Rigol HDO1000 and HDO4000 12bit oscilloscopes launched in China

[norbert.kiszka]

Hmmmm maybe I forget about one line in the script in the .GEL file...

Try to do this command:

Code: [Select]

pm enable com.rigol.scope

[lownoise]

I did it, that's the answer...

[norbert.kiszka]

Now Im sure what I did wrong. Wrong file name 

So the com.rigol.scope became uninstalled and later it's not installed.

As for now, for the quick fix, You can install it manually via adb:

Code: [Select]

adb install -r Auklet.apk
I will fix this .GEL in the next ~15 minutes.

[lownoise]

It can't be found in Adb shell, but I found it here in  rigol/app
Should I copy it from there to somewhere else?

[norbert.kiszka]

It can't be found in Adb shell, but I found it here in  rigol/app

This is the stock app. You can use apk from my Patreon page or You can install this stock one and use fixed .GEL, because I just fixed and uploaded it.

Edit: You should have modified one in /rigol/resource/mod/apk/

[lownoise]

I installed Auklet.apk from your Patreon page.
Now it has also found NOVA Launcher and it starts correctly 

p.s.

Now I found the mod version in rigol/resource/mod/apk.
Can I just install it over, or do I have to uninstall the current one first?

[norbert.kiszka]

I installed Auklet.apk from your Patreon page.
Now it has also found NOVA Launcher and it starts correctly 

p.s.

Now I found the mod version in rigol/resource/mod/apk.
Can I just install it over, or do I have to uninstall the current one first?

On the attached screen, I see either older version or the stock app. I removed this ugly grey background.

Since You have hacked Android, part of it called as Package Manager (pm) doesn't care if the apk is newer, older or if it has proper encryption keys (which was a huge barrier to install modified app).

On the side note, this is a huge pain in the Android, especially if You have not-rooted and unmodified Android on the phone or tablet. If You install some update from the vendor, You can't go back without some hacking. One of my two phones become practically not usable because of it - I need to take some time and flash some already hacked Android.

So You can install almost any apk (if it works on Android 7.1.2 and ARMv8) with any of those two:

Code: [Select]

pm install -r local.apk

Code: [Select]

adb install -r remote.apk

[lownoise]

Thanks for the help Norbert!

I just did a normal update with the new Gel version 0.1-1 and now everything works, including Self-Cal

Good work

[thm_w]

Installed GEL, 500M is a huge amount of data, so it can take time to process but from what I can see it works well.
Black background is nice. Thank you

[lownoise]

Hi thm_w,
I may be mistaken, but does 1GHz FFT only work on CH1 ? 
Could you please check this....

[norbert.kiszka]

FFT max bandwidth is the half of the sample rate. On those screens You have 1 GS/s instead of 2 GS/s because You have two channels enabled.

[lownoise]

Hi Norbert, thanks for the tip.
When I have all channels off, it shows 2Gs.
When I turn CH1 on, it shows 2Gs.
When I turn CH1 off again, it shows 2Gs.
However, when I turn CH2 (-4) on/off, it shows 1Gs.
Could it be that CH1 is not really off?

[norbert.kiszka]

Because You have trigger set to be on CH1. So the CH1 is always working, but not necessarily rendered.

[lownoise]

Yes, you're right 
Now it works

[TurboTom]

@ Norbert -
I tried your firmware 0.1-1 and I'm very impressed by the huge sample memory -- I never expected this to be possible. Yet, I found some other peculiarities that may relate to hardware limitations or other shortcomings. One of them are the input lowpass options: Here, 400MHz, 200MHz and 100MHz appear to be solid, 70MHz and 20MHz clearly show their digital origin, and 250MHz is plainly useless since it dampens a 200MHz signal stronger than the 200MHz filter... (see attached screenshots -- the level drop in the shot without any lowpass filter at 500MHz originates from the limits of the noise source as the following post proves).

(continued in next post)

[TurboTom]

And then I've got some very peculiar findings with the 50 Ohms input impedance mode:

First of all the bandwidth is fantastic -- that is, if you know what you're doing and you limit the signal to below Nyquist. I found the 3dB bandwidth of the input circuitry to be round about 900MHz. As you see in the attached screenshots, the single-channel useful bandwidth (due to aliasing artefacts) is somewhere between 700 and 800Mhz -- mighty impressive. Interestingly, the hardware frequency counter appears to sample at 1GHz since beyond 500MHz input frequency, the frequency reading gets mirrored, until the figures get unstable somewhere above 700MHz.

What I don't like is that the self-calibration of the 50 Ohms impedance branch doesn't seem to work properly. May it be related to a defect of my particular oscilloscope or is it a common problem of DHO1000's with Norbert's firmware? Did others make similar observations?

What's quite peculiar with my DHO1000 is that I get a failure of the self calibration whenever I check one of the following calibration jobs:
 - ADC Gain AC
 - DDR
 - ADC Clock
All others run without error messages even though they don't produce any results regarding the 50 Ohms branches.

You can see the effect in the two attached screenshots NK_v0.1-1_10M_100mV_H and NK_v0.1-1_10M_100mV_L which are taken with the same input level, yet one before the attenuator switches between the two 50 Ohm paths and the other after. The signal gets attenuated by more than half when switching to the less sensitive range...

Another peculiarity I observed on my DHO1000: With 50 Ohms active on all channels, CH1 switches the attenuator relay between the input sensitivity settings 50mV/div and 100mV while all other channels switch over between 100mV/div and 200mv/div.

When setting up the input impedance as 1MOhm, she switchover point is identical on all channels.

Can possibly someone cross-check these findings to make sure it's not just a defect on my specimen before I ring the "bug bell"? 

Thanks and cheers,
Thomas

[norbert.kiszka]

You should run self-cal with default checkboxes. Those are hidden in the stock app (available only after taping three times on about tab).

DDR in DHO800/900 reflashes another flash memory region and boots FPGA from there - more or less same FPGA image, but it does some magic with DDR memory and app grabs this data, which is saved into file and used at next start of app. This is quite useful after overclocking PLL (at least in DHO800/900). Im not sure if it's exactly the same in DHO1000/4000.

I will go back later to this. I just fixed my own scope (it was just some bad data in FRAM) and currently Im working with MHO900. It's hard to work with two different series at the exact same time.

[thm_w]

As Norbert has said its a combination of ADC and AFE to set a bandwidth limit, so maybe some are more useful than others.

Another peculiarity I observed on my DHO1000: With 50 Ohms active on all channels, CH1 swithes the attenuator relay between the input sensitivity settings 50mV/div and 100mV while all other channels switch over between 100mV/div and 200mv/div.

When setting up the input impedance as 1MOhm, she switchover point is identical on all channels.

For me on 50R I see: CH1, CH3, CH4 0.1-0.2V, CH2: 0.05-0.1V
1M: CH1, 2, 3, 4 = 0.05-1V

[lownoise]

Hi TurboTom, thm_w,
I just tested it and for me it's the same at 1M, but different at 50R:

50R
CH1, CH4: 100-200mV
CH2, CH3:   50-100mV

1M
Ch1-Ch4:     50-100mV

[RGF]

hello Thomas
  This is Richard Fentress, this doesn't have anything to do with this post. i didn't know how else to send a message to you.
About 8yrs or so ago i asked you if you would build me one of these clocks, you said yes but i didn't have the money.
i do now. would build one now & if so, how how much? I have a couple crt tubes   

[Kurt_!]

Hello
I'm looking for the "Board Test" option to read the temperature.
Is it possible that it disappeared with the last firmware update?

Can anyone help, please?

I've installed two quieter fans.

[skander36]

Yes, they removed these entry in the menu. You can go back 1 or 2 Fw. versions check the board components temperatures and after that come back to the last Fw. version.

[Kurt_!]

Yes, they removed these entry in the menu. You can go back 1 or 2 Fw. versions check the board components temperatures and after that come back to the last Fw. version.

Thanks


Yes, then I'll have to write to Rigol again. I already have quite a bit to say about the next firmware update.
I hope other people will write too... Maybe it will help?

[skander36]

Yes, they removed these entry in the menu. You can go back 1 or 2 Fw. versions check the board components temperatures and after that come back to the last Fw. version.

Thanks


Yes, then I'll have to write to Rigol again. I already have quite a bit to say about the next firmware update.
I hope other people will write too... Maybe it will help?

From my experience I can say that they never answer me. But if you can read through this thread there are people here that report that the support contacted them and answer to the questions. I think you should try, maybe things have changed in the meantime.

[Kurt_!]

Yes, they removed these entry in the menu. You can go back 1 or 2 Fw. versions check the board components temperatures and after that come back to the last Fw. version.

Thanks


Yes, then I'll have to write to Rigol again. I already have quite a bit to say about the next firmware update.
I hope other people will write too... Maybe it will help?

From my experience I can say that they never answer me. But if you can read through this thread there are people here that report that the support contacted them and answer to the questions. I think you should try, maybe things have changed in the meantime.

Rigol always wants the errors explained in detail.
That's a bit difficult for me, unfortunately, as my English isn't very good.

Regarding the fans...
I think it'll be fine. They're the 25mm thick Noctua fans.
Basically, it just blows air from one side to the other, right?
I can recommend it.