Rigol DHO800/900 new firmware... 1.03

[Moduvator]

But I went back temporarily to 1.03. It didn't happen with 1.03. After re-updating to 1.04 it was reproducable again. So it depends from the 1.04.

Hmm.... That's definitely important information, thank you for testing that! Haven't had the chance to roll back to 1.03 on mine.
Guess they broke something in the scripts and display controller remains stuck partially on somehow.

[norbert.kiszka]

But I went back temporarily to 1.03. It didn't happen with 1.03. After re-updating to 1.04 it was reproducable again. So it depends from the 1.04.

Hmm.... That's definitely important information, thank you for testing that! Haven't had the chance to roll back to 1.03 on mine.
Guess they broke something in the scripts and display controller remains stuck partially on somehow.

Between 1.03 and 1.04 only app was changed. Scripts remained unchanged.

[vano]

I switched the oscilloscope to sleep mode several times, from 30 minutes to 2 hours, no stains, everything is fine.

[AndyBig]

I also had the oscilloscope in standby mode for up to a day and there were no problems with the display.

[Moduvator]

Quite frustrating indeed! I was able to reproduce it again over the weekend, even though this was now a vertical strip of brighter pixels in the right part of the screen.

The scope then worked fine today over one several hours long sleep cycle after a fresh boot, however when coming out of sleep a second time that strip has reappeared. Moreover, I was also been able to spot a very faint RIGOL logo imprinted, as if that was the last frame sent to the LCD as it shut down but pixels got stuck under constant voltage.

Can't work out the pattern so far but this does seem to require a few sleep and wake up cycles to trigger, might not work on the first one!

[Moduvator]

I think I might be on to something. I have not cold booted the scope yet, having obtained the above pictures, and this time I was able to reliably reproduce that ghost logo after just some 40 minutes of sleep. This time you can even make out the spinner it shows as the system goes to sleep.

Overall it does seem you need a few sleep-wake cycles for it to manifest but once triggered it would then happen every time.

[Fungus]

Looks like some sort of "transient image persistance".

https://en.wikipedia.org/wiki/Image_persistence

AFAIK it's not permanent on LCDs.

[Moduvator]

AFAIK it's not permanent on LCDs though.

It would depend on how long the display is left in that state. Seems that when the OS enters sleep, the display refresh circuitry is switched off but whatever image was last in the buffer is still actively driving the pixels, causing them to be stuck like this. If that is true this could eventually permanently polarize the liquid crystals if left alone for long enough -- this is why drive LCDs with AC...
Here it seems that the charges also slowly "bleed" into surrounding pixels the longer the scope is left in sleep, explaining that 'halo effect' we saw earlier. Since this does not happen every time and certainly not between cold boots, this leads me to suspect it's a software/firmware issue...

I stand to be corrected =)

[norbert.kiszka]

Maybe somebody missed, but long time ago I shared here my reverse engineering (and modification) of this "sleep mode". In this state, app is killed and modules are unloaded. One added module in 1.03 when is unloaded, it switches off only one CPU GPIO output - that causes both fan and the screen to turn off.

I did this reverse engineering only on software side. So I don't know where this pin is connected.

[fearas38]

I checked SMB on the new firmware. It works perfectly with LAN, but it doesn't work via WiFi. Does anyone know why?

[norbert.kiszka]

I checked SMB on the new firmware. It works perfectly with LAN, but it doesn't work via WiFi. Does anyone know why?

Maybe bad ip mask. It's just a guess.

[fearas38]

Through the Wi-Fi - web interface works, I can connect to the oscilloscope. Through LAN, SMB also works. However, when I disable LAN, leaving only Wi-Fi, SMB immediately disconnects. I tried using static IP addresses on Wi-Fi, but it didn't help. Has anyone managed to get SMB working over Wi-Fi?[/img]

[norbert.kiszka]

when I disable LAN, leaving only Wi-Fi, SMB immediately disconnects

Sounds like a designed behavior (no lan = no network - Rigol employees didn't think some users will use WiFi via USB. Probably they don't think at all).