EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: PTR_1275 on November 11, 2018, 12:23:32 pm
-
I got a invite from the local keysight distributor for a Christmas sundowner and product showcase. They mentioned a 4 channel 1000x series scope, but I can’t find any sign of such a thing on the web. Has anyone else come across anything?
I have a rigol 1054z at home which has been fine, then we got a few 2 channel 1000x series scopes at work and they are in a whole different league while not being “that” much more expensive.
At work we only need 2ch, but at home I sometimes need 3 or 4 so I’d told myself that’s why I’m keeping the rigol. If they do bring out a 4 channel and it’s not much more than the 2 channel, would that be a game changer for the low end scopes or would it still not affect the super low price point?
-
I have an uncle that works at Nintendo, and he told me... ;)
(80's kid flashbacks, sorry)
But this distributor, do they also sell a certain other brand, starting with an S, that sells a 1000X series?
On your question regarding price, just looking very briefly at a local distribitur gives me a 25% price bump estimate.
-
Not the S brand, definitely keysight.
New Keysight 4-channel 1000X-series oscilloscopes - product launch!
That’s the line from the email.
I know it says product launch, but most of the time there’s info out on the web before any actual launch.
-
.. effectively killing off the 2000x series :box:
-
.. effectively killing off the 2000x series :box:
Yes, they should drop the 2000X line in that case, it's really struggling to hold any mainstream market share at all. But it could still be doing fine in the education sector?
-
.. effectively killing off the 2000x series :box:
Perhaps, or they are prepping for a refesh of the X-line? Say, pushing a new 10bit arch based on the S-series downstream to meet R&S new offerings?
But first things first.
-
It would make sense. IIRC the 2000X was the first MZ4 scope to be launched, so the first MZ5 product is likely to be at that level, too.
-
Have you heard anything about the 4 channel scope Dave?
The event is in a few weeks. I’ll try to get along and post info on here unless someone finds more info first.
-
Please do! I'm about to pull trigger on a 1000x scope but if they're releasing new models in a couple of weeks then I'll wait to see what they have to offer.
-
Have you heard anything about the 4 channel scope Dave?
The event is in a few weeks. I’ll try to get along and post info on here unless someone finds more info first.
Nope, but I've asked.
Keysight are usually keen to let me see stuff first, so it's probably unlikely they wouldn't tell me about a new low cost scope.
-
Just spoke to a senior Keysight guy at Electronica.
It exists, probably announced January, and will be in black.
-
Just spoke to a senior Keysight guy at Electronica.
It exists, probably announced January, and will be in black.
:o
Ya kidding us !
Not another one gone to the dark side ! :scared:
What's up with designers ? :-//
-
i should ask my lecroy guy if they're planning to put out new equipment in dark red :-DD
-
Just spoke to a senior Keysight guy at Electronica.
It exists, probably announced January, and will be in black.
Did he mention anything about price range?
-
:o
Ya kidding us !
Not another one gone to the dark side ! :scared:
What's up with designers ? :-//
While there have been examples of black devices before, Keysight made it their new standard look a while back. Since that moment all the Chinese are tripping over themselves to update the product line. You can't be any good if you don't copy the market leader, now can you? ;D
-
Yep, confirmed.
There have been a few delays, but target is now early Jan launch.
No other details yet.
-
:popcorn:
-
I’ll be curious to see if they have a unit to show at this evening in a few weeks if there have been delays.
Is it strange they haven’t notified you Dave, or is it because of the delays? Any chance of a tear down unit being sent your way?
-
Guy I spoke to has only seen pictures. Also hinted it may have LAN as standard
-
So, uh, @PTR_1275 can you forward me that invite? I'd love to see it >:D [e-mail removed]
I can also neither confirm nor deny any rumors at this point, but I can assure you our favorite YouTubers will get an early look at anything coming out in that space.
I can also say that Christmas will not come early, and - love it or hate it - this video may be appropriate:
https://www.youtube.com/watch?v=pAgnJDJN4VA (https://www.youtube.com/watch?v=pAgnJDJN4VA)
:popcorn:
-
By the way, if a Keysight representive may be reading along. This is the perfect class of instrument for a nice worldwide give-away, to have various people from various corners of the world with different interests show the world why this might be a very nice device for the serious hobbyist and professional alike.
-
By the way, if a Keysight representive may be reading along. This is the perfect class of instrument for a nice worldwide give-away, to have various people from various corners of the world with different interests show the world why this might be a very nice device for the serious hobbyist and professional alike.
They just give them to you tubers. Probably be the daily giveaway assuming they scopemonth again in 2019.
-
They just give them to you tubers. Probably be the daily giveaway assuming they scopemonth again in 2019.
That's a given, but why stop there? This is a release for the masses, as far as that applies to Keysight kit. That should be celebrated with the masses! R&S gave away their new oscilloscopes before and Keysight did the same with their power supplies and both times it's been received rather well I'd say.
-
I was told prices are already in the pricelists that the distributors have.
4channel 70MHz AU$1348 plus tax
70MHz with generator output AU$1626 plus tax.
Software license upgradable to 200MHz and the usual extra money for decoders.
Doesn’t look like it has individual vertical adjustment for each channel, but the photo I was sent looks like a blown up thumbnail.
-
Had a quick play with one at Electronica and a chat with one of their US guys.
I didn't find anything obvuously wrong, knobs maybe a little small, and might have been better in a colour that gave some contrast with the black.
Uart decode goes to 20Mbaud.
Screen feels a bit small once you have the MSO on
Mso probe has an acvtive head and 0.1" pin header, and they supply single wire probe leads, not paired with gnd so could have inductance issues.
Us guy also thought it nuts thay they did a 2ch version.
Has HDMI out, which for some reason has option to diwnscale to VGA and 720xsomething-is there any hdmi device that would need that?
sorry, wrong thread - was talking about the Rigol!
-
I was told prices are already in the pricelists that the distributors have.
4channel 70MHz AU$1348 plus tax
70MHz with generator output AU$1626 plus tax.
Software license upgradable to 200MHz and the usual extra money for decoders.
Doesn’t look like it has individual vertical adjustment for each channel, but the photo I was sent looks like a blown up thumbnail.
So pretty much double the price of 2 channel models and no hope for price drop on the old ones.
-
Had a quick play with one at Electronica and a chat with one of their US guys.
I didn't find anything obvuously wrong, knobs maybe a little small, and might have been better in a colour that gave some contrast with the black.
Uart decode goes to 20Mbaud.
Screen feels a bit small once you have the MSO on
I didn't know it had MSO. I've seen the front, so MSO connector must be on the back. And no Digital button.
Has HDMI out, which for some reason has option to diwnscale to VGA and 720xsomething-is there any hdmi device that would need that?
Maybe better scaling?
I'm starting to think we aren't talking about the same scope here?
-
I was told prices are already in the pricelists that the distributors have.
4channel 70MHz AU$1348 plus tax
70MHz with generator output AU$1626 plus tax.
Software license upgradable to 200MHz and the usual extra money for decoders.
Ouch.
It was a premium price product before, and I'm sure it's going to be very nice like the 1000X is, but this is getting pretty silly.
Doesn’t look like it has individual vertical adjustment for each channel, but the photo I was sent looks like a blown up thumbnail.
Button and knobs the same as the 2CH 1000X but with multiplexed vertical channel controls.
-
[...]
4channel 70MHz AU$1348 plus tax
70MHz with generator output AU$1626 plus tax.
[...]
That's a shame.
-
The 2 channel 70MHz with function gen is AU$1172 plus gst, so not double the price.
-
Had a quick play with one at Electronica and a chat with one of their US guys.
I didn't find anything obvuously wrong, knobs maybe a little small, and might have been better in a colour that gave some contrast with the black.
Uart decode goes to 20Mbaud.
Screen feels a bit small once you have the MSO on
I didn't know it had MSO. I've seen the front, so MSO connector must be on the back. And no Digital button.
Has HDMI out, which for some reason has option to diwnscale to VGA and 720xsomething-is there any hdmi device that would need that?
Maybe better scaling?
I'm starting to think we aren't talking about the same scope here?
Sorry wrong thread-was talking about the Rigol!
-
I can also say that Christmas will not come early, and - love it or hate it - this video may be appropriate:
Nah, for people bying the 1000X it's more like:
https://www.youtube.com/watch?v=-sUXMzkh-jI (https://www.youtube.com/watch?v=-sUXMzkh-jI)
-
Looks interesting.
-
Sorry wrong thread-was talking about the Rigol!
Haha @mikeselectricstuff - I was so confused!
-
Sorry wrong thread-was talking about the Rigol!
Haha @mikeselectricstuff - I was so confused!
In my defence I've just done 2 days doing every hall at Electronica,overdosing on gummy bears and posting on phone, with negligible sleep
-
with negligible sleep
They sound like a prog band.
-
When will a teaser be released?
-
@Hydrawerk, this isn't enough of a teaser for you?
-
Just to be clear, no MSO on the new 1000X?
-
Just to be clear, no MSO on the new 1000X?
Seems unlikely
-
Just to be clear, no MSO on the new 1000X?
Correct, no MSO. It's called the DSOX, not MSOX.
-
Will there be a MegaZoom V ASIC?
-
Will there be a MegaZoom V ASIC?
If there's a new ASIC, it's not going to be introduced on a lower-end model
-
If there's a new ASIC, it's not going to be introduced on a lower-end model
Moving the old ASIC to the budget tier and introducing the new one to differentiate the high end offerings makes sense, though.
-
Will there be a MegaZoom V ASIC?
If there's a new ASIC, it's not going to be introduced on a lower-end model
Maybe Keysight also went for a Xilinx Zync based solution. The Megazoom ASICs are seriously outdated / limited anyway so it makes no sense to bring any scope onto the market with a Megazoom ASIC unless it is a new type which packs a serious amount of memory.
-
If they did go with a Zynq it wouldn't be the first product Keysight has that uses it.
-
Any news?
-
Any news?
Soon ;)
-
It's here..
May not have much time to play but will take it apart soon!
-
Great :-+ Any datasheet? I'm very curious to see whether Keysight has come up with something radically new (with more memory) which can stand up against the cheaper Chinese scopes or it is just meh.
Either way it is interesting Keysight went for a shared controls.
-
Great :-+ Any datasheet? I'm very curious to see whether Keysight has come up with something radically new which can stand up against the cheaper Chinese scopes or it is just meh.
Either way it is interesting Keysight went for a shared controls.
No time to look atm - they've done it similar to R&S, RGB LED to indicate which channel the vertical controls control
-
If it is based on the same hardware / software as the 2 channel 1000X series, then I don't expect many changes except for the 1.5 additional channels (4 channel VS 2.5 channel)
The good news will be that the 1000X 2 channel software was hacked by one of the EEVBLOG forum members, so I assume the same hack technique can be applied to the new scope
Probably Keysight will release a single firmware for both 2 and 4 channel scopes on the 1000X series, like what they do with 2000X and 3000X
-
Great :-+ Any datasheet? I'm very curious to see whether Keysight has come up with something radically new which can stand up against the cheaper Chinese scopes or it is just meh.
Either way it is interesting Keysight went for a shared controls.
No time to look atm - they've done it similar to R&S, RGB LED to indicate which channel the vertical controls control
Looks pretty but seeing that KS went to the trouble of magnifying the 1mV/DIV to 500µV/DIV, I wonder what the actual noise performance is? Over the past year, I found the LeCroy HDO, R&S RTB / RTC / RTE / RTM all hover between 50µVrms and 100µVrms on actual testing with BWL on. Much to my surprise, none matched the SDS1104X-E at under 30µVrms despite all "Low Noise" claims in all their brochures.
When you have time, perhaps you can post µVrms and µVpp readings with BWL off and BWL on? Screenshots be great too! Thanks.
-
... I'm very curious to see whether Keysight has come up with something radically new (with more memory) which can stand up against the cheaper Chinese scopes or it is just meh.
Just 1Mpts!
-
... I'm very curious to see whether Keysight has come up with something radically new (with more memory) which can stand up against the cheaper Chinese scopes or it is just meh.
Just 1Mpts!
Which turns into 250kpts in a regular usage scenario with all channels on. I was hoping Keysight would finally see the light but I guess they are not. :'( IMHO Keysight makes great oscilloscopes but the memory is just too small.
-
Which turns into 250kpts in a regular usage scenario with all channels on. I was hoping Keysight would finally see the light but I guess they are not. :'( IMHO Keysight makes great oscilloscopes but the memory is just too small.
I assume it's a limitation of the ASIC.
-
Which turns into 250kpts in a regular usage scenario with all channels on. I was hoping Keysight would finally see the light but I guess they are not. :'( IMHO Keysight makes great oscilloscopes but the memory is just too small.
I assume it's a limitation of the ASIC.
There probably is more memory in the ASIC but if Keysight enables that they'll eat into the higher end models. But still the ASIC is likely to be limited to 4Mpts (1MPts in real usage) so -sadly- that still is nothing compared to the competition.
-
... I'm very curious to see whether Keysight has come up with something radically new (with more memory) which can stand up against the cheaper Chinese scopes or it is just meh.
Just 1Mpts!
Which turns into 250kpts in a regular usage scenario with all channels on. I was hoping Keysight would finally see the light but I guess they are not. :'( IMHO Keysight makes great oscilloscopes but the memory is just too small.
I am not completely sure but Keysight specifications are per channel. I assume it is 1Mpts per channel. On the 1000X series, it is 2GSa/s per channel with both channels active. If they added a second ASIC for the additional 2 channels, then probably they can maintain 2GSa/s per channel with all 4 channels active
-
AFAIK Keysight specs are always total memory shared by 2 channels and double buffering so the actual amount of memory available in run mode is 1/4 of the specified memory. Add digital channels to a model and you are left with 1/8. This information is usually in the fine print of the datasheet or user manual.
-
There probably is more memory in the ASIC but if Keysight enables that they'll eat into the higher end models. But still the ASIC is likely to be limited to 4Mpts (1MPts in real usage) so -sadly- that still is nothing compared to the competition.
To be fair, if I look at the responsiveness of the GUIs of the competition they're still miles ahead. Bigger numbers aren't everything. I'm not too impressed by the new Rigol, for example. It seems very laggy in comparison to the ASIC of the Keysight oscilloscopes.
-
... I'm very curious to see whether Keysight has come up with something radically new (with more memory) which can stand up against the cheaper Chinese scopes or it is just meh.
Just 1Mpts!
And according to the mouser website these scopes aren't exactly cheap:
DSOX1204A 2GS 70MHz: $998.00
DSOX1204G 2GS 70MHz: $1,204.00
DSOX1204A+D1200BW1A 2GS 100MHz: $1,348.00
DSOX1204G+D1200BW1A 2GS 100MHz: $1,554.00
DSOX1204A+D1200BW2A 2GS 200MHz: $1,698.00
DSOX1204G+D1200BW2A 2GS 200MHz: $1,904.00
:-DD
-
I didn't expect too much, but nice to see they pushed the bandwidth up to the maximum. They should do that with the sample rate too (2.5), because the BLT is capable and I bet they used the exactly same or very-very similar card. So the 1/4 of the specified memory is a real expectation.
Yes the GUI is very good, it was no brainer for me which want to choose when I compared to a rigol.. only two negatives: switching the trigger from one ch to another is a lot of button push, also reaching the segmented memories. These two what I use frequently, should be easier to reach, but heey I can't complain.
Btw love to see a serial output and some macros from the frontend, but probably nothing too exacting happening there.
-
Internal shots.
A quick side-by-side comparison with the 1102 :
Bus and Ref buttons have moved to the Analyse menu
Has LAN interface, but remote front panel not yet implemented. PHY is SMC 8710A, so 100Mbit
Ridiculously long boot time - 55 secs v.21 on the 1102 - maybe still some debug code in there ? FW is V2.00
No WinCE license sticker on the back
Low-level noise performance identical.
Ref and FFT have dedicated size and position knobs, presumably to match the knob layout of the 1102. Not sire I like that these are on the left.
Still has external trig input on the back, but can no longer display it, which is as expected
The BLT board has gone to rev 3 but looks pretty much identical at first glance, apart from J104 now being fitted.
Might be interesting to swap into the 1102G...
-
The fan lead is now routed though the main PCB - when cold it runs at 6.4v, but this increases when some warm air is blown over the board.
It is noticeably quieter than the 1102.
One criticism of this is that they should run the fan up to full speed briefly at startup to overcome any bearing stickiness as it gets full of dust with age.
-
It runs Linux
-
U-Boot 2010.03.01.B.01-AUTOINC+48ceb543a1-tlo-production-release (Oct 12 2018 - 09:07:13)Keysight P500
CPU: SPEAr600
DRAM: 128 MiB
Flash: 512 KiB
NAND: fsmc-ecc1 linux oob ef da 256 MiB
In: serial
Out: serial
Err: serial
SerNum:serial number not programmed
Chip: BD Board Rev: 4
Last address in xloader: 8d4421a3
Power-on reset
Setting watch dog timer to 45 seconds, count 0x29209080
Press space to stop autoboot: 0
NAND read: device 0 offset 0x180000, size 0x400000
4194304 bytes read: OK
## Checking Image at 02000000 ...
Legacy image found
Image Name: linux
Created: 2018-10-12 8:46:27 UTC
Image Type: ARM Linux Kernel Image (uncompressed)
Data Size: 3625377 Bytes = 3.5 MiB
Load Address: 02000000
Entry Point: 02000040
Verifying Checksum ... OK
## Booting kernel from Legacy Image at 02000000 ...
Image Name: linux
Created: 2018-10-12 8:46:27 UTC
Image Type: ARM Linux Kernel Image (uncompressed)
Data Size: 3625377 Bytes = 3.5 MiB
Load Address: 02000000
Entry Point: 02000040
XIP Kernel Image ... OK
OK
Starting kernel ...
Uncompressing Linux... done, booting the kernel.
tlo 2.4 k-dx1204g-67301 ttyAMA0
k-dx1204g-67301 login:
-
This does raise questions about whether they will move any of the other current products to Linux, or reserve it for new models only - seems like it would be a lot of work to test all the old hardware.
ISTR one issue is that when MS discontinued CE Embedded, KS wouldn't be allowed to just buy more licenses to ship existing products.
Let's hope they can get that boot time under better control - 55 secs is just unacceptable IMO.
and no, login with root/root doesnt work..!
-
Linux is a big twist and looks like they populated the missing connector on the BLT module for the 2 extra channel. Any chance you have a copy of the fw? or maybe u'd make one with jtag?
Just speculation, but looking at the fw number (FW2.00) maybe they roll it out to the 2ch models too. We have 1.10 now. A quick look at the frontend and it's kind of the same, the layout differ a little bit at the top, around the LMH6552, but nothing else. Should be everything compatible.
Choosing the SMC 8710A was obvious, (common) u-boot already supports it. Also included in the source code for the spear600 EB.
login with root/root doesnt work..!
what about skywalker? like on the 2000/3000 series.
-
Skywalker was working with 2ch. 1000x scopes. Maybe it's possible change file what contains root password while linux installation.
BLT board looks identical like in 2ch version. Unmounted connector J104 in 2ch version was handling Ethernet and extra USB. I don't expect that they change BLT board pinout.
-
There is a 10 pin header on the main board that has the serial lines (tx on 3 and rx on 1 from memory), not looked at what else is on there.
-
How is the UI responsiveness compared to the Windows CE based scopes?
-
How is the UI responsiveness compared to the Windows CE based scopes?
Not played extensively but didn't see any obvious difference. I doubt the UI has much to do with the OS.
-
Now we know what OS Keysight will be using going forward on what will likely be many products. It isn't the first time they have used linux, I wonder if MS will regret their silly license rules. The boot time on devices like the 3446x series of bench meters was also a minute when they were released with WinCE, thankfully they were able to optimize it to half that a few months after release.
-
Link to datasheet: http://www.farnell.com/datasheets/2714341.pdf (http://www.farnell.com/datasheets/2714341.pdf)
-
Let's hope they can get that boot time under better control - 55 secs is just unacceptable IMO.
It's pretty well known how to make Linux boot through to running app code in just a few seconds (talk from linux.conf.au 2018 on booting to usable in < 2 seconds (https://www.youtube.com/watch?v=d1v_o7ww43A)).
Hopefully Keysight have some good Linux people (I know *HP* had some excellent ones, sadly all retired or left) to do the Linux side of things, although the scope app itself was often not that quick to start and may well be where most of the time is spent.
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
-
What a twist, the new scope runs Linux. I guess that makes sense, with WinCE 6 reaching EOL, and STMicro advertises that their SPEAr600 supports Linux anyway. I wonder how Linux takes advantage of the processor's two cores (to my knowledge, WinCE 6 didn't support multi-core processing).
I guess that means we'll need a different approach to run Doom on Keysight's newer lineup of scopes. ;)
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
Linux doesn't have to be heavier. That is nonsense. Also the boot time of a Linux system is defined by the speed of the flash being used. Put a Sata SSD in an embedded Linux system and it wil start in seconds. Besides that the oscilloscope likely does a quick self test during starting.
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
Linux doesn't have to be heavier. That is nonsense. Also the boot time of a Linux system is defined by the speed of the flash being used. Put a Sata SSD in an embedded Linux system and it wil start in seconds. Besides that the oscilloscope likely does a quick self test during starting.
Ideally sure, but there's plenty of ways a default normal Linux system is slower to boot than it needs to be for fixed hardware configurations (scanning busses that will never be used etc.)
*Done well* there should be minimal practical differences between WinCE and Linux, but if you just use a generic kernel & image (or without systemd which for all its faults is normally faster) it probably is slower.
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
Linux doesn't have to be heavier. That is nonsense. Also the boot time of a Linux system is defined by the speed of the flash being used. Put a Sata SSD in an embedded Linux system and it wil start in seconds. Besides that the oscilloscope likely does a quick self test during starting.
What I meant is that PROBABLY the specific Linux distro in use is heavier than Windows CE on the SPEAR600 that the Keysight scope uses.
I know it can be made to boot in 10 seconds like the R&S or the Micsig TO1000 scopes, probably by using some battery backed memory to keep persistent code and data in memory saved for the next boot. I wonder if R&S scope takes longer to boot after a firmware update or the update process per se stores certain code & data in persistent memory.
But if they decided to go with the 60 seconds boot time, then there must be hardware limitations on the current SPEAR600 + flash + ASIC interface to implement specific techniques like smartphones or tablets.
-
Daniel says there is an update coming soon to improve boot time.
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
Linux doesn't have to be heavier. That is nonsense. Also the boot time of a Linux system is defined by the speed of the flash being used. Put a Sata SSD in an embedded Linux system and it wil start in seconds. Besides that the oscilloscope likely does a quick self test during starting.
What I meant is that PROBABLY the specific Linux distro in use is heavier than Windows CE on the SPEAR600 that the Keysight scope uses.
I know it can be made to boot in 10 seconds like the R&S or the Micsig TO1000 scopes, probably by using some battery backed memory to keep persistent code and data in memory saved for the next boot. I wonder if R&S scope takes longer to boot after a firmware update or the update process per se stores certain code & data in persistent memory.
But if they decided to go with the 60 seconds boot time, then there must be hardware limitations on the current SPEAR600 + flash + ASIC interface to implement specific techniques like smartphones or tablets.
The R&S are always fast. They don't use linux either.
-
R&S uses freeRTOS or something similar (it's listed in the open source acknowledgement thing). The boot time is always fast, and i don't think it saves anything between boots to do it. Definitely nice to have it that quick but I think Linux is also a reasonable choice if optimised well - boot times aren't everything.
-
I still love my Agilent DSOX2002A introduced 8 years ago. Is there a successor coming?
On the other hand, Tektronix still has no successor for that MSO2000B from 2008 or so. And they still produce TDS3000 from 1998! https://uk.tek.com/oscilloscope/tds3000
OK, they changed the floppy drive for an USB port.
-
it appears they launched it today.
-
"Protocol Trigger & Decodes I²C, SPI, UART and CAN, LIN (optional)" Oh?
Sounds like they're dropping their 2000 series. In the 1000X flyer it says "Need more bandwidth, sampling rate, and analysis? Consider the InfiniiVision 3000T X-Series"
-
"Protocol Trigger & Decodes I²C, SPI, UART and CAN, LIN (optional)" Oh?
Sounds like they're dropping their 2000 series. In the 1000X flyer it says "Need more bandwidth, sampling rate, and analysis? Consider the InfiniiVision 3000T X-Series"
The landing page for 1000x still shows 2000x as need more... Although the only thing the 1000x now don't have the 2000 do is 8ch MSO... I wouldn't blame them if the did drop the 2000.
Edit: My bad, the 2000 also has up to 4x the wfms/s
-
Like I said earlier, it wouldn't surprise me to see a new arch show up for the rest of the X-line within a year or two. But who knows. I've been wrong before, yhey might have decided that they want to thin out the linup also. :)
I could see a split like this being a thing, otherwise:
x1000 old arch, no probe compensation
x2000 new arch, +DSO, still no probe comp
x3000 new arch, +probe comp, +50 Ohm.
x4000 new arch, +bigger screen + really deep mem, +additional arb port.
x6000 new arch, up to 8 channels.
-
Well, I still do not see much reasons to buy this scope instead of DSOX2000 series, when it is based on the same ASIC. Maybe if you need a little 4 channel scope. 1000x has only 150Vrms input voltage. DSOX2000 has 300Vrms probably, the datasheet is a bit unclear. Yes the 1000x is a lot cheaper.
-
Like I said earlier, it wouldn't surprise me to see a new arch show up for the rest of the X-line within a year or two. But who knows. I've been wrong before, yhey might have decided that they want to thin out the linup also. :)
I could see a split like this being a thing, otherwise:
x1000 old arch, no probe compensation
x2000 new arch, +DSO, still no probe comp
x3000 new arch, +probe comp, +50 Ohm.
x4000 new arch, +bigger screen + really deep mem, +additional arb port.
x6000 new arch, up to 8 channels.
I assume by "probe comp" you're really meaning AutoProbe (at least the more limited version the 3/4/6 series has).
The list you have does roughly make sense, I'd probably say new 4k probably keeps at least a small bandwidth bump over 3k.
It's the 6k I'm not sure about, it's a wonderful scope (I love mine), but it's very close in many ways to the S-series, so it would need something special to keep it post-refresh, especially if the 4k was also being kept. One possibility might be to somehow allow the scope to be hooked up to a Windows PC via thunderbolt (external PCIe) to turn it into a full Windows-running scope when needed.
I'd also guess that we'd see the models released in the same order as they were for the current run, so a 2k & 3k first, then about a year later a 4k, and another year or two for the 6k, and as old as these models are, other than perhaps competing with the (lower end of the) new Tek 5k series at the high end, and some of the R&S units, I don't think the pressure is there to move as fast as they could.
Remember there's been no firmware update since late 2017 for most of these models.
-
We don't plan to drop the 2000 X-Series right now, but it definitely will make sense for a lot of people to go with the 1000X. A lot has changed in the last 8-10 years, so the 1000 X-Series makes more business sense today than it did in 2012 with the series was first launched.
The big draw for the 6000 X-Series is that it's a cheap way to get to 6 GHz. Most people in that space will opt for the S-Series, though.
-
Well, I still do not see much reasons to buy this scope instead of DSOX2000 series, when it is based on the same ASIC. Maybe if you need a little 4 channel scope. 1000x has only 150Vrms input voltage. DSOX2000 has 300Vrms probably, the datasheet is a bit unclear. Yes the 1000x is a lot cheaper.
There you go. You get almost the same for a lot less. Unless you require those differences or money isn't an object, the 1000X series is hard to beat.
-
It surprises me we haven't seen this one pop up in reviews or teardowns yet, as they seem to be for sale already.
-
It surprises me we haven't seen this one pop up in reviews or teardowns yet, as they seem to be for sale already.
Sorry, too busy at the moment !
-
https://twitter.com/eevblog/status/1088626285469782016
Looks like Dave has this scope.
-
Has anyone hacked this scope yet? Did some searching but have not found anything.
Considering the DSO5000 from Rigol or the 1204X if it can be hacked to 200Mhz and the Protocal Analyzer opened up.
-
I own the EDUX1002G, then purchased MSO5074 from Rigol... for me, it is not near the usability of the Keysight... it can be hacked to have tons of features, but it is slow and the serial protocol decoding a toy. I am about to return it
The EDUX1002G hacked and with some hardware mods can have 2.5 channels and it is a pleasure and fun to use. The Rigol is giving me headaches, you don't know what you are getting from the scope, you are always wondering if what is showing is accurate or not.
-
I can hardly wait to receive mine :popcorn: @Keysight DanielBogdanoff do you know wether the contest scopes are fully kitted out or do I need to contact a local distributor to buy serial decoding (SPI, I2C and UART)?
-
The Rigol is giving me headaches, you don't know what you are getting from the scope, you are always wondering if what is showing is accurate or not.
I said that shortly after rigol cheapies flooded the market and contaminated brains on this forum. And i provided hands on evidence in my "Project Yaigol" forensic post (which can be found by searching this forum). I said people will be puppy happy until this rigol junk will start affecting their performance by dragging them back, when instead of working on the projects they will be working on fighting their rigol junk scope. For which i was beaten by rigol fan boys. I am glad my predictions were correct.
-
Yes, my Rigol MSO5074 is on its way back to where it came from, for good. No more Yaigol for me.
-
Firmware for the DSOX1204x series is now posted on the Keysight website. As this scope runs linux it should be interesting to take a peak at.
Version 2.01
https://www.keysight.com/main/software.jspx?ckey=3023906&id=3023906&nid=-32110.1258590&lc=eng&cc=US (https://www.keysight.com/main/software.jspx?ckey=3023906&id=3023906&nid=-32110.1258590&lc=eng&cc=US)
Release notes:
Enhancements
- The following Frequency Response Analysis (FRA) enhancements were added:
o The sweep frequency range is no longer limited to decade values – it can now be
set to any frequency within the WaveGen’s limits.
o The “points per decade” setting has been changed to “total points” for ease of use.
o FRA Chart:
Minor gridlines for the horizontal axis are now shown to improve readiblity
of the plot.
The chart’s horizontal display range is now independent of the Start/Stop
Frequency setting in the Setup Menu.
- Boot-up time was reduced.
- Improved the color constrast for Channel 3 and Channel 4’s waveform persistence.
- Miscellaneous LAN enhancements.
Bug Fixes
- Fixed missing “Time Reference” softkey.
- Fixed an unused softkey briefly appearing in the Pattern Trigger menu.
- Fixed an issue of the inverted channel mode not displaying the waveform correctly in
Averaging Mode.
-
- Boot-up time was reduced.
Was 58 seconds, now 32 secs
1102G : 21 secs
-
I am curious about how much hardware it shares with the 2-channel 1000X scopes... no teardown yet?
-
I am curious about how much hardware it shares with the 2-channel 1000X scopes... no teardown yet?
It's been surprisingly quiet.
-
Well, Mike has one and he has posted internal pics earlier in this thread. It is based on the same hardware platform as the 1002X series scopes, very similar platform. I don't know if anyone has done a video tera down, though.
-
Please excuse my ignorance, but what does the acronym BLT mean for that board?
-
Internal shots.
A quick side-by-side comparison with the 1102 :
Bus and Ref buttons have moved to the Analyse menu
Has LAN interface, but remote front panel not yet implemented. PHY is SMC 8710A, so 100Mbit
Ridiculously long boot time - 55 secs v.21 on the 1102 - maybe still some debug code in there ? FW is V2.00
No WinCE license sticker on the back
Low-level noise performance identical.
Ref and FFT have dedicated size and position knobs, presumably to match the knob layout of the 1102. Not sire I like that these are on the left.
Still has external trig input on the back, but can no longer display it, which is as expected
The BLT board has gone to rev 3 but looks pretty much identical at first glance, apart from J104 now being fitted.
Might be interesting to swap into the 1102G...
I found an interesting article written by Daniel Bogdanoff about the making of an ASIC. There is a picture of the same one you show in the article (revision 2 instead of 3) but it does not have the heat sinks on the chips. In case you want to see them.
Thanks for a great read, Daniel!
https://www.electronicdesign.com/test-measurement/making-asic-secret-building-good-cheap-oscilloscope (https://www.electronicdesign.com/test-measurement/making-asic-secret-building-good-cheap-oscilloscope)
(https://www.electronicdesign.com/sites/electronicdesign.com/files/bog_asic_Fig2.jpg)
-
I am curious about how much hardware it shares with the 2-channel 1000X scopes... no teardown yet?
Dave did a teardown of the DSOX1102G in March 2017. You can see in the chassis holes the future planning for the DSOX1204G. Comparing Mikes pictures of the 1204 with the EEVblog teardown of the 1102, the Main boards are labeled different. DSOx1000AG vs DSOX1004AG. It seems the 1004 board is a bit more complex.
https://www.youtube.com/watch?v=9KcOQsVxtoU (https://www.youtube.com/watch?v=9KcOQsVxtoU)
-
Please excuse my ignorance, but what does the acronym BLT mean for that board?
Bacon, lettuce and tomato
-
Firmware for the DSOX1204x series is now posted on the Keysight website. As this scope runs linux it should be interesting to take a peak at.
Shouldn't be mission impossible to devise workarounds... ;)
The .ksx is a initrnamfs file which can be easily opened with 7zip decompressor.
-
Would someone who owns or uses a 4ch 1000X confirm if its automatic measurements also include statistics of these measurements, like min, max, current, mean, std dev, count etc? This was mentioned in the 1000X datasheet as "Measurements continuously updated with statistics" on page 19 but not mentioned at all in the User's Guide. I also have not noticed it from any demos, reviews or YouTubes I've seen.
Thank you.
-
Upon a quick check it looks like no, but I could be missing it somewhere..
-
Thank you for checking.
If stats were there, they would be somewhere buried in the measurement menu tree. I checked the user's guides of the 1000X, 2000X and 3000X and it is only referenced in the 3000X, and it is actually there on the scope. Funny it would be in the datasheets but not in the manuals or the machines.
Upon a quick check it looks like no, but I could be missing it somewhere..
-
Firmware for the DSOX1204x series is now posted on the Keysight website. As this scope runs linux it should be interesting to take a peak at.
Shouldn't be mission impossible to devise workarounds... ;)
The .ksx is a initrnamfs file which can be easily opened with 7zip decompressor.
There is a 128 byte sw-description.sig file. Anyone knows how it is generated? It is different on each install package (.ksx)
-
Lost (TV Series) Easter Egg on DSOX1204 and EDUX1052 scopes:
https://www.eevblog.com/forum/testgear/time-travelling-on-my-keysight-edux1002g/msg1280254/#msg1280254 (https://www.eevblog.com/forum/testgear/time-travelling-on-my-keysight-edux1002g/msg1280254/#msg1280254)
-
I own the EDUX1002G, then purchased MSO5074 from Rigol... for me, it is not near the usability of the Keysight... it can be hacked to have tons of features, but it is slow and the serial protocol decoding a toy. I am about to return it
The EDUX1002G hacked and with some hardware mods can have 2.5 channels and it is a pleasure and fun to use. The Rigol is giving me headaches, you don't know what you are getting from the scope, you are always wondering if what is showing is accurate or not.
Do you have specific example ?
It is not so easy to have scope on loan without losing money and I am open minded so if it is possible to have examples. I am interested.
I am also interested in what you criticize for the serial decoding on the Rigol ?
Before receiving a shitstorm on my head, I specify that I am really interested to know in what the Keysight is superior on these aspects.
I am not here to support Rigol, just to have examples from real life.
Thank you :)
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
I checked out the otherwise great new black scopes in the 1000x series, and the boot time is still (with latest 2020-06 firmware) awfully much slower than the old versions with Windows. I do not blame HP as they probably had no choice due to Microsoft's decisions, but half a minute boot to operation with the Linux scopes is not very good. I wish it had been 15-20 seconds like the old scopes. Much more acceptable.
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
Linux doesn't have to be heavier. That is nonsense. Also the boot time of a Linux system is defined by the speed of the flash being used. Put a Sata SSD in an embedded Linux system and it wil start in seconds. Besides that the oscilloscope likely does a quick self test during starting.
Not nonsens. You do not know any clue at all what you talk about. Your the one talking total nonsense and bull shit.
1. A low end scope normally do not have the same powerful hardware as a desktop PC.
2. It is to expensive (for this low end market) to but a desktop SSD in the scopes. Totally unrealistic. The Windows scopes did OK boot time with the current (non SSD) storage in the scopes.
4. It is not enough to boot just the OS. You also need to load the scope application. No "flash of light" here with the Linux scopes; more slow like molasses.
5. The old white Windows scopes also did self test during power on (like all normal scopes) during boot time. Nothing new for the Linux scopes here (besides being slower).
-
I think Spear600 is old compared to new CPUs running faster and with more cores and Probably the old Windows CE was a better fit. That is why I asked about the responsiveness as it seems Linux is heavier and the same reason for it to take 1 min to boot might impact on the performance of the scope app
Linux doesn't have to be heavier. That is nonsense. Also the boot time of a Linux system is defined by the speed of the flash being used. Put a Sata SSD in an embedded Linux system and it wil start in seconds. Besides that the oscilloscope likely does a quick self test during starting.
What I meant is that PROBABLY the specific Linux distro in use is heavier than Windows CE on the SPEAR600 that the Keysight scope uses.
I know it can be made to boot in 10 seconds like the R&S or the Micsig TO1000 scopes, probably by using some battery backed memory to keep persistent code and data in memory saved for the next boot. I wonder if R&S scope takes longer to boot after a firmware update or the update process per se stores certain code & data in persistent memory.
But if they decided to go with the 60 seconds boot time, then there must be hardware limitations on the current SPEAR600 + flash + ASIC interface to implement specific techniques like smartphones or tablets.
The R&S are always fast. They don't use linux either.
Yes, I prefer scopes like R&S which boots much faster.
R&S uses freeRTOS or something similar (it's listed in the open source acknowledgement thing). The boot time is always fast, and i don't think it saves anything between boots to do it. Definitely nice to have it that quick but I think Linux is also a reasonable choice if optimised well - boot times aren't everything.
Boot time is not everything, but it's clear that scopes with Linux is generally much slower to boot compared to scopes with more optimized systems. Might be some scopes who manages to boot fast with Linux (given more expensive hardware), but those I have seen has ben slow as molasses.
The good thing is that the "fatter" Linux OS on the black 1000-series at least seems to work as normal as the old white Windows scopes, one it's booted. That is the important thing for me (I really hate the slow user interface scopes from China and Taiwan).
Slow boot time is irritating (especially as older versions has been better), but it is not critical. Quick GUI and responsiveness are the critical things.
-
I own the EDUX1002G, then purchased MSO5074 from Rigol... for me, it is not near the usability of the Keysight... it can be hacked to have tons of features, but it is slow and the serial protocol decoding a toy. I am about to return it
The EDUX1002G hacked and with some hardware mods can have 2.5 channels and it is a pleasure and fun to use. The Rigol is giving me headaches, you don't know what you are getting from the scope, you are always wondering if what is showing is accurate or not.
Looked at the Rigol scopes as well, and I can only agree with your assessment. Really nice hardware (nothing to complain at all) and really low price, but not very polished or optimized software wise.
And ugly as hell! :-DD
-
I own the EDUX1002G, then purchased MSO5074 from Rigol... for me, it is not near the usability of the Keysight... it can be hacked to have tons of features, but it is slow and the serial protocol decoding a toy. I am about to return it
The EDUX1002G hacked and with some hardware mods can have 2.5 channels and it is a pleasure and fun to use. The Rigol is giving me headaches, you don't know what you are getting from the scope, you are always wondering if what is showing is accurate or not.
Do you have specific example ?
It is not so easy to have scope on loan without losing money and I am open minded so if it is possible to have examples. I am interested.
I am also interested in what you criticize for the serial decoding on the Rigol ?
Before receiving a shitstorm on my head, I specify that I am really interested to know in what the Keysight is superior on these aspects.
I am not here to support Rigol, just to have examples from real life.
Thank you :)
TK managed to achieve some 50000 triggers per second on Keysight 1000 triggering on 50 MHz SPI signal.
At first try Rigol MSO5000 couldn't go faster than some 3000. But after some discussion, and by setting the scope to Auto memory management (that is what Keysight does anyways) he managed to get it to capture 50000 triggers too..
His opinion of accuracy I have no idea where it comes from. It can only be more accurate on many measurements by virtue of 100x more memory and full memory measurements.
Problem is that 50000 SPI triggers on KS 1000 is USELESS, mostly. It can record maximum 50 segments, so you can capture first 50 of 50000 segments and the rest is lost. OTOH, MSO5000/7000/8000 and Siglent can actually capture all 50000 captures and save each for
later analysis. And I don't know about you but I can't read 50000 decodes per second by reading it from the screen in real time anyways. Getting slow with age, sorry..
Not to mention literally 10 times more analysis options on Rigol and Siglent. And that you can capture first and decode later without need to recapture because you forgot to enable decode or a setting was wrong.
There are many people that are seduced by idea that KS 1000 series is somehow smaller and cheaper, but basically almost equally powerful as 3000 series. They are nothing alike. KS 1000 is deliberately severely limited not to step on 2000 series which is also very limited subset of 3000A series. Cheapest scope that is very nice from Keysight is 3000A series, and a 3000T is a step up, being basically 4000 with smaller touch screen and only one ch of siggen. 3000/3000T have less memory and smaller screen than contemporary competition, but they are capable machines that hold their ground still. Not good choice for deep memory types of tasks, but probably best G.P. scope currently. My opinion.
1000 and 2000 are just pathetic, because 500 USD Siglents are so much more powerful. Compared with MSO5000 and Siglent SDS2000X+ they are just sad... That being said, that is if you use scope as a digital scope should be used, by using built in measurements, analysys, decodes etc..
If you use scope in a same manner you would use analog scope, counting divisions on screen and occasional use of cursors and such, and want scope to to be optimized for button twiddling speed, than those 1000 and 2000 series Keysights are going to be just fine. That is literally their target market: big brand, deliberately low on features (marketed as: easy to use), and made NOT to feel like digital scope as possible (again marketed to analog scope replacement crowd). If you want that (nothing wrong with it) and if you are willing to pay for it, perfect choice.
It is equivalent of Ferrari branded Fiat 500 sized car with all the visual Ferrari stuff, special exhaust that makes it sound angry. And whole 60 HP.
So in a parking lot full of Fiats, you will be noticed. Fact that it is in fact slower than Fiat 500 Abarth, makes no difference. Because you're not going to race it. You will use to go to grocery store in style every day, really. And that is fine, really. Nothing wrong with it.
Until you get delusional that since your car says Ferrari on the back, that means it is faster, better or in any way better value that Honda Civic type R or Golf GTI or whatnot, that costs the same.
It's not. That doesn't mean that for your (limited?) use case you will not enjoy every minute of driving it and be very happy with your purchase.
Because it is PERFECT for your priorities as a buyer.
-
Sinisa, this should be a sticky!
-
I own the EDUX1002G, then purchased MSO5074 from Rigol... for me, it is not near the usability of the Keysight... it can be hacked to have tons of features, but it is slow and the serial protocol decoding a toy. I am about to return it
The EDUX1002G hacked and with some hardware mods can have 2.5 channels and it is a pleasure and fun to use. The Rigol is giving me headaches, you don't know what you are getting from the scope, you are always wondering if what is showing is accurate or not.
Do you have specific example ?
It is not so easy to have scope on loan without losing money and I am open minded so if it is possible to have examples. I am interested.
I am also interested in what you criticize for the serial decoding on the Rigol ?
Before receiving a shitstorm on my head, I specify that I am really interested to know in what the Keysight is superior on these aspects.
I am not here to support Rigol, just to have examples from real life.
Thank you :)
TK managed to achieve some 50000 triggers per second on Keysight 1000 triggering on 50 MHz SPI signal.
At first try Rigol MSO5000 couldn't go faster than some 3000. But after some discussion, and by setting the scope to Auto memory management (that is what Keysight does anyways) he managed to get it to capture 50000 triggers too..
His opinion of accuracy I have no idea where it comes from. It can only be more accurate on many measurements by virtue of 100x more memory and full memory measurements.
Problem is that 50000 SPI triggers on KS 1000 is USELESS, mostly. It can record maximum 50 segments, so you can capture first 50 of 50000 segments and the rest is lost. OTOH, MSO5000/7000/8000 and Siglent can actually capture all 50000 captures and save each for
later analysis. And I don't know about you but I can't read 50000 decodes per second by reading it from the screen in real time anyways. Getting slow with age, sorry..
Not to mention literally 10 times more analysis options on Rigol and Siglent. And that you can capture first and decode later without need to recapture because you forgot to enable decode or a setting was wrong.
There are many people that are seduced by idea that KS 1000 series is somehow smaller and cheaper, but basically almost equally powerful as 3000 series. They are nothing alike. KS 1000 is deliberately severely limited not to step on 2000 series which is also very limited subset of 3000A series. Cheapest scope that is very nice from Keysight is 3000A series, and a 3000T is a step up, being basically 4000 with smaller touch screen and only one ch of siggen. 3000/3000T have less memory and smaller screen than contemporary competition, but they are capable machines that hold their ground still. Not good choice for deep memory types of tasks, but probably best G.P. scope currently. My opinion.
1000 and 2000 are just pathetic, because 500 USD Siglents are so much more powerful. Compared with MSO5000 and Siglent SDS2000X+ they are just sad... That being said, that is if you use scope as a digital scope should be used, by using built in measurements, analysys, decodes etc..
If you use scope in a same manner you would use analog scope, counting divisions on screen and occasional use of cursors and such, and want scope to to be optimized for button twiddling speed, than those 1000 and 2000 series Keysights are going to be just fine. That is literally their target market: big brand, deliberately low on features (marketed as: easy to use), and made NOT to feel like digital scope as possible (again marketed to analog scope replacement crowd). If you want that (nothing wrong with it) and if you are willing to pay for it, perfect choice.
It is equivalent of Ferrari branded Fiat 500 sized car with all the visual Ferrari stuff, special exhaust that makes it sound angry. And whole 60 HP.
So in a parking lot full of Fiats, you will be noticed. Fact that it is in fact slower than Fiat 500 Abarth, makes no difference. Because you're not going to race it. You will use to go to grocery store in style every day, really. And that is fine, really. Nothing wrong with it.
Until you get delusional that since your car says Ferrari on the back, that means it is faster, better or in any way better value that Honda Civic type R or Golf GTI or whatnot, that costs the same.
It's not. That doesn't mean that for your (limited?) use case you will not enjoy every minute of driving it and be very happy with your purchase.
Because it is PERFECT for your priorities as a buyer.
You are missing one very important point, the idea is not to capture all 50000 packets with a scope to do an analysis. A logic analyzer is better suited for such capture and analysis. The point I was trying to make is that by triggering 50000 times, it can capture a specific event and the Siglent will probably miss it. That is the whole point of using the protocol analyzer on the scopes. My SPI test setup has a user button enabled that sends a specific packet 0x3F when pressed, and 0x37 when not pressed. The KS gets it almost on every click, but the siglent misses a lot of them, the SDS1104X-E that I tested missed like 80-90% of them. Even the GDS-1054B can trigger on 90-100% of the time, the siglent is very slow handling serial decoding and triggering.
-
You are missing one very important point, the idea is not to capture all 50000 packets with a scope to do an analysis. A logic analyzer is better suited for such capture and analysis. The point I was trying to make is that by triggering 50000 times, it can capture a specific event and the Siglent will probably miss it. That is the whole point of using the protocol analyzer on the scopes. My SPI test setup has a user button enabled that sends a specific packet 0x3F when pressed, and 0x37 when not pressed. The KS gets it almost on every click, but the siglent misses a lot of them, the SDS1104X-E that I tested missed like 80-90% of them. Even the GDS-1054B can trigger on 90-100% of the time, the siglent is very slow handling serial decoding and triggering.
When scope is armed and waiting for trigger (in Normal not Auto mode), it will trigger instantaneously. It is literally waiting to pull the trigger. It is retriggering that is questionable, once it triggers and finishes the sweep, it will take time to rearm. That rearm time is very fast on Keysight, also very fast on new Rigols , and a bit slower on Siglent. And because of it, Keysight might trigger on every packet and Siglent might not be ready for next one and miss it...
For that kind of use you don't even need to have decoding enabled. You just trigger on SPI packet when it happens. You know what it shows already, you set the trigger. You will see visual confirmation of packets coming.
For that kind of work, i would press button few times, enable decode and go trough history to see what I got.
Also for button verification, I might connect button to one channel an trigger on it. And would be able to measure latency from button press/release and SPI packet...
As I said for interactive, analog like work Keysight is very good. But you can do so much more with MSO5000 or SDS2000X+.
If you're willing to learn new ways of accomplishing things, that is...
-
Thank you for checking.
If stats were there, they would be somewhere buried in the measurement menu tree. I checked the user's guides of the 1000X, 2000X and 3000X and it is only referenced in the 3000X, and it is actually there on the scope. Funny it would be in the datasheets but not in the manuals or the machines.
Statistics were also on the 2000X I checked a while back. Nothing in the manual but I suspect it got added in firmware but the manual was not updated.
-
You are missing one very important point, the idea is not to capture all 50000 packets with a scope to do an analysis. A logic analyzer is better suited for such capture and analysis. The point I was trying to make is that by triggering 50000 times, it can capture a specific event and the Siglent will probably miss it. That is the whole point of using the protocol analyzer on the scopes. My SPI test setup has a user button enabled that sends a specific packet 0x3F when pressed, and 0x37 when not pressed. The KS gets it almost on every click, but the siglent misses a lot of them, the SDS1104X-E that I tested missed like 80-90% of them. Even the GDS-1054B can trigger on 90-100% of the time, the siglent is very slow handling serial decoding and triggering.
When scope is armed and waiting for trigger (in Normal not Auto mode), it will trigger instantaneously. It is literally waiting to pull the trigger. It is retriggering that is questionable, once it triggers and finishes the sweep, it will take time to rearm. That rearm time is very fast on Keysight, also very fast on new Rigols , and a bit slower on Siglent. And because of it, Keysight might trigger on every packet and Siglent might not be ready for next one and miss it...
Yes, that looks like a reasonable explanation.
-
Firmware for the DSOX1204x series is now posted on the Keysight website. As this scope runs linux it should be interesting to take a peak at.
Shouldn't be mission impossible to devise workarounds... ;)
The .ksx is a initrnamfs file which can be easily opened with 7zip decompressor.
There is a 128 byte sw-description.sig file. Anyone knows how it is generated? It is different on each install package (.ksx)
Sounds like a 1024-bit RSA signature to me. Must be generated using a private key at the factory and verified with the public key residing inside the scope during software setup.
-
Some additional info:
- Firmware package description file sw-description is signed with a 1024-bit RSA key. The signarure is in sw-description.sig file. It can be verified using the following parameters:
Algorithm: RSASSA_PKCS1
Hash: SHA256
- The assumption is the signature will be verified during firmware update by the scope using the corresponding public key stored in the previous firmware. A copy of the public key is also shipped in the firmware update (a copy is attached). Technically, every time the firmware is updated, a new public key can be shipped for the next firmware update. Signing the firmware update package prevents the firmware files from being altered.
- Root account is disabled.
- The working account is called 'tlouser'. The password for it is stored as SHA-512 hash with salt.
- The filesystem is stored compressed and is of read-only type
- NAND flash memory size is double of that on 1000X (256MiB vs 128MiB)
-
- The assumption is the signature will be verified during firmware update by the scope using the corresponding public key stored in the previous firmware. A copy of the public key is also shipped in the firmware update (a copy is attached). Technically, every time the firmware is updated, a new public key can be shipped for the next firmware update. Signing the firmware update package prevents the firmware files from being altered.
Correct BUT they usually keep always the same key because, if you would change key from package to package, when doing an update an user would be forced to execute all intermediary updates between his version and the current version. Also, KS would be forced to make available all those packages in their website, etc. permanently.
So, they can do a key change but only when there is a major justification for that and, then, everyone will have to go through that update package.
-
The story continues here. (https://www.eevblog.com/forum/testgear/the-fort-is-falling-liberating-keysight-dsox-1200-(linux-black)-series-scopes/msg5328674/#msg5328674)