Rigol's New DHO800 Oscilloscope unbox & teardown

[StillTrying]

I agree, but I would hope they're doing 4096/8 which would give 512 levels to cover the ~400 Y pixels, which would be an improvement over the display of 200 levels covering 400 pixels that we usually see on an 8 bit.

Lets hope the extra resolution is used during MATH at least.

[souldevelop]

Guys, I'm back, I self-calibrated DHO914S and tested the noise floor, which is probably between 210uVpp~260uVpp, which is basically normal. I tested the BODE plot and found a bug that when the CH1 or CH2 channel gear is much smaller than the signal amplitude of the AFG, it runs without any display, which is unacceptable, and I guess it is not able to automatically adjust the vertical gear of the oscilloscope to adapt to the signal source. I didn't know if it was just me who appeared, and then I connected my homemade 16-channel LA and the test results were satisfactory. By the way, I got another 250M bandwidth for free, and now the only thing missing is VESA100 Didn't arrive, I'll make up later, thanks.

[EEVblog]

Teardown:
https://www.flickr.com/photos/eevblog/albums/72177720310962700/

[Fungus]

Teardown:
https://www.flickr.com/photos/eevblog/albums/72177720310962700/

I'm not seeing any dedicated RAM for sample memory and that Zynq chip doesn't have enough.

There's also no obvious wide signal paths coming out of the FPGA so where's the sample memory? Is it inside this ADC chip?


(Just wondering if the DHO800 will be hackable to 50Ms memory...)

[Warhawk]

Guys, I'm back, I self-calibrated DHO914S and tested the noise floor, which is probably between ......

Thanks for sharing. I am extremely interested in the HDMI output. Could you please connect the scope to an external monitor and show us how it looks? Also, does the scope support a mouse/keyboard when you plug it in to USB. (R&S scopes do).

[Fungus]

Guys, I'm back, I self-calibrated DHO914S and tested the noise floor, which is probably between ......

Thanks for sharing. I am extremely interested in the HDMI output. Could you please connect the scope to an external monitor and show us how it looks?

https://www.eevblog.com/forum/testgear/rigols-new-dho800-oscilloscope-unbox-teardown/msg5044444/#msg5044444

Also, does the scope support a mouse/keyboard when you plug it in to USB. (R&S scopes do).

It supports mouse.

[Warhawk]

Guys, I'm back, I self-calibrated DHO914S and tested the noise floor, which is probably between ......

Thanks for sharing. I am extremely interested in the HDMI output. Could you please connect the scope to an external monitor and show us how it looks?

https://www.eevblog.com/forum/testgear/rigols-new-dho800-oscilloscope-unbox-teardown/msg5044444/#msg5044444

Also, does the scope support a mouse/keyboard when you plug it in to USB. (R&S scopes do).

It supports mouse.

Excellent, thanks for sharing.

[EEVblog]

I tried to read the 32GB SD card that was glued into place on the board, but can't see anything. I tried using DiskINternals Linux Reader but it doesn't read it.
Anyone know what it's for?

[EEVblog]

Guys, I'm back, I self-calibrated DHO914S and tested the noise floor, which is probably between ......

Thanks for sharing. I am extremely interested in the HDMI output. Could you please connect the scope to an external monitor and show us how it looks? Also, does the scope support a mouse/keyboard when you plug it in to USB. (R&S scopes do).

Go back one page, I posted photos.

[Fungus]

All the menus and fonts are being rendered at the higher resolution?

They seem to be.

That makes sense if you think about it. Their new OS has to work on different size screens (eg. the DHO1000/4000 series have bigger screens) so it could be programmed to scale the UI properly.

The RK3399 chip which is driving the HDMI output is capable of 2160p@60fps, ie. 4096x2304 resolution.

http://rockchip.wikidot.com/rk3399

(It should run DOOM amazingly well, and quite a few other things, too)

The waveform display will be down to the FPGA so maybe that's not as flexible.

[2N3055]

On my 2K monitor, at first is chucked a wobbly and told me 1920x1080 with garbage but after a reboot I got 2560x1440
So it seems like the Rigol is detecting the resolution of the monitor and adjusting it's output?
I can say that it looks very nice and sharp. It's clearly not a fixed 1024x600 output.

Both resolutions you mentioned were scaled up 1280*720 :
x3   3840*2160   
x2 2560x1440.

It might be scaler in scope or monitor, but seems like upscaling of native 1280*720.

[Fungus]

Both resolutions you mentioned were scaled up 1280*720 :

Most resolutions are multiples of some smaller resolution, no surprises there.

OTOH 1280x720 doesn't match either the HDO800 or the HDO1000 built-in screens. The HDO800 is 1024x600 and the HDO1000 is 1280×800.

It looks to me more like the UI is fully scalable (which makes sense - it's an Android APP)

[2N3055]

Teardown:

I'm not seeing any dedicated RAM for sample memory and that Zynq chip doesn't have enough.

There's also no obvious wide signal paths coming out of the FPGA so where's the sample memory? Is it inside this ADC chip?

(Just wondering if the DHO800 will be hackable to 50Ms memory...)

You can see space for 2 chips above FPGA... This is extremely cheap design, and it is probably same thing as they did long time ago for some of their scopes: fast sampling only with small memory. They could use CSI to transport data between FPGA and CPU and just use RAM there.  1Mpts for 4 ch actually makes sense now.

It is cheapest 12bit scope ever made. Compromises will exist.

[Warhawk]

Is this 1MPts already confirmed? I am not able to follow the threads that quickly. Also true for 900 series?

[2N3055]

Is this 1MPts already confirmed? I am not able to follow the threads that quickly. Also true for 900 series?

No,  it was not confirmed, AFAIK.
DHO900 claims 10Mpts in 4 ch mode. It might have RAM chips installed.

[Nikki Smith]

No,  it was not confirmed, AFAIK.
DHO900 claims 10Mpts in 4 ch mode. It might have RAM chips installed.

Someone opened up a DHO924 and took photos of those 2 DRAM chips installed.
https://www.eevblog.com/forum/testgear/rigols-new-dho800-oscilloscope-unbox-teardown/msg5043223/#msg5043223

[Warhawk]

Could it be the result of the 16b LA?
Just speculating. 1Mpts is insanely low... (cough, cough, looking at you my Agilent with 32kpts)

[Howardlong]

If a boot time of about 1 minute on average causes a noticeable loss of productivity, you should rethink your time management in general.

Others weren't talking "about 1 minute", they were talking of 2-3 minutes and an hour to warm up.

Furthermore, it's not "about 1 minute" anymore if you're going to make a cuppa and going to the washroom.

I'm just going from personal experience: once you take you mind off the task at hand, a half hour piece of work can easily run into an hour and more as you suddenly find and start other unrelated jobs that need fulfilling. It's human nature.

Thus the longer it takes an interruption to complete, the more likely you are to go and fill the time with something else unrelated, upsetting your train of thought, and ultimately wasting time unnecessarily.

you should rethink your time management in general.

Thanks, dad.

[Aldo22]

OTOH it could have a sleep mode. This is Android...

Maybe the FPGA, front end, etc., can't go to sleep so they'd have to leave most of it powered on, including the fan. I wouldn't want a noisy 20W device constantly running on my desk.

Exactly.

The problem is that there is no standby mode that would at least turn off the fan and the display.
I only boot my PC or smartphone in emergencies or after system updates. Then it takes 22 seconds for the PC.
Otherwise, these devices are in standby (suspend to RAM).

If I have to decide whether I want to hear the annoying fan or whether I want to wait 54 seconds for the oscilloscope to run again in a few minutes, that could be a nuisance.

Some may still be used to this from the CRT, but it doesn't have to be anymore, does it?
Admittedly, I don't have much idea about "real" oscilloscopes yet.
The Rigol would be my first and I'm trying to get my head around it. Maybe I'm worrying too much about this.

My Zeeweii runs quasi immediately, like a multimeter. Therefore I am perhaps somewhat spoiled. 

[EEVblog]

Most resolutions are multiples of some smaller resolution, no surprises there.
OTOH 1280x720 doesn't match either the HDO800 or the HDO1000 built-in screens. The HDO800 is 1024x600 and the HDO1000 is 1280×800.
It looks to me more like the UI is fully scalable (which makes sense - it's an Android APP)

Yes, makes sense that it's scaleable. How the FPGA waveform mapping works is a different story.

[Fungus]

They could use CSI to transport data between FPGA and CPU and just use RAM there.

It's a fierce data rate.

1Mpts for 4 ch actually makes sense now.

No it doesn't. Data is data, doesn't matter if it's interleaved or not.

It is cheapest 12bit scope ever made. Compromises will exist.

Obviously.

Reducing it to only six chips will help a lot with the price.

[.RC.]

Imagine how much faster the electrons could travel if they did not have to slow down to go round those bends. It is like they have deliberately crippled it. 

[tv84]

I tried to read the 32GB SD card that was glued into place on the board, but can't see anything. I tried using DiskINternals Linux Reader but it doesn't read it.
Anyone know what it's for?

Can you post a picture of the card? Let's assume it ain't there only to consume power so, if the controller is not customized, one should be able to read it...

As always: "no coincidence, just lack of information".

[TurboTom]

The ram chip is 256Mx16 DDR3-2133, so it should pretty well get the job done for 50MSa of the scope alone. Since the digital (MSO) lines have to be sampled alongside of the ADC data output, I assume that the additional memory is primarily for the MSO function. Moreover, virtually all DSOs work in a kind of "hickup mode" of sampling to memory, processing the data and displaying it. During the latter two steps, the sampling engine usually pauses and gets restarted only after these are completed. Of course, things may work diffenently in segmented memory mode or other "fast sampling" schemes. Here, usually the ADC and trigger engine are running all the time, but this doesn't need any sample memory worth mentioning, and sampling to memory only starts after a trigger event has been detected.

So I'ld expect, memory-wise the 50Msa should be possible on the HDO800 as well but not the MSO functionality.

The missing small stuff appears to be one power supply and some auxiliry analog sample and hold circuits (CMOS OpAmps) to supply config/threshold voltages to the MSO / AWG sections.

BTW: Rigol still uses an FRAM chip to store configuration information and other frequently changed data, here it's a MB85RC64TA (64kbit), located at the "back" of the PCB -- Nice!

[EEVblog]

The missing small stuff appears to be one power supply and some auxiliry analog sample and hold circuits (CMOS OpAmps) to supply config/threshold voltages to the MSO / AWG sections.

There apprears to be no AWG harwdare except for a buffer driver on the output?