. but anyway, we'll see. Um, power supply will be under here we've got. The ventilation of course is fine.

The fan sucks in here and well, no, it sucks. sucks from this side and blows out that side. and Bob's your uncle. That's about it.

So there are whether or not there's any additional heat sinking on inside on the well, they won't probably won't be a six, they'll be FPGAs in there. so let's take it off. I'm not the best quality metal work I've seen a little bit how you doing there if you can see that. but anyway, it's built down our price.

so there you go. As you can see, we did get our threaded in metal inserts in there. Absolutely brilliant. Um, other scopes have just got us self tappers anyway, starting to take the screws out here.
Yes, they do have the thread Locker on them, but all this metal work seems like it's like not lined up properly and sort of squeezed in a little bit. It's almost like they were, you know, half a millimeter out on the drawings or something. Anyway, not a big deal, just something I noticed and there you go. We're in like Flynn and that's all.

rather neat and tidy. Power supply is a small no Thor expected like a big long one in there, but they've just got like half size in there. Um, it's in. it's only not shielded can and of course that's pretty much mandatory.

Shake proof our washer up there, is it? Anyway, No, it's not shake proof water. You're better than have the thread Locker on there. So the mains earth is done really well. Got no problems with that at all.

We'll try and get the cover off that power supply and have a squeeze of it. Just one one cable coming over one cable. for the Phantom there's the board. Neat.

Okay, this looks pretty much par for the course for any modern our digital scope. I've got an analogue front here and here. Hopefully we'll get the I can off that. We've got our external trigger circuitry over here.

Little relay in there as I think I When I did the review of this I didn't hear any relays click in the front section. so I'm not sure if it has any. Anyway, we'll have a squeeze under there. This is obviously our main acquisition.

FPGA under there one, you know I doubt it's an AC come pretty sure it's an FPGA. It looks like they're thermal adhesive that heatsink under that chip, so unfortunately we're not going to get that off. Sorry folks. I Really want to be careful with this one? That's the only one in the country.

Anyway, it looks like we've got our one will sample memory we'll take a look at in a minute and anyway, hopefully get the can off. What else are we got and totally unpopulated? BGA And memory here? Well, no surprises for guessing what that's for. Here's our logic analyzer connector down here, so they haven't populated any of that so you can forget about like Hardware enable in just working in the connector Hardware Enabling Of course they're gonna get the price down by. you know, the logic analyzer modules don't have the chip, the memory, or any of the passives or anything else fitted.

So yeah, that's actually fairly a simplistic. We've just got some input protection there, and unless there's stuff on the other side of the board I don't know whether or not we'll get to that today. But anyway, I'm fairly simplistic. logic analyzer stuff.

By the looks of it, we've got a little RF connecter I'm down here now what that's doing I don't know. some sort of external reference clock external our test connector or something. These you know. these.

these cost them a little pretty penny each. So I don't know why they're populated in it there if it's not being used at all. I Thought they'd be trying to shave every last cent off this puppy. Anyway, all of our DC to DC converter stuff around there, it's not worth it's.
not worth looking at. We've got ourselves. that's probably our Ethernet up here. So Ethernet and USB around there there nothing doing here.

We've got our JTAG header up here. fan control that'll I don't know if it's under firmware control or not or whether or not it's just running full tilt. Anyway, Analog devices DSP Anyways, got ourselves a lattice FPGA will have a taking, take a closer look. dead giveaway what that puppy is doing and the associated memory here.

That's that's our display connector. So clearly that's our display driver display processor. So this has got a reasonably quick display update right? In fact, in burst mode, it's up to like a 400,000 waveforms per second. but you have to watch my review video to see the caveats with that.

Anyway, so yeah, we've got ourselves the front end. Maybe Is that our ADC down there? Our dual channel ADC Anyway, we'll take a closer look at that going in there acquisition engine. There's our capture memory that's probably duplicate chips on the other side so it could double the amount of memory There will have to have a squiz that looks like our main oscillator in there and of course our analog devices our DSP up here. that'll be doing all the the main GUI type stuff, the operating system and the you know USB the user interface, all that sort of stuff and our display ASIC / FPGA here.

So it would be funneling of the data most likely directly from the acquisition ASIC over here into the display. ASIC Here, it'll be totally bypassing this process. this processor be doing not the waveform updating, but they've been doing stuff like our math and things like that I Don't think they're doing that inside the display. basic down here and there we go.

Got an analog devices black FEM DSP I Think it's almost exactly like the one used in the RAI gold scope. so it seems to be the go-to device for these types of scopes. I Wonder why is that The is that the programming environment? is that? the cost is that the you know what is it we've got here for our display processor FPGA It's kind of these of an FPGA use of the CPL did. It's kind of like a combination of both.

It's the latter saw LCM X 0 6 4 Oh it. it's part of the mark XO family and you won't see around the outside of this. You won't see a FPGA sorry I'm a flash memory for the configurable bit stream because this is a non-volatile device, boots instantly, it doesn't need an external memory to hold the bitstream, and it's pretty pedestrian. It's like 10 bucks of digi cares like 640 lookup tables.

Not much 6k of distributive ram in there, so nothing special. Pretty pedestrian. FPGA And it looks like we've got a span chin flash memory that X on there would be how one of the production operators saying yep, I've firmware programmed that puppy and the DSP of course has some memory coupled on to that. These would be logic translators by the looks of it.
haven't even looked at the part number for those, but that'd be my guess. Dead giveaway that you know there's three of them. It looks like they're you know, in the I/o path. and well, sure enough, at second glance, if you actually take out some of the numbers in here 2, 4, 5 at the end of it, you might recognize you know 7, 4, HC, 2, 4, 5 and AVC there will.

That's a family. So this is a simple AVC 2, 4, 5. So yep, definitely used for a logic level translation. So it seems like those logic level translators would be converting between a level for the DSP RP and or for the FPGA.

But these FPGA Czar have all the requisite interfaces. you know, 3.3 2.5 1.8 at 1.2 blah blah blah. So yeah, not sure what the deal is there anyway. logic level translation.

so there's oscillator there I Don't know. doesn't even seem to have a brand on that puppy. Hmm. Resistors everywhere and lots of unpopulated caps.

Don't know that part number off the top of my head. Um, interestingly look battery missing for the real-time clock? Yeah, where is the battery? Oh does this thing not have a real-time clock and I can't remember I Haven't actually checked the firmware to see if it has a date/time stamp for any function? Maybe it doesn't Hmm. So there you go. What is that connector doing there? I Mean this is obviously a little driver for it for outputting whatever signal it happens to be, but they've gone to a god or a bit of expense there in the world.

I Don't know why. So what we've got here is some SRAM coupled onto the main acquisition FPGA here and decoding that part and I've already never heard of Net Soul before I don't actually recall them Anyway, we're looking at a nine Meg bit synchronous pipeline burst SRAM cells designed for you know, really fast, low latency update All that sort of stuff. Exactly what this thing needs. for that.

You know that 4000 waveforms updates per second burst mode that this sampling mode that this sort of that this Scopes actually are capable of But of course it's in a 512 K bit by 18 bit arrangement. So I don't know. Are they doing a checksum with the extra two bits in there? Anyway, that only works out to 128 K sample memory for that. So this is not being used as the main sample memory because it's got it's supposed to have that 14 Meg's So presumably that's what that puppy on the side here is for.

So that puppy there. And if you know them, that number off the top of your head, well, my hats off to you lotty cryptic. You got to go into the Micro. Note that that's what the M stands for.

They're the that's the Micron trademark, their logo and you got to go. They've got like a part number decoder. you're waking to their website. so they got this weird ass part code numbers in there.
Anyway, this is an empty four 1j6, four M16, JT - 15. Of course it is bleeding. Lee Obvious. Um, it's a 64 Meg bit Ddr3 1333, 666 megahertz.

Um, yeah. DDR DRAM So there you go. Pretty quick. So 64 Meg bits, 8 Meg samples.

effectively. So this thing's got 14 Meg samples total, so there must be a second one on the other side. It looks like they've got two Dc-to-dc converters here locally right next to the SRAM. Well, they're actually are pairing not only the memory, but powering the main acquisition ASIC as well.

And haha here is the ADC wedged in between the main acquisition ASIC there and the can which we'll take a look at under shortly and I recognize that puppy that's the Head 1511 that's from a company called Hit ID I We may have seen a he died ad seen in another tear down somewhere I'm sure we have. Anyway, this is a one gig sample per second as the banner spec on this scope says one gig sample per second ADC Now it's actually for a DC's in here and it interleaves those to actually get the one gig bit. one gig sample per second total rate. So of course they've only got the one in here.

they haven't got one per channel. So of course, as I showed in the review, when you actually turn on the second channel, the sample rate drops from one gig sample per second to 500 Meg samples per second. So yeah, that's the downside of this cuz you know these things aren't cheap. It would have been nice to put two in there and then you know you've got to double the stuff in the acquisition ASIC and everything else.

But yeah, they decided nope, we'll just use the one and half the bet. the bit rate, which is pretty annoying for a 200 megahertz bandwidth scope. As I mentioned, one gig sample second is fine. For one channel, it meets that minimum four time times, four times, four multiplier requirement.

but it doesn't. When you have that two 500 make samples per second. It effectively limits that usable bandwidth down to about 125 megahertz or so. Therefore, those played along at home with the trigger input just a bit of a pan down, they're not gonna go into that.

It's a yeah, knock yourself out. Sorry for the jittery pan here. It's not very professional is it? Hmm. Lots of unpopulated stuff down there.

Look at that. Tons of it. What's missing? Where's Wally and on the side of the can? Here, we're gonna have some trigger pick off from the analog channels to presumably look a 79o five voltage Regulator surface mount Tucked in here is that for the low noise analog front end cuz this thing is supposed to have a true 500 micro volts per division front end. but uh-huh Look at what they have out here in addition to this negative 5 volt regulator look, they've got some LP 38 78.

haha. it's suspected they might. so I Had a closer look at these. These are at low noise, adjustable dropout, low dropout voltage regulators.
so they're low noise Of course, designed to get the low noise performance for powering the circuitry on the analog front end. Nice. Well, they granted they've only got two of those for. well, maybe there's one per channel I'm not sure whether or not to power both channels and they're common in that case.

Um, you might get a bit of a little teensy bit of crosstalk between channels. Oh Boohoo if we take a look at the analog front end here, I Won't go into a huge amount of detail I'll post some high-res photos for those who want to look at it, but it's all passive around here, so maybe there's some active circuitry on the bottom? I'm not going to take this board out today. We: yeah, there could be something on the bottom. not sure, but it's all passive networks.

right? Up until here, we've got a couple of couple of trainees in there. By the looks, are they diodes? Can't see the silks or diodes? I Lead d 1 d 2 Yeah, they could be diodes instead of transistors. Yep, on a Q There we go. Got a transistor over here? that's Q So anyway, if we go up here, here's where all the actions happening.

This is A if we can get that number. trust me, it's an analog devices are a DAT 370 and that's a variable gain amplifier / a DC driver. So that's the puppy that's doing most of her a good lot of the game. plus the main line driver straight to the ADC which isn't far.

by the way, it's just got - it's just gonna pop across there. no big deal and that thing down in there I'm not sure what it is, it's got h 1k hash, five one three. so I'm not exactly sure what that puppies are doing, but apart from that, nothing special. I mean we've got a set for HCV you know, Five Nine five up here.

that's just doing some driving stuff. We've got a Cosmo solid state relay there, so that's not uncommon. You find those all the time in these front ends. Not sure what that puppy there is can't quite make up the part number on that I'm afraid.

So yeah, there's not too much else exciting there. Yeah, unfortunately, there's nothing in there that's telling me that they're doing anything special for the 500 microvolt. Front end got a couple of our NEC who realized of course I was wrong on the realise they've got three of those puppies under this bed. No expense any.

C's there. not one of us and he sees make pretty good realise yeah, but apart from that unless there's stuff on the bottom side. but yeah, as I said, I probably don't want to take the whole board out today. Unfortunately, you and right next to our ADC here.

no surprises for finding the ATF 4360 - 7. This is a VCO It's the PLL for to generate the ADC clock and you can see some inductors in there. Look at those puppies there we go fueling there, so hopefully they've got it there or the looper filter components around here and whatnot to enable that. So hopefully they've done that right unlike our Rygel who goofed up their loop filter values.
and yeah, they got jitter. So there was a jitter problem on the right guys, which they fixed by tweaking the. luckily for them by tweaking the coefficients inside this thing so they were able to fix it anyway. Um, so unless they, well, somebody can actually reverse engineer that and have a look at, but we don't know the exact value.

so I'm not gonna go in there and measure. let's assume it works well. Hello Sailor is the power supply. At first glance, this looks pretty well done, nice and clean and tidy.

but our huge bypass cap button Wow there's the first thing Leland Brand didn't fail. Anyway, with our bridge rectifier in there, we've got our wife our common mode choke in there. we've got ya. our suppression caps is that our mob in there as well? But look at all the silastic sticking down those output caps they really don't want those are flapping around in the breeze.

but now upped a couple of noise isolation slots there. I'm sure they've all got the reckless it rated components, another isolation slot down in there. they're doing the business that's not too shabby at all. And there's the primary side switch in nice-looking heatsink.

they're nice-looking enough copper strap over the main transformer main switching transformer there and that's pretty decent. Check out the little heat sinks. have a bit of attention to detail there. They probably don't want it to run too hot, but they probably didn't want to you know, put any major.

Probably didn't require any major heat sink work in there. But yeah, that is a neat little power supply. What are those output caps? I Leave Rubicon's I think they're I think they're Ruby Cons: they've done a decent brand. there won't though they don't a decent brand for the main input cap.

Oh what a bummer. But apart from that, that is a neato looking power supply siglent brand of. whether or not they actually did it themselves or farmed it out as is most common in the industry to do, we don't know. So there you have it.

That's a not a total look inside the new segment 1000 x-series but it's probably as far as I'm going to go today. I don't have to take out there. You rip off the tape for this and get the flat flexes out and take it all out. Yes, sorry I want this thing that's the only one in the country? Yeah.

I'm a scaredy cat Anyway, no I just couldn't be bothered. I'm actually running out of time. So yeah, we're not going to see a huge amount. You know for people who want to analyze the front ends and stuff like that? Well, I'm sure somebody else will do a more thorough teardown with more professional pictures once they get one.

No big deal, We know there's an extra memory chip on the back of there. that's probably an extra one of those on the back of there. Nothing too much doing it will support from that on the back side. I'm sure.
So yeah, don't get all upset that you didn't see the bottom side. There you go. It was just too easy to access this top side. And anyway, and that is quite a nice little design.

I Don't mind it at all, you know. Apart from the sample rate of course, which you know is an issue with this thing. it's a 200 megahertz bear with one gig sample per second not shared. Apart from that, you know they haven't got much horsepower.

I'm surprised the amount of the lack of horsepower they got to drive the screen. There's not much there at all and not much to it. And once again, sorry, but we don't know what's under there? I've you know that thermal adhesive is really tough to get off the height? Really stand a good chance of damaging that puppy if I Take it off shade and they don't have it down with a clip or anything like that. But yeah, that's a fairly simplistic design.

You can see how it's being built down to a cost although they haven't really skimped. you know, huge meant. Once again, they've got you know, stuff like this connector and that's so populated for no apparent reason. And yeah, it's not like you know.

Madman Months has got in there and ripped out you know, components until I stopped working and then put the last one back in and said ship it, you know? So yeah, they've done a reasonable amount of engineering it. By the way, if you don't know who made me a month since I recommend you Google it. that's in there for the old-timers All right, just for kicks. I've powered this thing up.

we're gonna look. flashing heartbeat light there. We've got one over here and we all get the flower on this puppy and take a look at the thermal profile of this. Here we go.

No surprises for finding a hot little regulator over there. so that's part of the switch mode. 36 odd degrees, you know, not much doing on the on the main video processor chip there. the main DSP not much doing up there.

32 degrees. No surprises for finding all the actions happening down here. That's the real hot spot there, which I'm probably now is at 50 degrees. no fan by the way.

I haven't got the fan turn on this so there's no air flow so this is not a fair. you know, real-world test it actually be probably you know, at least 10 degrees lower than this or something like that. and then around you can see all the board as well. Around the main FPGA there the acquisition FPGA The reason why the FPGA looks cool is because that's the emissivity of the the silver heatsink there.

So yeah, it's not a calibrated for that emissivity so it appears cooler than it actually is. If that was a black heatsink, we'd have black anodized. He'd think it would Does show up just fine. But anyway, a DC is the real hot puppy in there.

and the board surrounding that is, you know, a good 10 degrees cooler. So not quite sure what the FPGA up here is is that? you know I can keep my finger on there? So it's under 50. It's it's in the high four, it's in the mid to high 40s. maybe? So anyway, that's not too shabby at all.
Just did that for a bit of fun. Oh, hang on I've got that upside down. What enough? What a nob. What a wooly thing.

And check it out. The stickers back, no one will be none the wiser. Look at that beauty. Let's power this puppy up.

See she still works. By the way, that power supply might look decent inside, but bloody hell takes as I shown in the thing. It tastes like eight watts or something on standby. Absolutely ridiculous.

It's not like my stupid heartbeat Led things which is always running. Oh God I was just to do a heartbeat lid. Gotta be kidding me. Anyway, there we go.

it's booting up. Have my old homemade clock here from the 1980s. We know we're no chicken dinner so as always tear down high-res tear down photos available on my Eevblog command forum links and leave youtube comments and all that sort of jazz as always, if you liked it, please give it a big thumbs up. Rate Comment: Subscribe: Support me on Patreon and all that sort of stuff.

Follow me on Twitter Bla bla bla bla bla by the merch, you know? Yeah. Catch you next time you.

Avatar photo

By YTB

17 thoughts on “Eevblog #800 – siglent 1000x oscilloscope teardown”
  1. Avataaar/Circle Created with python_avatars Higgins Gracie says:

    Thanks and Recommended for all

  2. Avataaar/Circle Created with python_avatars saintpine says:

    Whats wrong with the choice of Lelon capacitors.
    It's a Chinese Oscilloscope producer using Chinese capacitors.

    Lelon Electronics Corp, producers only capacitors since 1976.
    Surely they are experts in the field.

  3. Avataaar/Circle Created with python_avatars cricardol says:

    8bit ADC??? Am I missing something here?

  4. Avataaar/Circle Created with python_avatars Gautam Kumar Goswami says:

    Between Siglent 1000X & Rigol DS1054Z, I think built quality (inside & outside) is better in Rigol DS1054Z. Like to know what other viewers think on this aspect.

  5. Avataaar/Circle Created with python_avatars Xin Tian says:

    the board looks hand solder to me.

  6. Avataaar/Circle Created with python_avatars tkarlmann says:

    You get a Thumbs Up for the Proper Reference to Muntz! With a photo yet! Dave, you surprised me with that one; I cannot believe you knew of him!

  7. Avataaar/Circle Created with python_avatars Ernst Stavro Blofeld says:

    What is the best oscilloscope in the world?

  8. Avataaar/Circle Created with python_avatars FrankenPC says:

    I love how you peeled off the tamper sticker on a video Siglent will see. Classic Dave ๐Ÿ˜€

  9. Avataaar/Circle Created with python_avatars FrankenPC says:

    +1000 Dave for the Errol Flynn reference.

  10. Avataaar/Circle Created with python_avatars Dorota says:

    Thanks for the schematics when discuss parts eg. Micron RAM. Always too lazy to check datasheets myself…

  11. Avataaar/Circle Created with python_avatars Maurice Spelman says:

    Is there any good hobby grade waveform generators under $400?

  12. Avataaar/Circle Created with python_avatars Korytnackaseven English says:

    What is that black stylus or something that Dave uses for pointing at things?

  13. Avataaar/Circle Created with python_avatars maultrommel says:

    the pcb layout is pretty close to the DG2xxx series, now somebody hack it like the DS1,2,4k ๐Ÿ˜‰ /cybernet

  14. Avataaar/Circle Created with python_avatars BaZzZa says:

    Lovely work Dave

  15. Avataaar/Circle Created with python_avatars Joanna Hammond says:

    I noticed that some of that gunk was covering the vent holes on the caps. Could that have an effect on long term safety if they ever started to fail?

  16. Avataaar/Circle Created with python_avatars Joanna Hammond says:

    Really wondering what is the output on that mini connector on the board. Any chance of probing it???? ๐Ÿ˜€

  17. Avataaar/Circle Created with python_avatars TrainMasterMan says:

    Tear down of the bomb… I mean clock…

Leave a Reply

Your email address will not be published. Required fields are marked *