New Rigol DS7000

[drhex]

Thanks @memeruiz, this worked like a charm!

[mindcrime]

Awesome, now I just need to go back through this entire thread, re-read it all, really digest / understand what's going on here, and then I can take a stab at hacking my MSO70204. 

I'm hoping it goes more smoothly, since I get the benefit of learning from the pains you guys have had already! 

[drhex]

And managed to brick the scope. Somehow it didn't persist settings through reboot, so tried to install the latest from the website (v00.01.02.00.06), no change. The patched appEntry from v01.01.02.00.06 doesn't work with that, patching it with the same patch bricks the scope - ssh doesn't come up when it gets stuck at appEntry, at least for me. Now I can't reinstall from the bootloader after pressing "single" on power on, says the package is invalid? Doesn't seem to use the USB really, at least not for an extended time before complaining about the package. Does anybody have an idea?

[drhex]

OK, not being able to reinstall from the hidden menue was apparently due to a too big USB stick, a 2GB one worked fine in the end. Does anybody have the v01.01.02.00.06 install file?
Patch for v00.01.02.00.06 attached, please use bspatch to apply.

[memeruiz]

OK, not being able to reinstall from the hidden menue was apparently due to a too big USB stick, a 2GB one worked fine in the end. Does anybody have the v01.01.02.00.06 install file?
Patch for v00.01.02.00.06 attached, please use bsdiff to apply.

Awesome @drhex that you could get your scope working again and upgraded.

I'm at a loss with my scope firmware version. It is not on any website for downloading. Apparently is newer than all the downloadable ones.

When you applied bspatch, did the patching gave any errors or warnings? (Like patch not applied or failed)

The last binary data chunk is tricky. I think the binary asm instruction may change from firmware to firmware because it is a branch execution and the branch address is different. (This is what I noticed comparing with the 5k). Not sure really.

I hope this doesn't mean a different binary patch is necessary for each version.

[drhex]

I created the patch from a diff based on the 01.01.02.00.06 version. The symbols to be patched are the same in 00.01.02.00.06.
bspatch didn't complain when applying the 01 patch to the 00 version (think it would only complain if the target file is too small). Bit of a pity that patchfinder.sh doesn't display the actual offsets in the file - may figure that out at some point, I just searched with a hex editor and patched accordingly.
My scope came with the 01 version, too - so your original patch was very handy! Don't think the last change is an issue as such as there are no hard addresses in there. Worked without any change for me (but I haven't really looked for side effects).

[wat]

hi drhex,
I had the same problem of the persistance of the settings after reboot,
this solves it: https://www.eevblog.com/forum/testgear/new-rigol-ds7000/msg1863522/#msg1863522

And managed to brick the scope. Somehow it didn't persist settings through reboot, so tried to install the latest from the website (v00.01.02.00.06), no change. The patched appEntry from v01.01.02.00.06 doesn't work with that, patching it with the same patch bricks the scope - ssh doesn't come up when it gets stuck at appEntry, at least for me. Now I can't reinstall from the bootloader after pressing "single" on power on, says the package is invalid? Doesn't seem to use the USB really, at least not for an extended time before complaining about the package. Does anybody have an idea?

[drhex]

Thanks for mentioning, I did work that out in the end - I did expect this setting to influence scope channel setup but not basics like the IP configuration. Seems it resets EVERYTHING.

[normi]

Thanks for mentioning, I did work that out in the end - I did expect this setting to influence scope channel setup but not basics like the IP configuration. Seems it resets EVERYTHING.

The default setting looks like its there to ensure that all scopes will have a standard setting, so if you report an issue support can be sure that they are comparing a scope with a known configuration. The missing feature is to allow you to add your custom setup as an option to boot the scope. You can manually  store and load your settings though, sometimes it takes a while to get all the triggering correct for a decode so I have to store the config to prevent having to repeat process.

[normi]

Worked without any change for me (but I haven't really looked for side effects).

Did you see jitter analysis enabled under measure > Analyze

[drhex]

Yes, that is there and can be enabled - it isn't showing any results though (which may be entirely to me not using it correctly).

[BarsMonster]

Also find attached the bspatch for MSO7000 firmware version 01.01.02.00.06 (which is not on the Rigol webpage yet for some reason). It could be that this patch also works for older firmware versions. Please check!

This week I got MSO7014.
I am glad to report that patch files from memeruiz worked with no issues. Factory firmware version was matching.
Thanks to everyone involved in enabling this path

[Sighound36]

Yes, that is there and can be enabled - it isn't showing any results though (which may be entirely to me not using it correctly).

drhex

This is the position were found ourselves in two years (with the 5000 we even had the eye feature accessable (after opening up) BUT non functioning as with the 7000 as well. We managed to stretch the BW upto around 3/4 of a Ghz as well thaks to Tv84

One of the reasons I went over to an MSO8000 at that time.

The 7000 offers some great features 10G/s, 500Mpt memory, some nifty analysis tools, hdmi output 10 inch screen etc.
The chaps in the main lab still use a couple of these, but not for really low noise measurments.

Nice to see interested still in this under rated scope.

[BarsMonster]

Yes, that is there and can be enabled - it isn't showing any results though (which may be entirely to me not using it correctly).

I can confirm that both jitter & eye are there, but not showing anything for any clock recovery methods.
Only histogram is there working as crude jitter tool.

[tv84]

We did some experiments here with a MSO5000.

[BarsMonster]

We did some experiments here with a MSO5000.

This is the position were found ourselves in two years

On your screenshots I see that something is working better, than for me...
I see measurements of Jitter. Is it not correct, or they are deficient on some way? What settings you used to get these?

[tv84]

I see measurements of Jitter. Is it not correct, or they are deficient on some way? What settings you used to get these?

I just forced the MSO5000 to 500 MHz BW. Sighound36 did the rest with the settings.

[normi]

I have the 5000 and the Jitter analysis works as far as I can see, I am not able to see an issue. The clock is recovered and the histogram works, and the table is filled with results. The Eye analysis does not work, and does not detect clock speed.

Why is it reported as not working, also is it possible that the 8000 has some additional files which are missing from the 5000 and 7000 which could be preventing the Eye analysis from working.

[Sighound36]

I would just check the results on the 5000, this is what we had and it wasn't accurate shall we say (or not what I would like call reliable, repeatable consistently good readings).

The 5000 and 8000 boards are quite different, suspect both a bandwidth limitation as well as processing power possibly memory limitation?

[BarsMonster]

I've noticed interesting sampling artifacts of my unlocked MSO7000.
Signal from LeoBodnar's pulser (SMA->BNC) is fed into channel 3 (could be any other), and we enable other channels to enable 2.5/5/10G sampling.

With 10G sampling - rise time is 474ps, 10ps std deviation.
With 5G sampling - rise time is 459ps (!!!),  15ps std deviation.
With 2.5G sampling - rise time is 458ps, and some huge artifacts / jitter (52ps std deviation).

It is probably expected that when you feed high speed signal into 2 sampling blocks - everything gets slower. So highest bandwidth is achieved with 5G sampling.
But cause of artifacts with 2.5G sampling is not clear.

Any opinion?

[CookieMonster]

And managed to brick the scope. Somehow it didn't persist settings through reboot, so tried to install the latest from the website (v00.01.02.00.06), no change. The patched appEntry from v01.01.02.00.06 doesn't work with that, patching it with the same patch bricks the scope - ssh doesn't come up when it gets stuck at appEntry, at least for me. Now I can't reinstall from the bootloader after pressing "single" on power on, says the package is invalid? Doesn't seem to use the USB really, at least not for an extended time before complaining about the package. Does anybody have an idea?

Hi All,

I currently have the same issue as @drhex.
Had the same FW version and used hin patch file. Unfortunately I'm now getting stuck at the boot screen with the loading bar.
I tried restoring the firmware with the hidden menu, but it sais 'Upgrading firmware failed, please check the package."

I am using a 2gb usb stick as suggested, but no luck. I tried FAT and FAT32 formats.
Any Ideas ?

[CookieMonster]

I managed to find and fix the problem.

I tapped into the UART port in the device to get the log while updating.

Code: [Select]

Firmware upgrading
(Re)start USB...
USB0:   USB EHCI 1.00
scanning bus 0 for devices... 1 USB Device(s) found
USB1:   USB EHCI 1.00
scanning bus 1 for devices... 3 USB Device(s) found
       scanning usb for storage devices... 1 Storage Device(s) found
reading ds5000update.gel
** Unable to read file ds5000update.gel **
** Unable to read file ds5000update.gel **
0x4000000:0x2000000:fw4uboot.sh:
File not found fw4uboot.sh
Error:tar 0x4000000 0x2000000 fw4uboot.sh
as one can see, u-boot is looking for a file called "ds5000update.gel" even though this is a 7000 and the update file is called "ds7000update.gel".
I renamed the file into "ds5000update.gel", the update then preceded flawlessly.

[rsjsouza]

Congratulations in finding the solution!
If you haven't skipped any prior firmware updates (as the name of the file might have been changed in the boot code), this clearly shows the amount of testing Rigol invested in the firmware update.

[tv84]

If you haven't skipped any prior firmware updates (as the name of the file might have been changed in the boot code), this clearly shows the amount of testing Rigol invested in the firmware update.

Let's say that Rigol never advertised the use of the bootloader update menu. Since they must be aware of that part internally, I don't think there is a major error here.

BTW, nice find. As I did all my initial researches in the 5000, never had noticed that "small" detail.

The name seems impossible to change via update since our opinion is that you cannot update the bootloader via a FW update package  (almost sure).

[CookieMonster]

Alrighty, the hack worked on the second try.

All options are now enable.
That being said, there is another problem.

I think you guys all have the MSO7000, whereas I only have the DS7000 (no function gen and no logic analyser). Yet, with the hack, these features are showing up on the UI.
I guess one would need a different patch to disable these options.

Does anyone have any experience with this ?