Rigol DHO804 Test and Compare Thread

[Martin72]

Yes, but if you're going out to buy today... not even on the shortlist, right?

As a main device, not a single 7" scope would be on my list, the small screen would be too tiring for me in the long run.
If that weren't the case and I had a tight budget and no scope yet, the SDS1104X-E and the DHO804 would remain.
However, since I need Bodeplot and the Rigol doesn't offer one, I would have to opt for the siglent.
That's quite a shame, because I like the DHO804 better in some areas (what exactly will follow).
And even if I had more money to spare, a DHO900S would be out of the question for me, as I consider it a rather superfluous model.
What can it do better than the 800, almost nothing.
If you don't need an LA, the device is a more expensive 800.
As things stand, I think the 800 series is the best of all the new Rigol scopes, no joke.
The "oddities" and the sometimes rather meager features can be forgiven for the cheap 800, as the owner of a DHO1000 or even 4000 I would feel a bit stultified in view of the fact that most of the hardware used there is also available for a quarter or eighth of the price with the 800 and the software and thus the operation is the same anyway.

[nctnico]

Yes, but if you're going out to buy today... not even on the shortlist, right?

As a main device, not a single 7" scope would be on my list, the small screen would be too tiring for me in the long run.
If that weren't the case and I had a tight budget and no scope yet, the SDS1104X-E and the DHO804 would remain.
However, since I need Bodeplot and the Rigol doesn't offer one, I would have to opt for the siglent.
That's quite a shame, because I like the DHO804 better in some areas (what exactly will follow).
And even if I had more money to spare, a DHO900S would be out of the question for me, as I consider it a rather superfluous model.
What can it do better than the 800, almost nothing.
If you don't need an LA, the device is a more expensive 800.
As things stand, I think the 800 series is the best of all the new Rigol scopes, no joke.
The "oddities" and the sometimes rather meager features can be forgiven for the cheap 800, as the owner of a DHO1000 or even 4000 I would feel a bit stultified in view of the fact that most of the hardware used there is also available for a quarter or eighth of the price with the 800 and the software and thus the operation is the same anyway.

I agree. If you could modify the DH800 to have a bigger display, it would be more or less equal to the way more expensive DHO models.

[Fungus]

However, since I need Bodeplot and the Rigol doesn't offer one, I would have to opt for the siglent.
And even if I had more money to spare, a DHO900S would be out of the question for me, as I consider it a rather superfluous model.

With the Siglent you have to add the cost and extra bench space of a signal generator. A DHO900 could compare very favorably to that.

What can it do better than the 800, almost nothing.
If you don't need an LA, the device is a more expensive 800.

If you're not going to use the LA then the DHO900 makes no sense at all. Some people might think it has a higher bandwidth but that's not really true and it's just more opportunity for aliasing.

The DHO900 is a sweet deal for Rigol though. $200 higher sticker price could maybe make them three times the profit, all for the cost of a $1 edge connector.

(I think this is the reason they so massively underrate the official bandwidth of the DHO800)

As things stand, I think the 800 series is the best of all the new Rigol scopes, no joke.

Yep. The DHO804 is amazing value.

[Martin72]

With the Siglent you have to add the cost and extra bench space of a signal generator.

I wouldn't have the "problem" with the Rigol, of course.
But then I wouldn't have a bode plot either.
This is a good opportunity to buy yourself a nice generator.
And Rigol could of course also implement Bodeplot and then design it so that you can use external (Rigol) generators for it.
But I don't think that will ever happen.
That's just a point for me, if you don't need a Bodeplot, you can live very well with the 800.

[Fungus]

I wouldn't have the "problem" with the Rigol, of course.
But then I wouldn't have a bode plot either.

Was the Siglent Bode plot 100% perfect on launch? 

(I'm still waiting for an explanation of the Siglent's FFT problems shown in Dave's review of the DHO800...)

[Martin72]

Was the Siglent Bode plot 100% perfect on launch?

I didn't follow that back then, I just know that when I got to know it (2018), it was really good.
You can't compare that with what rigol almost reluctantly pushed after the MSO5000 via firmware update.
And that was years ago and what you can see with the 900S is exactly the same - they have no desire to make something good out of it.

I'm still waiting for an explanation of the Siglent's FFT problems shown in Dave's review of the DHO800...

What problem ?
That would be good if you could name it, because I have access to the scope at work.
This is also interesting in that the FFT performance of rigol is the biggest weakness for me, apart from the lack of a bodeplot function.

[Mechatrommer]

You people dont look dave's video arent you people? siglent has wobbly fft and dho800/900 can be connected to 99" tv with touch if you wish to btw.. going to sending 'less small sparrow afg' board to jlcpcb tonight after final checking stage... fwiw..

[Fungus]

I'm still waiting for an explanation of the Siglent's FFT problems shown in Dave's review of the DHO800...

What problem ?
That would be good if you could name it, because I have access to the scope at work.

See the video at 32:15:

[Fungus]

I didn't follow that back then, I just know that when I got to know it (2018), it was really good.

I don't follow Siglent threads much either but their firmware has release notes and it's
quite long:

https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2018/02/SDS1000X-Firmware-Revise-History.pdf

I know the Bode plot can glitch if it does one of its little internal recalibrations half way through:

https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-bode-plot-glitches!/


You probably don't remember Siglent mailing little bags of capacitors to people to install in their new 'scopes either?

https://www.eevblog.com/forum/testgear/siglent-sds1202x-e-compensation-issue/msg1350586/#msg1350586

Imagine if Rigol did that? We'd never hear the last of it.

[tautech]

Fuck off Fungus
https://www.eevblog.com/forum/projects/project-yaigol-fixing-rigol-scope-design-problems/

Shit PLL lock also:

[ebastler]

Fuck off Fungus
https://www.eevblog.com/forum/projects/project-yaigol-fixing-rigol-scope-design-problems/

Come on, man, get a grip. What kind of tone is that?

Regarding early scope issues: Sure, the "Yaigol" bug was disconcerting. But it was fixed within six weeks by a firmware update (improving the PLL settings). In contrast, Siglent's SDS1202X-E had an actual hardware issue. And they chose to address this not by a clean and simple "we'll send you a replacement scope, then you return your buggy scope to us", but by shipping DIY capaycitor kits. Not something to be proud of.

[tautech]

Regarding early scope issues: Sure, the "Yaigol" bug was disconcerting.

DS2000

But it was fixed within six weeks by a firmware update (improving the PLL settings).

DS1054Z

[ebastler]

DS2000
DS1054Z

The PLL issue was quickly addressed via firmware updates for both scopes, to my knowledge. I did not follow the DS2000 thread back then, but skimming it now it seems to me that Bud went off on a tangent. Which scope performance metric needed further improvement exactly?

[RoGeorge]

The PLL jitter bug was fixed with a firmware update, after Dave showed the bug.  Since then (~8 years ago) the PLL is working just fine on my Rigol DS1054Z.

[tautech]

Fuck off Fungus

your sale hurts?

Not at all, quite busy actually.

Those that throw dirt should expect some to be thrown back.

[Martin72]

@Fungus:

https://www.eevblog.com/forum/testgear/rigol-dho804-test-and-compare-thread/msg5171130/#msg5171130

I don't know where you're going with this...

[ebastler]

@Fungus:
https://www.eevblog.com/forum/testgear/rigol-dho804-test-and-compare-thread/msg5171130/#msg5171130
I don't know where you're going with this...

I think this whole digression is based on a misunderstanding. When you wrote,

I wouldn't have the "problem" with the Rigol, of course.
But then I wouldn't have a bode plot either.

Fungus assumed you were claiming that the DHO900 "does not have a Bode plot" since that functionality is not working properly yet. So he felt compelled to point out that other scopes had bugs upon their release as well.

Whereas I understood that you were focusing on the DHO800, which obviously does not have a Bode plot at all. -- Maybe Rigol wil eventually implement Bode plots using external Rigol function generators, now that they have announced function generators (and DMMs) in a DHO-style package? Would be a nice incentive to collect them all...

[nctnico]

Fuck off Fungus
https://www.eevblog.com/forum/projects/project-yaigol-fixing-rigol-scope-design-problems/

Come on, man, get a grip. What kind of tone is that?

Trouble in kindergarten test equipment land  The toddlers are having a diaper fight.

[Fungus]

Fungus assumed you were claiming that the DHO900 "does not have a Bode plot" since that functionality is not working properly yet.

Yep.

I was just pointing out that most other 'scopes don't "work" on launch day either.

(maybe even "none of them"...  )

[Martin72]

They had nearly four years time to polish the function since it cames out for the MSO5000.
So please forgive my scepticism that something bigger is happening.
I would be delighted if I were wrong, you can take my word for it.

[Fungus]

So please forgive my scepticism that something bigger is happening.
I would be delighted if I were wrong, you can take my word for it.

I don't pretend to understand Rigol internally. If you've just spent 100 million (or whatever) on developing a chipset and bringing these new devices to market then it seems a no-brainer to put a couple of people on EEVBLOG to gather info and do a monthly firmware update for the first few months.

The few bugs that we've found in the DHO800 seem to me like they could be fixed with a couple of morning's work.

It's a mystery. 

[Martin72]

Today I tested the waveform updaterate.
Based on the first table of rf-loop measurements, with a sine of 600mVpp and frequencies from 1khz up to 40Mhz.
As you know, with rigol you can't make any settings regarding vector/dots or lin / sinx/x, it only works as it is.
Memory mode is set to Auto, beforehand I made sure that the aux output is actually set to trigger out.
The trigger out signal has an amplitude of 3.3V and a single pulse width of 100ns.
The updaterates appear as bursts with pauses in between, which was to be expected.



Things worth noting:
from 5ns to 500ns the rate remains almost constant, between 1µs and 5µs it drops only slightly and between 5µs and 20µs it shows the halving at every timebase change that I know of.
But then from 50µs it drops drastically from 4 digits to 2 digits, for whatever reason.
The early minimisation of the sample rate is also interesting, it is already reduced from 100µs.
The memory points remain constant at 625 kpts from 50 µs - only at 500 µs does it rise again to 781.25 kpts and then drop again to 625 - whatever triggers this, especially as 781.25 is a rather crooked number.
I will have a look at the ultra acquiremode, how to use it and then create a table from it.
First attempt was "unsuccessful", I also had max 20000wfms/s in ultra mode.
Can't be right, so I'll look for what went wrong.

[Martin72]

The few bugs that we've found in the DHO800

Whereby the emphasis is on found.
And on bugs, there are also things that could be done better.
And how easy it is to fix things, I have no idea, but the scope is still very young, there's still something to come.

[Martin72]

I will have a look at the ultra acquiremode, how to use it and then create a table from it.
First attempt was "unsuccessful", I also had max 20000wfms/s in ultra mode.
Can't be right, so I'll look for what went wrong.

I'm beginning to understand what this mode is supposed to represent and how to adjust it, I also wrote down the values earlier, but I still have to transfer them to a table.
From 5ns to 20ns it is a constant 42000 wfms/s.
At 50ns something happens, the numbers shoot up to actually 1 million, fluctuates a bit.
Between 50ns and 500ns it remains in descending 6 digits and between 50µs and 100µs there is this dramatic drop again.
And the strange 781.25Kpoints are also present again at 500µs.
More of this soon.

[Martin72]

See the video at 32:15:

I just did, it doesn't look nice, but I don't know the settings, nor are they revealed there.
And I switched off the video at the latest when the bodnar pulser is attached to the results and its comments.