The cutouts around the edge of that mesh appear to match up with the fan mounts and the power connector, almost like it was supposed to go over there instead. And in case you haven't heard, you might be interested to hear that "Warranty Void If removed" stickers have been ruled illegal over here in the US .
Yup, its the same here. Companies can put whatever stickers they want on there, they are not illegal(even if media might tell you that, assuming you not actually gonna look it up), but it doesn't actually void the warranty when you break them. Just like article 13 of the EU doesn't ban memes, they didn't ban warranty void stickers either. Media outlets just tend to exaggerate and assume that people don't actually look at the rulings or laws themselfs, and that usually works. They have to prove that you caused it to fail. Opening a product up doesn't automatically break it, therefore the sticker is completly useless. It'd be ruled that way in most places probably.
People and Dave. The sticker are illegal if the manufacturer made it restriction to third party part (or small business), cost to repair device or no waiver from FTC. It is part of US Federal law since there have been unfairly history in American. Where the primary company had hurt both consumers and small business a lot. So, to have "warranty void if seal" broken sticker, it must not costing to repair it or providing the part. That is Magnuson-Moss Warranty Act, a federal law.
@@TheTyrost either way, if I see that message coming out of the serial port, I would expect to send a character over the serial port, not an attached keyboard. Maybe I would try a physical keyboard afterwards, but not first off the bat. 😆 That one made me facepalm a little too.
Even if he could get TX/RX to work - that would only get him as far as the u-boot bootloader. It'd be much easier just to read the linux image from flash, then "binwalk -e" on it and go from there.
Rx is probably working, but using different logic levels than Dave's USB serial adapter. If the USB serial adapter's default logic level screws with the receive, the CPU may not be willing to continue booting.
"Press key to stop boot" will be looking for serial chars, not usb keyboard. Hardware links will be mostly so the software can know board rev and deal with any changes over time with a single Fw version
My best guess is the resistors in hw-id is for the software to read the board revision if something changed etc - not to lock down the firmware. I still think you can purchase software option codes and enter and it would unlock that way.
“Déjà vu review” = Dave made this video and simply “forgot to publish” then a few weeks later he started his research and recording all over again and started having some flashbacks... checked his hard drive and the video was all done ready to go!! Hahahaa one of these moments... =]
I wonder if a hack for these will ever be published. After-all they are more expensive so not as many people can get their hands on these to start hacking.
That hardware version is probably just that. The hardware version, so they can run the same firmware images across multiple scopes, or at the very least the same firmware for different PCB versions of the same model.
me too, the feeling of voiding the warranty as soon as you get the scope home while having a cup of coffee while opening the box ahhhhhhhhhhh heaven BUT THERE IS A BETTER FEELING Having the manufacturer suspect that you've unlocked your scope then they challenge you on it WHILE YOU ARE ABOUT TO BUY ANOTHER PRODUCT Awww let me tell you what that is a feeling beyond compare because then 1. You get to tell them straight up - Stop being a fucking smart arse - No i'm not the criminal..... YOU ARE 2. You get to debate it with them WHILE THEY MAKE YOU A COFFEE 3. You get to put them in a spot or a corner that they can't debate out of because your'e just about to buy another product and if they fuck you around then you won't buy it 4. and when they back down you get to say ,...... Now for giving me shit i want a substantial discount on this product and they can't refuse otherwise you don't buy it I can't count how many times i've done this and saved shit loads BEST FEELING IN THE WORLD and oh.... almost forgot then... YOU COME HOME........AND IMMEDIATELY VOID THE WARRANTY ON THE PRODUCT THAT YOU GOT A STUPID DISCOUNT ON ONLY BECAUSE YOU VOIDED THE WARRANTY ON THE PRODUCT BEFORE.............. while enjoying a cup of coffee LMFAO
The mesh on the back of the can looks like what you'd find on the front door of a microwave. Considering it's right over the main switching power supply it might have been a last-minute fix to pass EMI testing. I realize it's on the outside of the can, but from what I've learned about antenna theory it wouldn't be out of the question for excessive noise to permeate through.
🤦 No no no no no no, just NO. NO. That mesh is 100% aesthetic for when you're looking at the vent holes with your eyes. It's so you don't see so much shiney aluminum. Usually it's done with paint.
Because the datasheets for the FPGAs require it. Because if the voltage was any lower, the transistors wouldn't switch, and if they were any higher, they'd dissipate too much power. It's just a consequence of really gunning for maximum performance, which a typical 32-bit ARM chip really isn't doing.
@@TheHuesSciTech "because datasheet says so" doesn't seem like an answer to this question. I would say that the FPGAs contain some distinct blocks, that do have various electrical characteristics, thus needing different voltage supplies - but it still doesn't sound like a sensible answer.
Smaller chips like a typical ARM SoC usually do have internal voltage regs for the core voltage, AFAIK? FPGAs generally have a much larger die, can take a fair bit of power when fully configured, but also need to support a whole slew of different IO voltage standards. (like 3V3 LVTTL, LVDS, PECL, PCIe, 1V8, 1V2 etc. Hence VCCIO is broken out for each bank.) The Cyclone III needs 1V2 for the core, 2V5 for the PLLs, and 3V3 for VCCIO (if you're just using LVTTL or whatever). Again, I would guess that there's just a ton of logic to support on modern FPGAs, and a very large clock tree, so the regulators need to be external? The regs would also take up die space, of course, and FPGAs need as much as they can get.
FFS. Even the most cutting edge processes like 5 or 7 nm require only "low" voltage for core and "high" voltage for I/O. Interface IP has its own regulators anyway. It's ridiculous.
Coz there is a lot of stuff in them which is quite universal to fit different applications. Core is one thing, multiple I/O banks with different standards might be powered at different voltages depending on connected stuff (lvds, jesd, TTL/CMOS), PLLs and then SERDES are once again beasts on their own.
Nice tear down. I love seeing the inside of electronics and learning how they work. BTW: I really like that chart of electromagnetic radiation you have there in the back.
Front panel design (especially the jagged lines that visually groups together buttons and knobs) looks just terrible! Has Rigol laid off designer of their previous products?
Fast-forward to 2023-2024... Their ~400$ DHO800 series (that is the replacement of 1000Z series Dave mentioned) uses custom, 800MHz-capable front-end ASICs (like ones found in HDO4000). Of course I doubt the rest of the front-end - from discrete components - can handle that, the chip might be SW-limited, and the sampling rate (1.25 GHz) wouldn't allow for anything over 600 MHz (Nyquist) on a single channel, but still, the front-end ASIC is the same.
Great vid 👍 and love the Mercedes Benz ad on changing the inside filter for the car, even though I drive a Ford 😏 and better then a dog food ad for a dog I don’t have and make up I don’t use as I’m male 😊
Dave it would be fantastic if you could make a video / videos about shielding and ground! :-) Copper, sink, cast-aluminium, steel and so on, and grounding of via capacitors, diodes, grounding of hum, RF, EMI and so on! :-)
Dave, Try adding a resistor to the TX line. I have some MCU's that are picky about it and don't boot when I have TX connected without an extra resistor I think due to some type over voltage / cheap MCU. :-) The press any key u-boot message will probably only respond to serial input.
I bought a 4 channel Tektronix scope that covers up to 33GHz and it was just shy of a half million dollars. You can't compare the price of your economy car to a F1 race car. 🤦 They're not the same. Ever heard the phrase "apples and oranges?" [shakes head]
@@kr854 Oh the irony.... it takes a special kind of moron to exist in pure consumer space and yet act like they're a professional that they're not. Commercial test gear and consumer commodity gear are in different leagues. I'll bet you've never designed a product and probably never set foot in a development lab. That's just my guess.... am I correct? Am I? I've been designing electronic products for over 30 years and have gear that's being commercially produced right now. My engineering passes Intel's high speed design lab review so I couldn't care less about your uninformed opinion. You can buy my designs at major retailers world-wide, and that includes places like Walmart and B&H Photo and Video. I've bought individual scope probes that cost $6000 and can only be used about a dozen times before their characteristics deteriorate. Yeah... you tell me allllll about it.... 🤣🤣🤣🤣🤣
Eh, ADDA fans are ok. I’ve found them in some very nice computer power supplies. I’ve had one of them die, but it wasn’t the end of the world. They’re *reasonable* - still a brand name though. I think CapXon are ok as well. They’re no Panasonic capacitors, which in my humble opinion are the best in the business, but capacitors are also nowhere near as bad as they used to be, across the board. *Everyone* got stung by that scandal, and even capacitors you’ve never heard of are probably better than the shit from the early 2000s.
Last year I bought some new capacitors on eBay and within a year even the ones left in the drawer had dropped badly in value, and the one in a machine had died. Threw them all out. Bad capacitors are still being manufactured.
Benedikt Müssig fair enough. I am by no means a *nix expert - although I’d like to be. But I have cancelled some boots in some operating systems by doing this before. Cisco you can cancel boot with the keyboard, but that’s hardly Linux I guess.
Your approach surely works with a lot of machines and it's certainly not wrong to believe that it may work here, but I do have experience with U-Boot and know it's in's and out's enough to confirm this will not be the case here ;-)
Benedikt Müssig that’s fair enough! I have cancelled boots in some embedded devices by doing as described, but I guess the point I should’ve been trying to make instead, is the fact that embedded devices almost always have weird and wonderful ways of doing more or less the same thing. It’s a source of infuriation when reversing, but it’s also a source of great interest to me. Everything is subtly different. And every new piece of kit I play with even for just a few minutes, teaches me something brand new. But I guess some of what I do has the same problem of “no one learns how to use Linux by watching progress bars advance when installing it in a virtual machine”. which is some of the problem I e been having. I honestly can’t settle on a distro to use, because I don’t really know what I want to do yet! I want to do so many different things! I like networking, so something BSD based would be useful. I have about 40TB of old 2TB drives that is yet to be consolidated, I know I’ll be using FreeNAS here, more than likely. But I want to learn sysadmin and network admin stuff too, so... red hat or CentOS? I’m sure you get my point. I might as well ask someone who is well versed in this world; where is a good place for me to start? I know effectively forcing myself to use the desired OS I want to learn is a great way to start - this is one of the primary reasons I bought a Mac a few years ago, and I love it! But the point of OS X is really to reduce command line usage as much as possible, and make computing more accessible, which I can’t help but admire. Regardless of which side of the fence you sit on, being able to put more computers into the hands of more people is only a good thing. I do find myself playing around in the command line from time to time, and installing home brew has made said command line *a lot* more fun to play with. But it is still a Mac. As a Linux fan friend puts it, it is basically BSD on training wheels. It might be hard to concede to that, I think OS X is a very capable and very powerful OS, even if it does pick and choose its components to the point where it may as well be it’s own bespoke thing and the fact that it uses a Unix subsystem is a mere technicality at this point. I’m rambling on far too much, and I left the topic in the dust ages ago - but I’d love to know what you think. You seem to be quite well versed in this subject, certainly far more than I am. Sorry for rabbiting on so much, but *nix has always been one of those things that I want so much to get in to, but there is so much to it I literally don’t know where to start. I’m at precisely the same point with coding right now, for precisely the same reason. Thanks for reading, if you made it this far! (I won’t blame you if you didn’t. :P)
I know this isn’t the kind of video to ask, but I’m sorry, I can’t wait for the review! I mentioned on the second channel that I’d be interested in this as a first DSO/MSO combo. If the thing is hackable, do you think it’d still be too much for a first scope? I’ve only played with the old Rigol at Hackerspace, but I know that I will be using the MSO feature a lot (which the HSBNE scope does not have, unfortunately) and it would be amazing if I could have the bandwidth necessary to probe and capture USB traffic. Thoughts? P.S. Most of what I will be doing is things like internal debugging, so RS232, UART, I2C, stuff like that. But being able to step up to USB more or less straight away would be extremely handy. I’d love to hear what you guys have to think.
That is a rather good point, Dave. I didn’t think about that. I had a squizz at the 1000 family, and I really want something that is basically a 1054Z with the logic analyzer. But I’m having trouble figuring out which one I need. Because these things can be hacked, I want to make sure the hardware is there to enable it. Could you perhaps point me into the right direction? Cheers! Also, I’ll be buying my first lab power supply soon, that little module you can get.
Putting a really high end, high frequency "front end" on an inexpensive 'scope does not make a lot of sense ! What most "low end" users want are deeper buffers, more samples/second/channel and more bits of resolution. 500MS/sec/channel would give you the capability to capture "one time" events up to close to 50MHz. And with 10 or 12 bits, now we have something "to write home about" !
Hopefully those power supply caps dry out in a few years, and I can afford to pick one up for 'parts or repair' and toss a few dollars worth of nichicon caps at it.
Capxon caps are so terrible I almost always replace them on sight. Everything from leaking to exploding, sometimes within the first six months of a new power supply.
does dave have any good videos on rf, rf testing and Bluetooth? there is this school project coming up.... ...also I chuckled when you called the insides very 'Spartan'
man I would kill for one of these... how do hobbyists come up with that kind of change anyway? I'm lucky to afford a decent soldering iron, I don't know how people do it man.
Id have to assume so... but that's not me. I'm just a hobbyist goofing around. I've no aspirations of doing it for money. I suppose that's why I can't afford the tools, eh? hehe
I hate the placing of the backup battery. It is not the first time i need to repair and replace tracés and chips from battery leaks. I really would like to see a spaced out room for those battery’s . It is a expensive gear
i had 2, coin cells leaking at the age of 8 years. and some nimh cells as well. would be a shame to lose a nice scope for that reason. it also could be that i had 2 bad case cells, and with the innovation of these days leaking is impossible?
I think some computer PSU manufacturers use ADDA fans. Although I'm not a expert in computer psu's, on Johnny Guru's site they are regarded as average (noise wise).
Heya Dave Got a question for you about frequency in power transmission systems. I wonder if you could break down for a layman why for example power systems in air craft are 400hz where as utility or mains power from your house socket is 50-60hz and how they decided at the design stage to use what they do in there respective systems. Long time watcher first time posting on a vid awesome content by the way keep up the good work with the chanel all the best from the UK.
In college was told the 400hz originated in aircraft as a weight saving strategy, 400Hz requires smaller and therefore lighter components especially inductor and capacitors in power supply smoothing circuits. This was back in tube days. As to the 60 HZ it seems for whatever reason it was smart cause of the 60 sec clock, makes a simple clock motor that locks onto the frequency of the line very accurate since the power is tightly regulated at that freq.
No, they put a piece of camouflage in the right place. The alternative would have been painting which is more costly in production and isn't as effective visually.
@@Peter_S_ why would it have cutouts matching the layout the mains plug, and all of the slots that are cut in the sheet metal case? A die or cutting process would be more time consuming and expensive for that shape than just a square. And no manufacturer I have seen paints under vent holes to disguise bare sheet metal.
Very simple.... The cutouts allow for a single die to make pieces for either side and this gives complete flexibility in the final thermal management solution. This way the final thermal design can get tweaked quite a bit at the last stages of product testing without having to rework the plastic which is both expensive and time consuming. Everything like that camo piece is made with a die to produce lots of units at once from a sheet or a roll and everything has a subassembly part number. It's just how it's done in manufacturing. There's no need to have two designs when a single one will work for either side. The metalwork is done with a CNC punch-press so making alternate versions is easy as pie and I'll guarantee they went through multiple revisions of metal during testing for thermal reasons. They probably even looked at a crossflow design with no rear facing fan which would need two camo panels. Somewhere in the lab is probably a prototype with big heat sinks and no fan done as a test, not because they thought it would be the final but to see how close they could reasonably get. Consider the open area of that mesh which is roughly 23% if it's the 3/64" holes on 3/32" centers (of course the metric equivalent) that I think it is (a standard pattern), and now compare the hit a fan would take in efficiency trying to draw air through it. Clearly it's not a filter and if you place it across the intake of a fan, you're only going to get 23% performance out of the fan, other turbulence aside. Taking a hit of 33% is common and acceptable but 77% will get you demoted or fired from a design team. You also have to consider the noise profile... the plastic and metalwork reflect high flow, low noise which is what you get from the big honeycomb pattern. Little round holes would not only be inefficient, but noisy. Speaking as a manufacturer, I'll assure you that painting and dyed anodization is very commonly used behind grills to reduce the visual impact of the device guts. Stickers are used less often but you see a lot of them too. I won't say who I've worked for buy you can buy my gear here: www.bhphotovideo.com/ and there's a place in the enclosure painting instructions in each applicable case specifying which other spots to touch with paint on the inside of the enclosures.
Trace length matching does two things: (1) timing equalization: if you're looking at two signals you need them to arrive at exactly the same time and the speed of light is a real limit in copper so lengths must match. (2) signal integrity: Most high speed signals are sent as a "differential pair" where you have your signal and an inverted copy of the signal running in parallel down a pair of traces. Ideally the signal and the anti-signal travel shoulder to shoulder down the pair of traces and if one wave gets ahead of the other, the waveform gets "smeared" by one wavefront getting capacitively coupled to the other trace in the pair. If all the lengths are matched then any interference coming from that transmission line will be composed of a signal and an anti-signal which (hopefully) cancel each other out.
Did you watch the video? it was mentioned and shown multiple times that the frontend is capable of up to 4ghz(according to rigol anyway), so they will certainly use that capability at some point.
The cutouts around the edge of that mesh appear to match up with the fan mounts and the power connector, almost like it was supposed to go over there instead. And in case you haven't heard, you might be interested to hear that "Warranty Void If removed" stickers have been ruled illegal over here in the US .
They must make a right hand drive and left hand drive version.
The stickers aren't illegal, they just don't have a legal basis.
Yup, its the same here. Companies can put whatever stickers they want on there, they are not illegal(even if media might tell you that, assuming you not actually gonna look it up), but it doesn't actually void the warranty when you break them.
Just like article 13 of the EU doesn't ban memes, they didn't ban warranty void stickers either. Media outlets just tend to exaggerate and assume that people don't actually look at the rulings or laws themselfs, and that usually works.
They have to prove that you caused it to fail. Opening a product up doesn't automatically break it, therefore the sticker is completly useless. It'd be ruled that way in most places probably.
People and Dave. The sticker are illegal if the manufacturer made it restriction to third party part (or small business), cost to repair device or no waiver from FTC. It is part of US Federal law since there have been unfairly history in American. Where the primary company had hurt both consumers and small business a lot. So, to have "warranty void if seal" broken sticker, it must not costing to repair it or providing the part. That is Magnuson-Moss Warranty Act, a federal law.
"hit any key" needs to be on the RS232 port, not a usb keyboard. Standard u-boot linux bootloader.
That one made me shake my head.
I hope he tries to work it out and makes a modding video of sorts.
@@TheTyrost either way, if I see that message coming out of the serial port, I would expect to send a character over the serial port, not an attached keyboard. Maybe I would try a physical keyboard afterwards, but not first off the bat. 😆
That one made me facepalm a little too.
Even if he could get TX/RX to work - that would only get him as far as the u-boot bootloader. It'd be much easier just to read the linux image from flash, then "binwalk -e" on it and go from there.
Rx is probably working, but using different logic levels than Dave's USB serial adapter. If the USB serial adapter's default logic level screws with the receive, the CPU may not be willing to continue booting.
"Press key to stop boot" will be looking for serial chars, not usb keyboard.
Hardware links will be mostly so the software can know board rev and deal with any changes over time with a single Fw version
My best guess is the resistors in hw-id is for the software to read the board revision if something changed etc - not to lock down the firmware. I still think you can purchase software option codes and enter and it would unlock that way.
Thank You EEVblog Keep up the Good Work! :)
Thumbs up Indiana USA.
Lol, so i'm guessing Rigol sent dave the scope? of course they sent the top end model, so he can't use it to figure out how to hack a lower end one :P
but now someone else just has to buy the cheaper version and we can see the differences
You can always disable things, also gives you a known maxed out state.
“Déjà vu review” = Dave made this video and simply “forgot to publish” then a few weeks later he started his research and recording all over again and started having some flashbacks... checked his hard drive and the video was all done ready to go!! Hahahaa one of these moments... =]
Was bound to happen eventually...
Dave ja vu
Should have a small desk fan to keep airflow on the heatsinks when the back is off, just in case..
Even at the PSU you can see they planned to use an extra fan. There is a second fan connector.
Great Video, love scope teardowns :D
I like the black color more shown on their prototype machine.
"... That makes it look, I dunno ... Meshy." @3:27 is such a wonderfully quotable line.
I wonder if a hack for these will ever be published. After-all they are more expensive so not as many people can get their hands on these to start hacking.
Chicken and egg kinda thing.
That hardware version is probably just that. The hardware version, so they can run the same firmware images across multiple scopes, or at the very least the same firmware for different PCB versions of the same model.
I love that feeling of voiding the warranty!
me too, the feeling of voiding the warranty as soon as you get the scope home while having a cup of coffee while opening the box
ahhhhhhhhhhh heaven
BUT THERE IS A BETTER FEELING
Having the manufacturer suspect that you've unlocked your scope
then they challenge you on it WHILE YOU ARE ABOUT TO BUY ANOTHER PRODUCT
Awww let me tell you what
that is a feeling beyond compare
because then
1. You get to tell them straight up
- Stop being a fucking smart arse
- No i'm not the criminal..... YOU ARE
2. You get to debate it with them WHILE THEY MAKE YOU A COFFEE
3. You get to put them in a spot or a corner that they can't debate out of
because your'e just about to buy another product and if they fuck you around then you won't buy it
4. and when they back down you get to say ,...... Now for giving me shit i want a substantial discount on this product and they can't refuse otherwise you don't buy it
I can't count how many times i've done this and saved shit loads
BEST FEELING IN THE WORLD
and oh.... almost forgot
then... YOU COME HOME........AND IMMEDIATELY VOID THE WARRANTY ON THE PRODUCT THAT YOU GOT A STUPID DISCOUNT ON ONLY BECAUSE YOU VOIDED THE WARRANTY ON THE PRODUCT BEFORE.............. while enjoying a cup of coffee
LMFAO
that black mesh looks like a misplaced dust filter
@9:28 My friend works for ADDA. Their DC brushless fan is not bad for its price. A cheaper alternative to Delta.
Looks nice and clean. I dig the black development prototype!
Wow, this looks cool, thanks Dave!
Nice to see Coilcraft inductors in the power supply section.
The mesh on the back of the can looks like what you'd find on the front door of a microwave. Considering it's right over the main switching power supply it might have been a last-minute fix to pass EMI testing. I realize it's on the outside of the can, but from what I've learned about antenna theory it wouldn't be out of the question for excessive noise to permeate through.
🤦 No no no no no no, just NO. NO.
That mesh is 100% aesthetic for when you're looking at the vent holes with your eyes. It's so you don't see so much shiney aluminum. Usually it's done with paint.
I thought the same for a moment, but I doubt that beafy can was insufficient. The cosmetic reason seems very plausible.
20 volt rail fanboy here :-)
10Gsamples/S for 500MHz seams quite a lot, could they have a 1 GHz model on the way?
Try placing a piece of Kapton tape on the shiny tin cans. This will allow you to 'see' their temperature with your thermal camera 😊
main application processor is xilinx zynq-7000 series. see ttyPS0, pl330, ps7-ddrc, zynq-edac in serial output.
What is the reason FPGAs need all those different voltages when typical CPUs don't? Possible Fundamentals Friday episode?
Because the datasheets for the FPGAs require it. Because if the voltage was any lower, the transistors wouldn't switch, and if they were any higher, they'd dissipate too much power. It's just a consequence of really gunning for maximum performance, which a typical 32-bit ARM chip really isn't doing.
@@TheHuesSciTech "because datasheet says so" doesn't seem like an answer to this question. I would say that the FPGAs contain some distinct blocks, that do have various electrical characteristics, thus needing different voltage supplies - but it still doesn't sound like a sensible answer.
Smaller chips like a typical ARM SoC usually do have internal voltage regs for the core voltage, AFAIK?
FPGAs generally have a much larger die, can take a fair bit of power when fully configured, but also need to support a whole slew of different IO voltage standards.
(like 3V3 LVTTL, LVDS, PECL, PCIe, 1V8, 1V2 etc. Hence VCCIO is broken out for each bank.)
The Cyclone III needs 1V2 for the core, 2V5 for the PLLs, and 3V3 for VCCIO (if you're just using LVTTL or whatever).
Again, I would guess that there's just a ton of logic to support on modern FPGAs, and a very large clock tree, so the regulators need to be external?
The regs would also take up die space, of course, and FPGAs need as much as they can get.
FFS. Even the most cutting edge processes like 5 or 7 nm require only "low" voltage for core and "high" voltage for I/O. Interface IP has its own regulators anyway. It's ridiculous.
Coz there is a lot of stuff in them which is quite universal to fit different applications. Core is one thing, multiple I/O banks with different standards might be powered at different voltages depending on connected stuff (lvds, jesd, TTL/CMOS), PLLs and then SERDES are once again beasts on their own.
Nice tear down. I love seeing the inside of electronics and learning how they work.
BTW: I really like that chart of electromagnetic radiation you have there in the back.
you can actually order that from him
i have one, it's cool
I'm thinking the jumpers are just to tell it the board revision or something, not to configure options
Adda fans are what I see in OEM PC PSUs and I've like never seen one fail...
Big ass fan is probably more for the quietness of it than raw airflow
Really want to see the power consumption and the HDMI out!
*Shouldn't turn it on for too long without the fan in place Mr. Jones!*
oh yeah, because he doesn't know what he's doing!?
it looks like that mesh was supposed to go over the vent holes on the other side
What a great idea to shoot a video of the teardown !
7 minutes in: "it's incredibly spartan, isn't it?" Well, with all the Xilinx FPGAs, of course it's Spartan...
Rigol sends Dave newest scope to tear down and promote. One of the first things Dave does is muse about hacking it to have higher bandwidth.
That's part of the marketing
Why not!
Hey, it works. I bought my Rigol 2072 because of one of your videos knowing I could turn it into a 2202 thanks to the eeblog forum.
I bought a DS1054Z because of that, and I haven't even hacked anything on it yet.
And I got dso5102 for the same reason
Still only 8 bits? You would think that they would be moving towards higher vertical resolution by now. Not everyone works in a digital world.
Quick Respinse: great video!
Front panel design (especially the jagged lines that visually groups together buttons and knobs) looks just terrible! Has Rigol laid off designer of their previous products?
ADDA Fans are a good brand !
Used them in mass production multiple times.
VERY COOL!!! ITS BEEN SO LONG!!!
Lol at the ANSI escape sequences in the serial debug interface's prompt
The HW Ver solder jumpers goes from 0 to 9 if read from right to left using a tilted squint eye. =)
Thanks Rigol!
Fast-forward to 2023-2024...
Their ~400$ DHO800 series (that is the replacement of 1000Z series Dave mentioned) uses custom, 800MHz-capable front-end ASICs (like ones found in HDO4000).
Of course I doubt the rest of the front-end - from discrete components - can handle that, the chip might be SW-limited, and the sampling rate (1.25 GHz) wouldn't allow for anything over 600 MHz (Nyquist) on a single channel, but still, the front-end ASIC is the same.
Ok the ad chosen for this video is for the TechTronics oscilloscope. I wonder if that was planned
Very instructive ,Thank you 👍👍👍👍
With all the space, why put in such a non user serviceable battery?
Nice, thanks for sharing:-)
Nice design to look at :-)
Dave Can you please make a detail video on Open drain and push pull concepts and their advantages?
Huh. What happened to the function gen? I must have had a blackout when that part was discussed.
top quality dave
The fan looks like a standard 120mm type. So a good quiet computer case fan (eg Scythe or such) could be a cheap upgrade.
Useful video 👍 Excellent 👍
Great vid 👍 and love the Mercedes Benz ad on changing the inside filter for the car, even though I drive a Ford 😏 and better then a dog food ad for a dog I don’t have and make up I don’t use as I’m male 😊
Swan has a flamingo on er
also
wouldnt be surprised if someone nicked the back ticks on those duplicate rails and they were actually negative rails
Dave it would be fantastic if you could make a video / videos about shielding and ground! :-)
Copper, sink, cast-aluminium, steel and so on, and grounding of via capacitors, diodes, grounding of hum, RF, EMI and so on! :-)
Dave, Try adding a resistor to the TX line. I have some MCU's that are picky about it and don't boot when I have TX connected without an extra resistor I think due to some type over voltage / cheap MCU. :-)
The press any key u-boot message will probably only respond to serial input.
$100 per MHz? so my 100MHz oscilloscope i bought for $450 would cost $10,000, sounds like a rip off to me
I bought a 4 channel Tektronix scope that covers up to 33GHz and it was just shy of a half million dollars.
You can't compare the price of your economy car to a F1 race car. 🤦 They're not the same.
Ever heard the phrase "apples and oranges?" [shakes head]
my comment was a sarcastic comment based on what Dave said at the end of his video.
@@Peter_S_ You don't seem to be very bright. Even with your example it's no where near $100 per MHz.
@@kr854 Oh the irony.... it takes a special kind of moron to exist in pure consumer space and yet act like they're a professional that they're not. Commercial test gear and consumer commodity gear are in different leagues. I'll bet you've never designed a product and probably never set foot in a development lab. That's just my guess.... am I correct? Am I? I've been designing electronic products for over 30 years and have gear that's being commercially produced right now. My engineering passes Intel's high speed design lab review so I couldn't care less about your uninformed opinion. You can buy my designs at major retailers world-wide, and that includes places like Walmart and B&H Photo and Video. I've bought individual scope probes that cost $6000 and can only be used about a dozen times before their characteristics deteriorate. Yeah... you tell me allllll about it.... 🤣🤣🤣🤣🤣
@19:20 A hundred dollars per MegaHertz is gonna be one helluva expensive scope. I think you mean a 100 dollars per 100 Meg Hertz, ^____^
"$100 per MHz" would be a bit expensive
Did I say that? Doh.
i assume you meant $100 per 100MHz or $1 per MHz
Phew. Luckily someone spotted the mistake and posted a pedantic comment about it!
Did that black flamingo version of this say 4GHz 20GSa/s on the front!? I want a video on that one.
Eh, ADDA fans are ok. I’ve found them in some very nice computer power supplies. I’ve had one of them die, but it wasn’t the end of the world. They’re *reasonable* - still a brand name though. I think CapXon are ok as well. They’re no Panasonic capacitors, which in my humble opinion are the best in the business, but capacitors are also nowhere near as bad as they used to be, across the board. *Everyone* got stung by that scandal, and even capacitors you’ve never heard of are probably better than the shit from the early 2000s.
Last year I bought some new capacitors on eBay and within a year even the ones left in the drawer had dropped badly in value, and the one in a machine had died. Threw them all out. Bad capacitors are still being manufactured.
Have you tried plugging a USB keyboard in to the scope? Usually that will can autoboot in Linux.
I doubt it would. Cancelling autoboot is done by transmitting any character via said serial port. U-Boot does not have keyboard support.
Benedikt Müssig fair enough. I am by no means a *nix expert - although I’d like to be. But I have cancelled some boots in some operating systems by doing this before. Cisco you can cancel boot with the keyboard, but that’s hardly Linux I guess.
Your approach surely works with a lot of machines and it's certainly not wrong to believe that it may work here, but I do have experience with U-Boot and know it's in's and out's enough to confirm this will not be the case here ;-)
Benedikt Müssig that’s fair enough! I have cancelled boots in some embedded devices by doing as described, but I guess the point I should’ve been trying to make instead, is the fact that embedded devices almost always have weird and wonderful ways of doing more or less the same thing. It’s a source of infuriation when reversing, but it’s also a source of great interest to me. Everything is subtly different. And every new piece of kit I play with even for just a few minutes, teaches me something brand new. But I guess some of what I do has the same problem of “no one learns how to use Linux by watching progress bars advance when installing it in a virtual machine”. which is some of the problem I e been having.
I honestly can’t settle on a distro to use, because I don’t really know what I want to do yet! I want to do so many different things! I like networking, so something BSD based would be useful. I have about 40TB of old 2TB drives that is yet to be consolidated, I know I’ll be using FreeNAS here, more than likely. But I want to learn sysadmin and network admin stuff too, so... red hat or CentOS? I’m sure you get my point.
I might as well ask someone who is well versed in this world; where is a good place for me to start? I know effectively forcing myself to use the desired OS I want to learn is a great way to start - this is one of the primary reasons I bought a Mac a few years ago, and I love it! But the point of OS X is really to reduce command line usage as much as possible, and make computing more accessible, which I can’t help but admire. Regardless of which side of the fence you sit on, being able to put more computers into the hands of more people is only a good thing. I do find myself playing around in the command line from time to time, and installing home brew has made said command line *a lot* more fun to play with. But it is still a Mac. As a Linux fan friend puts it, it is basically BSD on training wheels. It might be hard to concede to that, I think OS X is a very capable and very powerful OS, even if it does pick and choose its components to the point where it may as well be it’s own bespoke thing and the fact that it uses a Unix subsystem is a mere technicality at this point.
I’m rambling on far too much, and I left the topic in the dust ages ago - but I’d love to know what you think. You seem to be quite well versed in this subject, certainly far more than I am. Sorry for rabbiting on so much, but *nix has always been one of those things that I want so much to get in to, but there is so much to it I literally don’t know where to start. I’m at precisely the same point with coding right now, for precisely the same reason.
Thanks for reading, if you made it this far! (I won’t blame you if you didn’t. :P)
That’s a flamingo.
How about some testing vs a tear down.
Good video. 👏🙌👍👌
I know this isn’t the kind of video to ask, but I’m sorry, I can’t wait for the review! I mentioned on the second channel that I’d be interested in this as a first DSO/MSO combo. If the thing is hackable, do you think it’d still be too much for a first scope?
I’ve only played with the old Rigol at Hackerspace, but I know that I will be using the MSO feature a lot (which the HSBNE scope does not have, unfortunately) and it would be amazing if I could have the bandwidth necessary to probe and capture USB traffic. Thoughts?
P.S. Most of what I will be doing is things like internal debugging, so RS232, UART, I2C, stuff like that. But being able to step up to USB more or less straight away would be extremely handy. I’d love to hear what you guys have to think.
Yes, you wouldn't spend this one a first scope. Use the money you save on other gear for your lab.
That is a rather good point, Dave. I didn’t think about that. I had a squizz at the 1000 family, and I really want something that is basically a 1054Z with the logic analyzer. But I’m having trouble figuring out which one I need. Because these things can be hacked, I want to make sure the hardware is there to enable it. Could you perhaps point me into the right direction? Cheers!
Also, I’ll be buying my first lab power supply soon, that little module you can get.
They really seem to have went out of their way to use brand-name parts on the main board. Coilcraft inductors, ADDA fan, TDK buzzer
Rigol should give the low end market a good shake once again :)
Putting a really high end, high frequency "front end" on an inexpensive 'scope does not make a lot of sense ! What most "low end" users want are deeper buffers, more samples/second/channel and more bits of resolution. 500MS/sec/channel would give you the capability to capture "one time" events up to close to 50MHz. And with 10 or 12 bits, now we have something "to write home about" !
Hopefully those power supply caps dry out in a few years, and I can afford to pick one up for 'parts or repair' and toss a few dollars worth of nichicon caps at it.
Capxon caps are so terrible I almost always replace them on sight. Everything from leaking to exploding, sometimes within the first six months of a new power supply.
"Quick respinse code" at 15:36 lol.
dose the 7 stand for £7,308.00
does dave have any good videos on rf, rf testing and Bluetooth? there is this school project coming up.... ...also I chuckled when you called the insides very 'Spartan'
I guess Dave found his files "oh the stupidity " LOL
Is there a normal review of this item to accompany the teardown?
I'm definitely a 20v rail fanboy.
I have kind of an "already seen that" feeling...
man I would kill for one of these... how do hobbyists come up with that kind of change anyway? I'm lucky to afford a decent soldering iron, I don't know how people do it man.
If your making money fixing or doing things, it'll pay for itself in less than a month
Id have to assume so... but that's not me. I'm just a hobbyist goofing around. I've no aspirations of doing it for money. I suppose that's why I can't afford the tools, eh? hehe
A good multimeter is good enough.
I'm happy with my 1054z. (hacked) enough for most repairs and playing around. But sometimes I wish I had an extra 100MHz... Still love this scope!
Nah fam, gotta get a basic oscilloscope
Code Sourcery is a toolchain from Mentor. That's most likely where the "Sourcery" tags come from.
13:28
@@Graham_Wideman , ha, I must have been looking away when he said it. Totally missed it on the screen.
I see it's running Linux 3.12, from 2013, a bit surprising for a new scope..
Nice. I've been meaning to upgrade to 4CH for ages... now just need to rob a bank.
"those filthy buggers at Rigol" lmao
Well that looks nice but it was already hard enough to afford my $20 Tektronix 2440 that makes the whole place smell like burning CRT
I hate the placing of the backup battery. It is not the first time i need to repair and replace tracés and chips from battery leaks.
I really would like to see a spaced out room for those battery’s . It is a expensive gear
I have never seen a lithium coin cell leak. Definitely not the same as NiCd that used to be very common.
i had 2, coin cells leaking at the age of 8 years. and some nimh cells as well. would be a shame to lose a nice scope for that reason. it also could be that i had 2 bad case cells, and with the innovation of these days leaking is impossible?
I think some computer PSU manufacturers use ADDA fans. Although I'm not a expert in computer psu's, on Johnny Guru's site they are regarded as average (noise wise).
Heya Dave
Got a question for you about frequency in power transmission systems. I wonder if you could break down for a layman why for example power systems in air craft are 400hz where as utility or mains power from your house socket is 50-60hz and how they decided at the design stage to use what they do in there respective systems.
Long time watcher first time posting on a vid awesome content by the way keep up the good work with the chanel all the best from the UK.
In college was told the 400hz originated in aircraft as a weight saving strategy, 400Hz requires smaller and therefore lighter components especially inductor and capacitors in power supply smoothing circuits. This was back in tube days. As to the 60 HZ it seems for whatever reason it was smart cause of the 60 sec clock, makes a simple clock motor that locks onto the frequency of the line very accurate since the power is tightly regulated at that freq.
Please do a modding video!
By finding a way to change the hardware configure /version, Dave is about to put Rigol out of business too 😂😂
Yay it made it!
Actually, I think someone has just put the dust filter in the wrong place... the cutouts on the mesh perfectly match the actual fan location.
No, they put a piece of camouflage in the right place. The alternative would have been painting which is more costly in production and isn't as effective visually.
@@Peter_S_ why would it have cutouts matching the layout the mains plug, and all of the slots that are cut in the sheet metal case? A die or cutting process would be more time consuming and expensive for that shape than just a square.
And no manufacturer I have seen paints under vent holes to disguise bare sheet metal.
Very simple.... The cutouts allow for a single die to make pieces for either side and this gives complete flexibility in the final thermal management solution. This way the final thermal design can get tweaked quite a bit at the last stages of product testing without having to rework the plastic which is both expensive and time consuming. Everything like that camo piece is made with a die to produce lots of units at once from a sheet or a roll and everything has a subassembly part number. It's just how it's done in manufacturing. There's no need to have two designs when a single one will work for either side. The metalwork is done with a CNC punch-press so making alternate versions is easy as pie and I'll guarantee they went through multiple revisions of metal during testing for thermal reasons. They probably even looked at a crossflow design with no rear facing fan which would need two camo panels. Somewhere in the lab is probably a prototype with big heat sinks and no fan done as a test, not because they thought it would be the final but to see how close they could reasonably get.
Consider the open area of that mesh which is roughly 23% if it's the 3/64" holes on 3/32" centers (of course the metric equivalent) that I think it is (a standard pattern), and now compare the hit a fan would take in efficiency trying to draw air through it. Clearly it's not a filter and if you place it across the intake of a fan, you're only going to get 23% performance out of the fan, other turbulence aside. Taking a hit of 33% is common and acceptable but 77% will get you demoted or fired from a design team. You also have to consider the noise profile... the plastic and metalwork reflect high flow, low noise which is what you get from the big honeycomb pattern. Little round holes would not only be inefficient, but noisy.
Speaking as a manufacturer, I'll assure you that painting and dyed anodization is very commonly used behind grills to reduce the visual impact of the device guts. Stickers are used less often but you see a lot of them too. I won't say who I've worked for buy you can buy my gear here: www.bhphotovideo.com/ and there's a place in the enclosure painting instructions in each applicable case specifying which other spots to touch with paint on the inside of the enclosures.
Let's see this vs...idk 1054z the ultimate shoot out!
whats the need for trace length matching ?
....... Signal critical timing.
Trace length matching does two things:
(1) timing equalization: if you're looking at two signals you need them to arrive at exactly the same time and the speed of light is a real limit in copper so lengths must match.
(2) signal integrity: Most high speed signals are sent as a "differential pair" where you have your signal and an inverted copy of the signal running in parallel down a pair of traces. Ideally the signal and the anti-signal travel shoulder to shoulder down the pair of traces and if one wave gets ahead of the other, the waveform gets "smeared" by one wavefront getting capacitively coupled to the other trace in the pair. If all the lengths are matched then any interference coming from that transmission line will be composed of a signal and an anti-signal which (hopefully) cancel each other out.
brand spanking new?!...its only a month old :P
very good scope i will buy hantek
They put a finned heat sink underneath the EMI shield on the front end? How is it supposed to get air flow? That's pretty lame thermal design.
It was a surprise for sure, connecting thermally to the can should be much better.
No that’s a flamingo
"20v fanboys" I laughed so hard
not to complain about the power, but 10Gs/s is more like in the range of a 1GHz frontend... more to come?
Did you watch the video? it was mentioned and shown multiple times that the frontend is capable of up to 4ghz(according to rigol anyway), so they will certainly use that capability at some point.