New Rigol DS1054Z oscilloscope

[alsetalokin4017]

Boot version 0.0.1.2
Build date Sep-11-2015 (lol @ that date)

With Boot Version 0.0.1.2 you _may_ have  the Freeze Bug that I discovered and that is discussed here:
https://www.eevblog.com/forum/testgear/rigol-ds1054z-freeze-up-bug/

[poorchava]

Yeah, I was able to lock the scope up, by playing with delayed time base, but I decided not to pursue warranty, as I'm not using delayed timebase much and I don't feel like losing it for a month os so before anybody resolves the claim.

If it's possible to update that on my own, then I will do that, but sending the scope away is a big no.

What I kinda fail to understand though, is why they have incorporated at least some part of hardware driver into bootloader that is designed to not be user-replaceable. Bad design IMO.

[Fungus]

What I kinda fail to understand though, is why they have incorporated at least some part of hardware driver into bootloader that is designed to not be user-replaceable.

Whatever the problem is, it's not caused by the bootloader.

Plenty of people have that particular bootloader and don't have the problem.

Bad design IMO.

Not really. If all the bootloader does is load the main firmware into memory, then ... what's to replace? 

"Bad design" would be system that can be bricked by users trying to update their bootloader when there's no need.

[Wirehead]

It would be nice if Rigol could just chime in and at least give an update of what they're working on atm. It's not that much of a deal to do so, and they have a huge following in hobbyist market, it would be the "nicest thing to do"® 

Wish they would give us full-screen FFT and X-Y mode as well though.. FFT can be done from memory anyway, so it should be possible to hide the original traces. Same for X-Y...

[Deuze]

Good topic, specially interested in the hack methods. 

[alsetalokin4017]

What I kinda fail to understand though, is why they have incorporated at least some part of hardware driver into bootloader that is designed to not be user-replaceable.

Whatever the problem is, it's not caused by the bootloader.

Plenty of people have that particular bootloader and don't have the problem.

Bad design IMO.

Not really. If all the bootloader does is load the main firmware into memory, then ... what's to replace? 

"Bad design" would be system that can be bricked by users trying to update their bootloader when there's no need.

Er.... no. Every case that I know about of a scope that can be "locked up" due to the Freeze Bug is running Boot Version 0.0.1.2, no matter what "firmware" the scope is running. As you can see poorchava has the latest "firmware" revision but the older Boot Version, and he is able to reproduce the Freeze Bug. It _may_ be the case that _some_ scopes with Boot Version 0.0.1.2 do not have the Freeze Bug, but I am still not convinced even of that much.
 
"Plenty of people have that particular bootloader and don't have the problem." .... the Poll indicates that nearly half the people who tried it were indeed able to get their scopes to freeze. But....
Many people did not seem to be able to follow the precise instructions, and some people have contaminated the Poll in that thread by reporting results from other models of the scope, such as 1074z and MSO versions and actual 1104z scopes. When the thread and poll was started we did not yet know how to display the full System Information screen that indicated which Boot Version was in use so we are missing data, but in every case that I am aware of where we _do_ have the full data, the scopes which freeze are running Boot Version 0.0.1.2.

And you and I have no idea just what the Boot Version actually does. A reasonable software design would have it just "load the main firmware from memory" but is anyone really confident that Rigol's software is "reasonable" at this point? 

Meanwhile -- if you know of a DS1054Z scope that is subject to the Freeze Bug but does NOT have Boot Version 0.0.1.2 installed, please report it. And if you know of one that DOES have Boot Version 0.0.1.2 installed but does NOT have the Freeze Bug, please let me know and I'll send a .stp setup file that contains all the conditions so that scope can be tested without having to worry whether the user actually did try the proper combination of conditions. Until we have such data, I'll continue to believe that the Freeze Bug is associated with Boot Version 0.0.1.2, since all the data we _do_ have -- including poorchava's test -- supports that conclusion. It may be the case that some combination of Boot Version 0.0.1.2 plus some hardware instability is required for the Freeze Bug, making the Boot Version 0.0.1.2 a _necessary_ condition but not a _sufficient_ one.

[Fungus]

Er.... no. Every case that I know about of a scope that can be "locked up" due to the Freeze Bug is running Boot Version 0.0.1.2, no matter what "firmware" the scope is running.

So? 

Think harder: The bootloader is roughly related to the date of manufacture. It might be a manufacturing problem for a single batch of oscilloscopes (this seems more likely to me). A bad batch of chips, a cockroach nest fell in the solder bath, ...

If it was purely the bootloader then all scopes with that bootloader would have the problem. They don't.

I really can't imagine that any code in the bootloader is being used once the system is up and running.

[borjam]

I really can't imagine that any code in the bootloader is being used once the system is up and running.

Unless it performs some hardware initialization, as it can happen with peecees and their "BIOS".

I mean, some poorly done initialization (or something that triggers a hardware fault in certain series) can indeed provoke an effect even though the boot loader is no longer running. Maybe the boot loader does something more than, well, loading the firmware, and different versions perform different initialization tasks so that what they call "firmware" doesn't require different files or hardware version checking.

We don't know

[alsetalokin4017]

Er.... no. Every case that I know about of a scope that can be "locked up" due to the Freeze Bug is running Boot Version 0.0.1.2, no matter what "firmware" the scope is running.

(snip)

If it was purely the bootloader then all scopes with that bootloader would have the problem. They don't.

(snip)

And I'm still not convinced that this is the case. As I asked before, if you know of a scope that is running Boot Version 0.0.1.2 and does not have the Freeze Bug, please let me know, and I'll send you (or the scope's owner) a setup file with the proper conditions to produce the bug. If you followed the original thread on that topic, you will see multiple cases where people said they didn't have the bug, but only later _when properly following directions_ found that they did indeed have it.

I'm not saying absolutely that all scopes with Boot Version 0.0.1.2 have the bug, but I'd love to properly test one such that is reported not to have it. And as I also said before, the Boot Version 0.0.1.2 seems to be _necessary_ but may not be _sufficient_, so I'm open to the idea that some interaction between the Boot Version and some hardware or other problem might be required for the Freeze Bug to occur. Only _proper_ testing will answer that question, and I'm not convinced yet that such proper testing has happened.

Furthermore, how do we actually know that the "Boot Version" functions as what we commonly know as a bootloader? After all.... what we have been calling "firmware" isn't even what the scope calls firmware -- the scope calls it "software" -- and what the scope calls "firmware" is something else entirely.

By the way, Fungus.... do you actually own a Rigol DS1054z yourself?

[ProBang2]

Just for fun I had compared the system information in the upper post with mine.
It has changed since last time. Now my scope displays the following:

(No trick, no fake, no joke!)   



What the hell is going on?    

[smithnerd]

I agree with Fungus. I think that those blaming the bootloader are falling into the logical trap of assuming that correlation implies causation.

Is the datasheet for this processor part available without NDA? The errata list would be an interesting read.

[alsetalokin4017]

I agree with Fungus. I think that those blaming the bootloader are falling into the logical trap of assuming that correlation implies causation.

Is the datasheet for this processor part available without NDA? The errata list would be an interesting read.

Feel free to present evidence that Boot Version 0.0.1.2 is _not_ the cause of the Freeze Bug problem.

As I've said multiple times now.... 1. The Boot Version 0.0.1.2 is _necessary_ but may not be _sufficient_ to cause the issue; and 2. if you or anyone else knows of a scope with Boot Version 0.0.1.2 that the user thinks does _not_ have the Freeze Bug, let's test it with a setup file that I'll provide which contains all the conditions necessary to produce the Freeze on scopes that are susceptible to it.

[alsetalokin4017]

Just for fun I had compared the system information in the upper post with mine.
It has changed since last time. Now my scope displays the following:

(No trick, no fake, no joke!)   



What the hell is going on?    

Better check inside to make sure the board is still there.....   

Welcome to the wonderful world of Rigol!  I've even had mine miscount the startup counts, reporting a smaller count value at a later date.

This is where someone usually chimes in and says "What did you expect, it's only 400 dollars..."   

[xrunner]

This is where someone usually chimes in and says "What did you expect, it's only 400 dollars..."   

I don't know what you people expect - the thing is only $400 ... 

[Fungus]

Feel free to present evidence that Boot Version 0.0.1.2 is _not_ the cause of the Freeze Bug problem.

It makes no sense that it would be*, so:
a) You're the one making extraordinary claims
b) The burden of proof lies with you

(*) A bootloader by definition does just enough to get the machine running, and no more.

[Fungus]

This is where someone usually chimes in and says "What did you expect, it's only 400 dollars..."   

I don't know what you people expect - the thing is only $400 ... 

Well, it's true. It is only $400.

[Gabri74]

Meanwhile -- if you know of a DS1054Z scope that is subject to the Freeze Bug but does NOT have Boot Version 0.0.1.2 installed, please report it. And if you know of one that DOES have Boot Version 0.0.1.2 installed but does NOT have the Freeze Bug, please let me know and I'll send a .stp setup

Hi alsetalokin4017, boot version 0.0.1.2 here, never been able to reproduce the freezing bug, happy to test your .stp file :-)

[poorchava]

Its possible,  that the connection is opposite. Boot 0.0.1.2 doesn't  cause freeze bug. Maybe fixing the bug required some change in hardware and this hardware needed changes in bootloader,  and hence 0.0.1.3 came to be.

Sent from my HTC One M8s using Tapatalk.

[ProBang2]

The three zero´s were somehow familar to me.
But from where?
And now I got it!

From here: https://youtu.be/WCSKXugQrP0?t=1m51s

At least I know now what is coming next...   

[alsetalokin4017]

This is where someone usually chimes in and says "What did you expect, it's only 400 dollars..."   

I don't know what you people expect - the thing is only $400 ... 

I expect it to do "what it says on the box". The manufacturer says it will do certain things, and sets the price point. So people purchase it at that price point and then find out that it does not perform as the manufacturer said. Is the purchaser at fault for expecting it to perform as claimed, whatever the manufacturer decides to sell it for?
It seems that some people think so.

[alsetalokin4017]

Its possible,  that the connection is opposite. Boot 0.0.1.2 doesn't  cause freeze bug. Maybe fixing the bug required some change in hardware and this hardware needed changes in bootloader,  and hence 0.0.1.3 came to be.

Sent from my HTC One M8s using Tapatalk.

That's a pretty big stretch, since Rigol evidently didn't know about the Freeze Bug until I reported it, and by that time all the scopes in the lab at RigolUSA were already running Boot Version 0.0.1.3 and couldn't reproduce the bug.... so they asked me to return my scope with 0.0.1.2 and replaced it with one having 0.0.1.3. Note that I did not initiate this swap, it was offered by Rigol USA and they even sent me the replacement _before_ I sent the earlier version back, so I had three days or so to test the two scopes side-by-side.

[alsetalokin4017]

Meanwhile -- if you know of a DS1054Z scope that is subject to the Freeze Bug but does NOT have Boot Version 0.0.1.2 installed, please report it. And if you know of one that DOES have Boot Version 0.0.1.2 installed but does NOT have the Freeze Bug, please let me know and I'll send a .stp setup

Hi alsetalokin4017, boot version 0.0.1.2 here, never been able to reproduce the freezing bug, happy to test your .stp file :-)

OK... try this. You'll have to unzip it because I can't upload a .stp file directly to the forum. If your scope _does_ freeze you can get it unfrozen by using the "5th grey button on left while booting" procedure to get it back to factory defaults, and the scope will come up in Chinese....

You may also want to review some of the videos I posted back in the day, like this one:


ETA: If people are hesitant to try the setup file here's a simplified set of instructions starting from the Default setup (Storage>Default) that worked for many people in the FreezeBug thread:
1. Display a signal on CH1 from the probe Calibrator, at 1V/div, trigger type Edge, trigger level 2.00 V, sweep mode Auto
2. Set Horizontal Scale to 1 us/div
3. Select "Auto" from Acquire>Mem Depth menu
4. Select "100 ms" from Display>Persis. Time menu
5. Enter Horizontal Zoom mode by pressing the Horizontal Scale knob

[xrunner]

I expect it to do "what it says on the box". The manufacturer says it will do certain things, and sets the price point. So people purchase it at that price point and then find out that it does not perform as the manufacturer said. Is the purchaser at fault for expecting it to perform as claimed, whatever the manufacturer decides to sell it for?
It seems that some people think so.

Yep, and I've backed you on that too. If they want to add measurement "bling" to help market the thing, and the additions don't work right, then they should think about other things to add to help it sell, such as different case colors (consult Martha Stewart) or perhaps a talking menu system. 

[alsetalokin4017]

OOohh.... I think a Pink or Purple case would look great ....     

[Gabri74]

OK... try this. You'll have to unzip it because I can't upload a .stp file directly to the forum. If your scope _does_ freeze you can get it unfrozen by using the "5th grey button on left while booting" procedure to get it back to factory defaults, and the scope will come up in Chinese....

Nope, can't reproduce 
Loaded your setup file, tried with 1,2,3 and 4 channels, with and without signal applied.
Repeated every single step in your video starting from minute 1:20 and also followed your 5 points instructions
and abused Horizontal knob with and without zoom on...but no freeze here.

Attached my scope infos for reference.