you gotta have an aluminium handle to lift this kinda processing grunt. 4 FPGAs. Whoa! ;) No seriously, the mechanical engineering (as well as their electronics) improved a lot since I first bought that PSU with a rusty metal frame inside a slightly brittle plastic case about 15 years ago! Most likely its demanded by their OEM customers like LeCroy thats giving us better quality!
The vent holes in the case and shielding would align with each other if all of the other holes were lined up too, you just have that piece turned around 180 degrees.
I like to use CH1 and CH3 way more than CH1 and CH2, because then the sockets are further apart, and it's easier to fiddle around with the cables. The sample rate is obviously another factor, and the board layout, too, but it is simply more practical to use sockets that are not right next to each other.
Great teardown video Dave! I think there's a much simpler and sensible reason why the channels are not organized 1324 iso the usual way... Though trace routing concerns are indeed a thing that would have to be considered, the most important thing is easiness of configuration in case of a dual channel option. Sure enough no one wants to be messing with varying signal routing depending on the option, and even less using 2 ADCs on a dual channel scope... The only sensible choice is to organize the signals coming out of the front end 1&2 to the first ADC, and 3&4 to the 2nd ADC. As for the additional FPGA with "mystery" use.... Wouldn't it be for Sync? I don't have much experience with digital scope architectures, but at least on the analog ones, Sync usually takes quite an important chunk of the circuitry... And maybe it was too much for them to accomodate into the main, larger FPGA... Or it can be dedicated to the logic analyzer section.
I could be way off here, but at first glance, the board just looks like it repeats a known good one channel configuration that was made to work for two channels. Then, there are two of these side by side for four probes. The five BGA RAM IC's in the upper right might be a way to use cheaper older RAM and get easy cheap storage to feed the main CPU with stored data chunks
yep! they have gone all the way 1980s-SONY construction! pretty cool! it seems the air duct is there to prevent the hot air of the PSU mix with air flowing over the main board.
DCDC chips are never really 'chunky' because of mosfets, those usually best sink their heat into a copper plane and are better off left flat(some do have exposed pads at top for thermal as well though). Integrated inductors are usually the reason for the added height.
Hats off to Dave on maintaining the physique, even the reflection on the oscilloscope screen shows his nerves popping out on his forearm! Inspiration to many hacker couch potatoes out there.
Siglent has come a long way in terms of quality since its beginnings and it shows. I hope they'll still make scopes for the low end /hobbyists market, though.
This is low end, at least in cost. It's $1.2k. That's not a lot more than the basic scopes were a decade ago. I still have my Rigol DS1054, half the price but good enough for my use. If I were buying now however, this is a lot better specified and looks very well made.
@@JohnDoe-wb7ng You mean the 12 bits version ? If there is a 10 bits version at $1.2k, then it's a pretty good deal and still in the reach of advanced hobbyists.
@@lolilollolilol7773 Correct. I am ok with my rigol but the 10b version for 1.2k would be a nice upgrade of my lab. I was just reacting to Mike Willis saying that this very version is not $1.2k device.
Looks like they paid quite alot of attention to detail, very nice. Not sure if I like the cramped area on the RH side with all the buttons and knobs. However...I could see how that could make for faster selection, not having to move your finger/thumb around alot. Very curious to see how it performs given its price point. I'm really impressed at the construction of the scope.
@@JFirn86Q 👋how are you today and I hope this message meets you in good health? I believe in the saying that you can't meet better people if you don't meet strangers. I admire your profile very much and it will be great to have you as a friend please can we be friend?. Thank you and remain blessed 🙏
Speaking of having Ch1 and Ch2 on the same ADC: Ch1 and Ch2 are special channels comparing to the others. They are designated for X and Y in the X-Y mode. For example, with SDS2000X Plus you can only use Ch1 as X, and Ch2 as Y. I would prefer to have full sampling rate in X-Y mode! I don't know how hard it would be to make X and Y selectable channels.
The more elegant way of channel arrangement (instead of trying to route the traces around each other) would be to just have the channels numbered 1,3,2,4
How are the remote interfaces (USB/LAN)? I tire of taking screen shots to document circuit behavior and have shifted almost completely to instruments with complete and performant remote interfaces, greatly simplifying documentation (screen grabs), data capture (including live) and analysis (e.g., Matlab). At my current gig (an RF semiconductor startup), engineers were making frequent trips into the well-equipped lab, then back to their desks, then back to the lab. I put a Digilent Analog Discovery 2 (DAD2) on my desk-bench, and for basic stuff was instantly running productivity rings around them, occasionally using the lab only for higher-bandwidth stuff. The DAD2 paid for itself just in recovering the "out of chair" time. And without a stack of equipment taking up room on the small bench next to my desk!
I'm really not a fan of replacing the warranty void sticker before resale. It's quite disingenuous to the buyer. Myself, I really want to know if something has been opened or repaired before if I buy second hand. I've seen the most horrible repair attempts done to stuff, half the screws missing and it may even have some ESD-damage, if it was handled poorly. Especially if I were to spent the big bucks to get one of these non-entry-level scopes.
It's not nice if some actual repair work was done to it, but if it was just opened for curiosity, like this one, no big deal. Although I would suggest just to remove the broken warranty sticker. A missing warranty sticker is better than a broken one.
To put it into perspective, if you add up the price of just the Spartan 7, Zynq, Artix 7, and the two ADCs, it's already more than the cost of the entire scope itself. Those ADCs cost $1,500 USD _each_ My guess is that the Spartan 7 handles the logic analyzer input capture, the Artix 7 handles the ADCs and any math on the data, and the Zynq is used to tie it all together and run the main OS.
8:00 that looks like a Samxon KM-series cap. And yes, MeanWell seems to be switching from exclusively using Japanese caps to allowing good quality Chinse caps in their products as well. I have come across Lelon and a brand I forget in more recent batches of Mean Well PSUs.
I always remember the first time I saw a modern scope and how amazing I found it. The fact that it was only a couple of inches deep as apposed to all the old analog scopes that were always about 18 inches deep.
Perfect timing. I've been on the fence about getting one. $1400 is a lot of money for a scope, but this seems to be so good that I'll realistically never need or even want to upgrade. What do you guys think? Input welcome
i know it isnt 1400 bucks but so wish it was... so much because then i could actually afford to buy one! (rather than the rigol hdo 1000, with its unfinished bugs and lack of mixed signal / awg). man would i love to have this as a purchase option, if it were not triple (maybe double)
Lattice has shifted toward serving the low-end/medium size FPGAs, while Xlilinx/Altera has mostly abandoned developing new products for the "does not need a heatsink" market.
I spent 2k around Xmas st Canada's major siglent reseller and they treated me like a child and missed out on a 2500 dollar purchase at least. I sold my siglent scope and kept the psu as it was an Xmas gift from the wife.
Their airflow is a bit drunk indeed. Zynq is running operating system for sure, they are all the rage with the ZyBo Digilent Boards. Spartan is for the sampling of the digital stuff. The function gen seems to have its own processor. If they use the Lattice as the over scope, my guess would be glue logic, from status bits to buttons to pretty much everything.
Dave, I feel you, I also hate when that happens 0:05 we all can relate, the hard knock life of a social media techinfluencer, who gets swallowed in gifted 12 bit scopes.. Keep pushing thrue, see ya on the other side.
Hm... no heatsink compound or did Dave wipe it off before putting the exposed parts on camera? It was interesting to see the mix of makers for the FPGA devices.
On directing ch1 and ch2 to different chips... is there a similar issue just swapping the trace or wire from the bnc for ch2 and ch 3? Other than changing what people have become accustomed to. I don't usually stress bandwidth anyway, so I use 1 and 2 first.
So where is the justification for the Siglent being so much more expensive than the Rigol? Any reason why anyone should buy the Siglent over the Rigol with the same specs?
Interesting so many FPGA's. I would have thought the giant FPGA would have been able to do everything? Also interesting that they'd use the obsolete ADC12D1000 from National.
I can see a few potential reasons. Does the big ARTIX have the capability to do everything all in one? Probably. Would it be easy to design? Almost certainly not. You wouldn't want Linux or your logic analyzer sharing access to your timing-sensitive acquisition memory controller, and the ARTIX might not have a second or third memory hard-core. As for the little MachXO controlling the display, my guess is the ARTIX doesn't support the right I/O voltage standard? Newer FPGAs like ARTIX-7 don't support 3.3V and 2.5V I/O (which are commonly used for display panel interfaces) on the high-performance type I/O banks.
@@asm_nop its more likely that they used their already stable working system modules rather than develop a new one, especially in this price region so the MachXO was a no brainer. The Zync is the obvious platform for linux application processor if you already using the Xilinx ecosystem and the Artix has the necessary power to not only do the data acquisition but also processing like FFT… The Spartan is ten used as Dave suspected for the logic analyzer, external triggering, maybe the arb wavegen stuff. All of these parts may have come out of other products like spectrum analyzers, signal gererators and such they're doing as well, probably on the same FPGAs.
Great teardown. Hello dave if you can help me. I have wavejet334 its like bootlop , posted on eevblog forum but theres no one reply. Thanks. Please reply me 😊
Odd to have a Zynq, Artix-7, Spartan-7 and that Lattice in there, you'd think a bigger Zynq could have done it all. But who knows, with how illogical FPGA pricing is, it may have been cheaper to go that route..
Most probably crosstalk, yes, plus other losses, like on-state resistance. I have a Sony ICF-7600DA radio, in which a 4066 (or similar analog MUX) does the band selection. That radio is pretty mediocre in SW sensitivity, and I presume that is caused by the crosstalk and other losses in the MUX IC at those 10-20MHz frequencies. It performs gradually worse at upper SW bands.
Void your warranty, void it all to hell and beyond! Ha. A sign of our times, you don't see diecast alloy parts anymore while they were pretty mainstream in pro gear a few decades ago... Sigh. These days ain't coming back, I guess. One of the reasons why I love tinkering with vintage gear! I love the separate wavegen board. Modularity and separation of concerns... you know what's what rather than having to figure out the jumbotron PCB. 8:28 Facebook? LOL Alignment pin, nice one! The thermal design is really interesting too. I don't get the idea of criss-crossing the input channels. Enjoyed the teardown big time :)
On way they could label the channels to indicate that they share an ADC is to label them Ch 1-A, Ch-1 b, Ch2-A, Ch2-b. But the marketing wankers would likely object since they would say that it makes sound like a 2 channel scope.
Really, Delta fans are the *only* fans I have ever seen disintegrate in normal operation. The bit were the fan blades came off, it kept spinning and then proceeded to spew brass shavings all over the inside of the server was particularly spectacular. We had at least 7 Delta fans disintegrate, and in the end the server vendor just replaced them lock stock and barrel because another on the ground the bearings up would cost more. The repair bill was in the thousands of dollars. Pictures can be found on servethehome forums.
@@jonathanbuzzard1376 Maybe a bad batch. Delta fans are really excellent. At work, we had an ancient HP Pentium II server that was running from 1998-2008 in 24/7, both the PSU and CPU fans were Delta, and had no problems when the server got retired. Also at work, CCTV DVR running from 2013-2022 in 24/7, it had Delta fans as well, but smaller ones, so higher RPM. They were a bit noisy already when the DVR got pulled out of service, but they still were running (also the two WD hard drives were perfect after almost 10 years of continous use, quite remarkable).
You will never have a 4 channel scope with 1:3 and 2:4 sharing channels simply because 1:2 and 3:4 has been used for so long it would just be confusing
Re: Why aren't channel 1 and channel 2 on different dual ADC chips? The obvious way to do it would surely be not to have them in order on the front panel. Just label them "Channel 1", "Channel 3", "Channel 2", "Channel 4" on the front panel and change the software a bit and you're done with still the same tidy track layout.
It would be even more retarded than having channels 1 and 3 as mains. If people are really so out with this, then most logical way would be having channel 1.5 and 2.5 as "half" channels, so scope channels would be 1, 1.5, 2 and 2.5. Don't see any reasons to do so, manufacturer could just print graphical design print on front panel which would suggest that channels 2 and 4 are just subchannels of 1 and 3.
@@darausus8315 First part: silly because you can probably still use channel 2 as full channel if you don't use channel 1 etc. Second part: fair enough.
No i am doing the same as well. The only fan i am not changing is PSU fans on PC's, but they are Seasonic Titanium with a Hybrid mode and they rarely turn on.
i was just thinking about this adc channel length matching: is this not just a complete waste of time as length of the probe will play a role and we know that a few mm will not make any difference?
More "wow"s per minute than an Owen Wilson movie. I think I had a boom box back in the day with a split handle into brackets like that, but I can't be certain. At my age, I don't trust my memory for shit.
I had to stop the video at the metal handle........and raise a glass to Siglent. Absolutely topnotch.
That handle really ties the room together.
you gotta have an aluminium handle to lift this kinda processing grunt. 4 FPGAs. Whoa! ;) No seriously, the mechanical engineering (as well as their electronics) improved a lot since I first bought that PSU with a rusty metal frame inside a slightly brittle plastic case about 15 years ago! Most likely its demanded by their OEM customers like LeCroy thats giving us better quality!
even rohde and schwartz used a plastic handle
I suspect the Lattice is a serialiser for the LVDS LCD interface
Lattice is cheaper if considering the cost per LVDS pair, Xilinx is more complicate and pricey.
The vent holes in the case and shielding would align with each other if all of the other holes were lined up too, you just have that piece turned around 180 degrees.
Yeah, I was screaming at the screen. Can't believe Dave couldn't see it.
THIS is the yolo stuff finally arriving from the Siglent boss. I love it!
Are you referring to the CEO dude saying Yolo frequently?
haha I'm still wondering why he said that.
@@PiitaaDerbez Yes! 😁
@@Indiskret1 I'm guessing maybe it's his equivalent of "um". In Japanese they say "eto" and "ano", which are the equivalents.
@@Indiskret1 although I took a semester of Chinese and never heard that 🤔
You had that back shield the wrong way around at 12:31. You could see the fan through the slats at 4:20.
D'oh, yeah, that was dumb.
@@EEVblog no worries, mate. Spotted that from experience. Been there done that.
Wow, that looks neat! None of the old time Siglent oxide/rust coating :D
Those were the days!
Is there a video of this? I wanna see haha.
I like to use CH1 and CH3 way more than CH1 and CH2, because then the sockets are further apart, and it's easier to fiddle around with the cables. The sample rate is obviously another factor, and the board layout, too, but it is simply more practical to use sockets that are not right next to each other.
Not gonna lie, the alloy handle that connects to the metal frame definitely sold me!
Это силумин
I love how the handle is made of 2 identical parts
Almost. One is threaded.
@@therealjammit alright but they do come from the same mold.
Yeah, that's great little detail :D
Great teardown video Dave!
I think there's a much simpler and sensible reason why the channels are not organized 1324 iso the usual way...
Though trace routing concerns are indeed a thing that would have to be considered, the most important thing is easiness of configuration in case of a dual channel option.
Sure enough no one wants to be messing with varying signal routing depending on the option, and even less using 2 ADCs on a dual channel scope... The only sensible choice is to organize the signals coming out of the front end 1&2 to the first ADC, and 3&4 to the 2nd ADC.
As for the additional FPGA with "mystery" use.... Wouldn't it be for Sync?
I don't have much experience with digital scope architectures, but at least on the analog ones, Sync usually takes quite an important chunk of the circuitry... And maybe it was too much for them to accomodate into the main, larger FPGA...
Or it can be dedicated to the logic analyzer section.
Yep, forgot that.
I could be way off here, but at first glance, the board just looks like it repeats a known good one channel configuration that was made to work for two channels. Then, there are two of these side by side for four probes. The five BGA RAM IC's in the upper right might be a way to use cheaper older RAM and get easy cheap storage to feed the main CPU with stored data chunks
That's a nice handle. I like it. Hope it starts a trend.
yep! they have gone all the way 1980s-SONY construction! pretty cool! it seems the air duct is there to prevent the hot air of the PSU mix with air flowing over the main board.
Thou shall follow Ghostbusters cannon and not cross the streams.
DCDC chips are never really 'chunky' because of mosfets, those usually best sink their heat into a copper plane and are better off left flat(some do have exposed pads at top for thermal as well though). Integrated inductors are usually the reason for the added height.
This one is chunky as it has a built in inductor as well, the datasheet shown briefly was for the wrong part
I’m buying it for the handle design. Brilliant 😊
Hats off to Dave on maintaining the physique, even the reflection on the oscilloscope screen shows his nerves popping out on his forearm! Inspiration to many hacker couch potatoes out there.
Any time their sending you fancy kit starts getting on your nerves, you can just send them my way. No really, it's no problem!
Siglent has come a long way in terms of quality since its beginnings and it shows. I hope they'll still make scopes for the low end /hobbyists market, though.
This is low end, at least in cost. It's $1.2k. That's not a lot more than the basic scopes were a decade ago. I still have my Rigol DS1054, half the price but good enough for my use. If I were buying now however, this is a lot better specified and looks very well made.
@@mikewillis1592 This is the HD version. It starts at 3.5k
@@JohnDoe-wb7ng You mean the 12 bits version ? If there is a 10 bits version at $1.2k, then it's a pretty good deal and still in the reach of advanced hobbyists.
@@lolilollolilol7773 Correct. I am ok with my rigol but the 10b version for 1.2k would be a nice upgrade of my lab. I was just reacting to Mike Willis saying that this very version is not $1.2k device.
@@JohnDoe-wb7ng That looks like a damn good scope for the price, for sure !
I think that metal plate @13:00 is using a venturi effect to suck air past the power supply from the side vent.
Yes, the fan is press fit (bot no seal) on the back case, so the suction will draw air through the thin plate from the outside over the PSU.
Looks like they paid quite alot of attention to detail, very nice.
Not sure if I like the cramped area on the RH side with all the buttons and knobs. However...I could see how that could make for faster selection, not having to move your finger/thumb around alot.
Very curious to see how it performs given its price point. I'm really impressed at the construction of the scope.
I'm impressed Siglent, wow. Did not expect that level.
Hello Jonathan, Can I ask you a question?
@@danm18835 Just ask, no need for permission on youtube.
@@JFirn86Q 👋how are you today and I hope this message meets you in good health? I believe in the saying that you can't meet better people if you don't meet strangers. I admire your profile very much and it will be great to have you as a friend please can we be friend?. Thank you and remain blessed 🙏
@@JFirn86Q Where are you from?
@@danm18835 USA, you?
Thats quite a tasty bit of gear! Thanks for open it.
The memory chips with SEC = Samsung Electronics Corporation.
Speaking of having Ch1 and Ch2 on the same ADC:
Ch1 and Ch2 are special channels comparing to the others. They are designated for X and Y in the X-Y mode. For example, with SDS2000X Plus you can only use Ch1 as X, and Ch2 as Y. I would prefer to have full sampling rate in X-Y mode!
I don't know how hard it would be to make X and Y selectable channels.
The more elegant way of channel arrangement (instead of trying to route the traces around each other) would be to just have the channels numbered 1,3,2,4
How are the remote interfaces (USB/LAN)? I tire of taking screen shots to document circuit behavior and have shifted almost completely to instruments with complete and performant remote interfaces, greatly simplifying documentation (screen grabs), data capture (including live) and analysis (e.g., Matlab).
At my current gig (an RF semiconductor startup), engineers were making frequent trips into the well-equipped lab, then back to their desks, then back to the lab. I put a Digilent Analog Discovery 2 (DAD2) on my desk-bench, and for basic stuff was instantly running productivity rings around them, occasionally using the lab only for higher-bandwidth stuff. The DAD2 paid for itself just in recovering the "out of chair" time. And without a stack of equipment taking up room on the small bench next to my desk!
Haven't even used it yet.
If it's anything like the 2000X Plus series I have on my bench, the LAN interface is outstanding. Full speed low latency web based GUI!
REALLY COOL BUILD. METAL HANDLE N ALL. THANKS DAVE GREAT VIDEO
The SDS2000X Plus has a 10 bit mode, it is under the Acquire menu, it reduces the memory depth when in use. I have a SDS2504X Plus.
Not a true 10 bit ADC though.
I'm really not a fan of replacing the warranty void sticker before resale. It's quite disingenuous to the buyer.
Myself, I really want to know if something has been opened or repaired before if I buy second hand.
I've seen the most horrible repair attempts done to stuff, half the screws missing and it may even have some ESD-damage, if it was handled poorly.
Especially if I were to spent the big bucks to get one of these non-entry-level scopes.
Yep, pretty poor play to actively suggest that sellers should lie to their would-be customers.
It's not nice if some actual repair work was done to it, but if it was just opened for curiosity, like this one, no big deal. Although I would suggest just to remove the broken warranty sticker. A missing warranty sticker is better than a broken one.
The blank board by the PSU is for a battery option maybe?
To put it into perspective, if you add up the price of just the Spartan 7, Zynq, Artix 7, and the two ADCs, it's already more than the cost of the entire scope itself. Those ADCs cost $1,500 USD _each_
My guess is that the Spartan 7 handles the logic analyzer input capture, the Artix 7 handles the ADCs and any math on the data, and the Zynq is used to tie it all together and run the main OS.
8:00 that looks like a Samxon KM-series cap. And yes, MeanWell seems to be switching from exclusively using Japanese caps to allowing good quality Chinse caps in their products as well. I have come across Lelon and a brand I forget in more recent batches of Mean Well PSUs.
I always remember the first time I saw a modern scope and how amazing I found it. The fact that it was only a couple of inches deep as apposed to all the old analog scopes that were always about 18 inches deep.
Wave gen on the back side? What an ingenious solution. They should have put some knobs on the back as well.
And display as well.
It seems that the days of the Siglent infamous rust are long gone :)
Perfect timing. I've been on the fence about getting one. $1400 is a lot of money for a scope, but this seems to be so good that I'll realistically never need or even want to upgrade. What do you guys think? Input welcome
if you are a hobbyist, for sure it's an excellent instrument.
$1400? I believe you were looking at the plus models. These HD 12 bit models look to start at about $3200.
The SDS200X Plus are still great though.
@@Woodturner88 yes! I was looking at the plus, not the HD. my bad, thanks for the correction
i know it isnt 1400 bucks but so wish it was... so much because then i could actually afford to buy one! (rather than the rigol hdo 1000, with its unfinished bugs and lack of mixed signal / awg). man would i love to have this as a purchase option, if it were not triple (maybe double)
So is this just a Teledyne Lecroy Wavesurfer super dumbed down?
Nice machine, nicely built.
AWG output at the front would have been better though.
Hi, Dave, which 12bits one do you prefer more? Siglent SDS2000X HD or Rigol HDO4000? or just MSO5000 & SDS2000XP?
Lattice has shifted toward serving the low-end/medium size FPGAs, while Xlilinx/Altera has mostly abandoned developing new products for the "does not need a heatsink" market.
I spent 2k around Xmas st Canada's major siglent reseller and they treated me like a child and missed out on a 2500 dollar purchase at least. I sold my siglent scope and kept the psu as it was an Xmas gift from the wife.
Their airflow is a bit drunk indeed. Zynq is running operating system for sure, they are all the rage with the ZyBo Digilent Boards. Spartan is for the sampling of the digital stuff.
The function gen seems to have its own processor. If they use the Lattice as the over scope, my guess would be glue logic, from status bits to buttons to pretty much everything.
You'd think they would have just mounted the fan on the back cover, then removed the fan extension and duct.
Those National logo's on the ADC's.
Probably the same stuff that goes into Tektronix?
Dave, I feel you, I also hate when that happens 0:05 we all can relate, the hard knock life of a social media techinfluencer, who gets swallowed in gifted 12 bit scopes..
Keep pushing thrue, see ya on the other side.
Hm... no heatsink compound or did Dave wipe it off before putting the exposed parts on camera? It was interesting to see the mix of makers for the FPGA devices.
He wiped it off, you can see the residue.
On directing ch1 and ch2 to different chips... is there a similar issue just swapping the trace or wire from the bnc for ch2 and ch 3?
Other than changing what people have become accustomed to. I don't usually stress bandwidth anyway, so I use 1 and 2 first.
So where is the justification for the Siglent being so much more expensive than the Rigol? Any reason why anyone should buy the Siglent over the Rigol with the same specs?
I didn't watch the video for that wou at the beginning of the video but because i like the electronics teardown 😛
The wavegen, is effectively a USB device, Delta is a taiwanese company.
In your Professional opinion, what would be the Cheapest and most accurate Oscilloscope to buy 2023?
That's nuts!
I now want to modify my digital scope by swapping out the handle to an all aluminum cast handle. Lol😂
1. Remove screen protecting film
2. Put it face down and scrape it against the table thoroughly
Sounds like a plan! 🤦♂️🤦♂️🤦♂️
My thoughts as well, hillarious..looked quite a lot on the gren surface..hmm would probably not be a problem..
That AWG is a pure USB2.0 device
Interesting so many FPGA's. I would have thought the giant FPGA would have been able to do everything? Also interesting that they'd use the obsolete ADC12D1000 from National.
I can see a few potential reasons. Does the big ARTIX have the capability to do everything all in one? Probably. Would it be easy to design? Almost certainly not. You wouldn't want Linux or your logic analyzer sharing access to your timing-sensitive acquisition memory controller, and the ARTIX might not have a second or third memory hard-core.
As for the little MachXO controlling the display, my guess is the ARTIX doesn't support the right I/O voltage standard? Newer FPGAs like ARTIX-7 don't support 3.3V and 2.5V I/O (which are commonly used for display panel interfaces) on the high-performance type I/O banks.
@@asm_nop its more likely that they used their already stable working system modules rather than develop a new one, especially in this price region so the MachXO was a no brainer. The Zync is the obvious platform for linux application processor if you already using the Xilinx ecosystem and the Artix has the necessary power to not only do the data acquisition but also processing like FFT… The Spartan is ten used as Dave suspected for the logic analyzer, external triggering, maybe the arb wavegen stuff. All of these parts may have come out of other products like spectrum analyzers, signal gererators and such they're doing as well, probably on the same FPGAs.
They made this with the idea that it would be opened.
Great teardown. Hello dave if you can help me. I have wavejet334 its like bootlop , posted on eevblog forum but theres no one reply. Thanks. Please reply me 😊
Odd to have a Zynq, Artix-7, Spartan-7 and that Lattice in there, you'd think a bigger Zynq could have done it all. But who knows, with how illogical FPGA pricing is, it may have been cheaper to go that route..
The Rigol looks awesome though
straight to the pool room with that one!..i havent seen quality like that in..30 years! :P
Looks like a SAMXON cap in that meanwell
I've been really happy with mine. Except the high pitch SMPS whine in standby.
What? It produces wine during standby? How? You have to feed it with grapes?
@@mrnmrn1 haha Whine! :)
Good teardown video! No review?
Can someone explain, why have low current switching relays not been superseded by silicone, the 4066 has been out forever, is it a crosstalk issue?
Most probably crosstalk, yes, plus other losses, like on-state resistance. I have a Sony ICF-7600DA radio, in which a 4066 (or similar analog MUX) does the band selection. That radio is pretty mediocre in SW sensitivity, and I presume that is caused by the crosstalk and other losses in the MUX IC at those 10-20MHz frequencies. It performs gradually worse at upper SW bands.
Dave! You will have to do some chip muncing to fimd out what the Lattice chip actually does!!!
Just realized that this scope look just like Tek series 3 scope
Where is the power supply fuse?
Void your warranty, void it all to hell and beyond! Ha.
A sign of our times, you don't see diecast alloy parts anymore while they were pretty mainstream in pro gear a few decades ago... Sigh. These days ain't coming back, I guess. One of the reasons why I love tinkering with vintage gear!
I love the separate wavegen board. Modularity and separation of concerns... you know what's what rather than having to figure out the jumbotron PCB.
8:28 Facebook? LOL
Alignment pin, nice one! The thermal design is really interesting too. I don't get the idea of criss-crossing the input channels. Enjoyed the teardown big time :)
The criss-cross is for user psychology. People prefer the use CH1 and CH2, not CH1 and CH3.
08:25 Ferrite Bead, sometimes.
Could it be ARM's wannabe JTAG, SWD, present in that connector named JTAG? Does match the pincount (SWDIO / CLK / Ground)
Whats total ram?
Almost looks like the wavegen is usb by the pinout.
Forget about 12-bit resolution, buy it for its magnificent handle 😂
Alloy handles are gonna make or break the deal for me now....cheers.
lol, good plastic is strong enough
4 FPGAs in one device ?? WOW :) That's why it costs more than Rigol :)
But the real question is... can you sell it used on eBay?
They could just label them 1, 3, 2, 4 - but what's the fun in that?
So many FPGAs - nice ;-)
I saw this one coming 😎
On way they could label the channels to indicate that they share an ADC is to label them Ch 1-A, Ch-1 b, Ch2-A, Ch2-b. But the marketing wankers would likely object since they would say that it makes sound like a 2 channel scope.
This wave gen looks like generic signal generator that you can buy for sds 1204xe oscilloscope. It even connects via USB
What’s the Teledyne affiliation?
Siglent is the OEM for some Teledyne LeCroy scopes
Hallelujah!
Capacitor in power supply is a cheap Chinese made capacitor. Sam Young or Sham Wha or Sun Xo brand.
Delta fans are the best out there. I like the triple blades. A little loud., but excellent cooling.
Really, Delta fans are the *only* fans I have ever seen disintegrate in normal operation. The bit were the fan blades came off, it kept spinning and then proceeded to spew brass shavings all over the inside of the server was particularly spectacular. We had at least 7 Delta fans disintegrate, and in the end the server vendor just replaced them lock stock and barrel because another on the ground the bearings up would cost more. The repair bill was in the thousands of dollars. Pictures can be found on servethehome forums.
@@jonathanbuzzard1376 perhaps he unknowingly bought a bunch of fake copies ?
@@lolilollolilol7773 we are talking a major tier 1 server vendor here, fake fans is not the explanation
@@jonathanbuzzard1376 Maybe a bad batch. Delta fans are really excellent. At work, we had an ancient HP Pentium II server that was running from 1998-2008 in 24/7, both the PSU and CPU fans were Delta, and had no problems when the server got retired. Also at work, CCTV DVR running from 2013-2022 in 24/7, it had Delta fans as well, but smaller ones, so higher RPM. They were a bit noisy already when the DVR got pulled out of service, but they still were running (also the two WD hard drives were perfect after almost 10 years of continous use, quite remarkable).
You will never have a 4 channel scope with 1:3 and 2:4 sharing channels simply because 1:2 and 3:4 has been used for so long it would just be confusing
Re: Why aren't channel 1 and channel 2 on different dual ADC chips? The obvious way to do it would surely be not to have them in order on the front panel. Just label them "Channel 1", "Channel 3", "Channel 2", "Channel 4" on the front panel and change the software a bit and you're done with still the same tidy track layout.
It would be even more retarded than having channels 1 and 3 as mains. If people are really so out with this, then most logical way would be having channel 1.5 and 2.5 as "half" channels, so scope channels would be 1, 1.5, 2 and 2.5. Don't see any reasons to do so, manufacturer could just print graphical design print on front panel which would suggest that channels 2 and 4 are just subchannels of 1 and 3.
@@darausus8315 First part: silly because you can probably still use channel 2 as full channel if you don't use channel 1 etc. Second part: fair enough.
Honestly, this scope looks ridiculously overengineered. Plus, I wonder what's the actual ENOB for the 12-bit version.
18:00 SEC - Samsung electronics corporation
Delta Electronics is a Taiwanese electronics manufacturing company. Not China.
Meehh. You forgot the Wavegen. Or did I fall asleep?
Wait a sec..... There is no more Silent rust signature? Yoloscope become better and better!
I wonder if I'm the only person who changes every fan in every piece of gear with noctua.
No i am doing the same as well. The only fan i am not changing is PSU fans on PC's, but they are Seasonic Titanium with a Hybrid mode and they rarely turn on.
i was just thinking about this adc channel length matching: is this not just a complete waste of time as length of the probe will play a role and we know that a few mm will not make any difference?
More "wow"s per minute than an Owen Wilson movie.
I think I had a boom box back in the day with a split handle into brackets like that, but I can't be certain. At my age, I don't trust my memory for shit.
that ain't Siglent, there is no RUST!
A USB like scope with a phone built in.
😢not akurat osciloscope
Remov the SPO film nau !!!
Hello, Can I ask you a question?