Rigol's New DHO800 Oscilloscope unbox & teardown

[Fungus]

I'm certainly not going t buy it just because it has a 12bit ADC.

Neither am I.

I'm buying it for the form factor, the user interface, and, yes... the fact that it has a very good front end and high resolution ADC (much better than 8 bits) all for less than 400 Euros.

It also looks like it will be extremely hackable and fun to mess around with. It also has windows, perspective views, graphical front end setup, and all sorts of neat tricks.

[Fungus]

This thing is badly in need of a good and thorough review.

There's been reviews of the HDO4000 series. This is basically the same thing, but smaller.

[DEiE]

This thing is badly in need of a good and thorough review.

Yeah I'm also eagerly awaiting the first impressions, and especially how it stacks up against the SDS1104X-E.

I'm a beginner hobbyist who is looking into buying their first scope, because I would like to be able to visualize the circuits I'm playing around with (mostly Arduino stuff, beginner circuits, etc.). I initially set my eyes on the SDS1104X-E and was about to purchase that one considering it's 10% off in Europe this month, when this one popped up.

I know on one hand that for my use it doesn't really matter which one I would go for because I'll probably only use a fraction of what it can do.
On the other hand, it would feel quite bad if after saving up and buying the Siglent, it gets dethroned by the DHO800 as the best bang-for-buck scope for hobbyists a month later (especially if the early discoveries turn out to be true and the bandwidth and memory depth of the DHO804 can be easily uphacked to those of the DHO914/DHO924).

[Howardlong]

Once you've tried a touch screen (or mouse) then using a twisty knob for navigation is a complete joke. I went from a Rigol DS1054Z to a Micsig and I'd never go back even though the Rigol can do more things.

I’m almost completely the opposite: I have two Micsigs, one with the knobs, one without. I also have an MSO1074Z. I always prefer using the Rigol over the Micsigs, I just don’t find working with touch screens works particularly well for me.

The same applies to the other touch screen scopes I use, MSO5000, SDS2000X+ and MDO3104T: if I can do it with the knobs, I prefer to use the knobs. I do have a wireless mouse hooked up to the MDO3104T which occasionally comes in useful.

I also have an early touchscreen scope, an MSO8104A, at my soldering workbench. You have to use a keyboard and mouse or touchscreen for anything but primary functions. I can’t remember ever using the touchscreen. I very recently replaced it with a 2nd hand MSO6104A, with no touch screen, keyboard or mouse, and consider my life improved!

[2N3055]

Once you've tried a touch screen (or mouse) then using a twisty knob for navigation is a complete joke. I went from a Rigol DS1054Z to a Micsig and I'd never go back even though the Rigol can do more things.

I’m almost completely the opposite: I have two Micsigs, one with the knobs, one without. I also have an MSO1074Z. I always prefer using the Rigol over the Micsigs, I just don’t find working with touch screens works particularly well for me.

The same applies to the other touch screen scopes I use, MSO5000, SDS2000X+ and MDO3104T: if I can do it with the knobs, I prefer to use the knobs. I do have a wireless mouse hooked up to the MDO3104T which occasionally comes in useful.

I also have an early touchscreen scope, an MSO8104A, at my soldering workbench. You have to use a keyboard and mouse or touchscreen for anything but primary functions. I can’t remember ever using the touchscreen. I very recently replaced it with a 2nd hand MSO6104A, with no touch screen, keyboard or mouse, and consider my life improved!

Hehe, we people are different, who knew .... 

I personally am somewhere in the middle: I have Micsig STO (one with the knobs) MSOX3000T, SDS2000X HD and SDS6000 H12Pro and few Picos..

On all touchscreen scopes I use combination: menus, virtual keyboard etc on the screen and setting some stuff on knobs (vertical, time base, cursors).
Horizontal zoom position I use screen for coarse moves and knob for fine...
I do that on all the scopes, whatever fells more natural for operation.

On MSO3000T I use it all:  soft keys around the screen, touch and knobs...

I personally would not like to use only knob based or only touchscreen scope if I had a choice.
And before someone comments, on Picoscopes you have touch, keyboard, and mouse knob that combined works... I run it on large touchscreen BUT have mouse connected. I need that mouse wheel.

[NE666]

if this Rigol with it's 12 bit ADC still has a 3% accuracy for on screen RMS measurements, then that's pretty atrocious in my view.

I think you have unrealistic expectations of oscilloscopes, if you regard them as precision measuring instruments.  You could spend 4-5 times the amount on a current A-brand scope and not achieve better.

with it's 12 bit ADC

resolution != accuracy

[Howardlong]

Hehe, we people are different, who knew .... 

Exactly!

[TomKatt]

if this Rigol with it's 12 bit ADC still has a 3% accuracy for on screen RMS measurements, then that's pretty atrocious in my view.

I think you have unrealistic expectations of oscilloscopes, if you regard them as precision measuring instruments.  You could spend 4-5 times the amount on a current A-brand scope and not achieve better.

with it's 12 bit ADC

resolution != accuracy

That seems reasonable.  Which then makes me curious how much that improved resolution actually benefits the user.  Because is the resolution can't really be conveyed through the wave image on the screen - and the measurements are still not 'accurate' - then where is the real benefit? 

I'm guessing a 12bit will be more 'accurate' than an 8bit scope, even though neither is good enough as a precision instrument?

EDIT - one area where I expect more bits does make a big improvement is that frequently signals are not measured at their full scale, so many bits go unused and the displayed waveform is based on only a portion of the bits available - in that case, you probably get more bits on a higher resolution scope?

[2N3055]

if this Rigol with it's 12 bit ADC still has a 3% accuracy for on screen RMS measurements, then that's pretty atrocious in my view.

I think you have unrealistic expectations of oscilloscopes, if you regard them as precision measuring instruments.  You could spend 4-5 times the amount on a current A-brand scope and not achieve better.

with it's 12 bit ADC

resolution != accuracy

That seems reasonable.  Which then makes me curious how much that improved resolution actually benefits the user.  Because is the resolution can't really be conveyed through the wave image on the screen - and the measurements are still not 'accurate' - then where is the real benefit? 

I'm guessing a 12bit will be more 'accurate' than an 8bit scope, even though neither is good enough as a precision instrument?

Even DHO4000 has only 2% ...

It depends on the class of the scope.
Some 12 bit scopes will have tighter DC accuracy specifications.

SDS2000X HD is more expensive but has :
0.5 mV/div ~ 4.95 mV/div: ±1.5 %;
5 mV/div ~ 10 V/div: ±0.5 %

SDS2000X HD also have much larger offset range...

PicoScope 4262 (16 bit) has best accuracy of 0,25%, worst 2%, depending on range...

For best results, datasheets should be analysed and details worked out..

[dreamcat4]

If i understand the purpose / value of 12-bit (which i don't profess i do, so please correct me here)...

For a higher bandwidth 10-bit scope you are analyzing waveforms, the more bandwidth then it can resolve those higher freqs. And you are in effect trading lower ADC bits for higher time base resolving capability.

Wheras on 12-bit (at lower rates, such as these ones, like HDO800 we are discussing here). You are analyzing a lower freq waveform. Perhaps very lower Hz down into Kilohertz... then you get nicer resolving image on the screen. That is 12-bits = 4096 vertical scanlines for the maximum peak-2-peak adjustment. Rather than a 10-bit which would (under zooming the vertical) - that would resolve to 1024 scan lines of resolution.

But that's just the ADC part. Within the longer pipeline, including the probes, and the analoge input amplifier and all that.

Then behind the ADC is the memory aquisition storage and firmware. Whatever it's doing to either not-store, or to mess up in the Math functions. To not take proper advantage of what the ADC is giving.

And assuming noise floor has to be below the resolving capability of the ADC.

This picture paints only 2 words:

PROPER REVIEW

[TomKatt]

Considering that scopes are not precision measurement devices, and seem to be most useful for visually identifying signal issues and time based analysis, I have to admit that the whole 12bit vs 8bit thing has me confused.  Technically, I get it and my gut says more bits is better.  But I feel the same way with audio and a lot of people say there isn't a huge difference in 24bit vs 16bit audio (I've never bothered to do any real A/B testing).

In the scheme of things, at this price class I guess you just can't argue with getting 'more' for less.  And battery power possibilities make floating measurements so much safer (for the equipment lol).  As a consumer I'm thrilled to see this coming, but I don't see the 12bits being the primary motivator.

EDIT - it's also amazing how few components now make up the analog front end...  No doubt there is some cost savings there.

[TomKatt]

Who buys a DHO814 when the DHO804 is hackable?

I searched around a bit and haven't been able to find information on that hack...  Do you have a link?  Thanks.

[DEiE]

Who buys a DHO814 when the DHO804 is hackable?

I searched around a bit and haven't been able to find information on that hack...  Do you have a link?  Thanks.

Earlier in the thread, the following was mentioned, although it's less of a hack and more of a complete reflash:

Thank you,hubert.
My DHO804 is DHO924 now.
Two extra drams are only for LA.It can be unpopulate if doesn't use the LA .
The bandwidth is increased successfully.it has 800ps rise time now.
And the memory depth is 50M now.

[Fungus]

where is the real benefit? 

It's like having "hires" mode on all the time but without the loss of bandwidth.

I'm guessing a 12bit will be more 'accurate' than an 8bit scope, even though neither is good enough as a precision instrument?

Oscilloscopes are for looking at  the shapes of things, not for measuring them to many decimal places. More bits = more precise capturing of the shape.

EDIT - one area where I expect more bits does make a big improvement is that frequently signals are not measured at their full scale, so many bits go unused and the displayed waveform is based on only a portion of the bits available

Watch this video at time 23:11:
https://youtu.be/gremHHvqYTE?t=1391

nb. That video also clearly shows that the Rigol chipset has more bits than a 10-bit R&S.

It's the exact same ADC/frontend as the DHO800 so results should be the same. 

[tv84]

More bits = more precise capturing of the shape.

... in vertical terms.

[Fungus]

Who buys a DHO814 when the DHO804 is hackable?

I searched around a bit and haven't been able to find information on that hack...  Do you have a link?  Thanks.

Earlier in the thread, the following was mentioned, although it's less of a hack and more of a complete reflash:

If the 924 firmware works on a 804 then it's a very safe bet that the 814 firmware will work. It might even be that the 4000-series firmware works and we can do power analysis on a $400 'scope.

It's also obvious that the system is wide open to hacking. It will probably be done via USB within 30 minutes of the first wave of deliveries at the end of this month.

(And before you ask: No, Rigol is NOT going to do anything about this ... hacking is their bread and butter)

[madires]

Considering that scopes are not precision measurement devices, and seem to be most useful for visually identifying signal issues and time based analysis, I have to admit that the whole 12bit vs 8bit thing has me confused.  Technically, I get it and my gut says more bits is better.  But I feel the same way with audio and a lot of people say there isn't a huge difference in 24bit vs 16bit audio (I've never bothered to do any real A/B testing).

The question is if the DSO can make good use of the higher ADC resolution. In general, with a higher resolution you can see smaller changes or get a higher dynamic range. The latter is the benefit of 24bit audio, especially for classical music (if done right). At the same time it's not hard fit less demanding music into 16 bits without clipping, but some studios manage to screw up even that (loudness wars). More bits can help if used properly.

[Fungus]

Even DHO4000 has only 2% ...

You can probably do better than that if you warm it up properly and do a self-cal just before the measurement. 

[Bicurico]

More than 16bits in audio is normally used to have some margin for mixing and digital editting. This way, the rounding errors are still above the 16bits resolution.

In an oscilloscope, 8 bits resolution means that a signal interval from 0V to 5V can be represented in steps of 0.02V, i.e. you can differenciate a 0.02V difference in your signal.

If you have 12 bits, that equates to: 0.0015V (approx.), which is more than 10 times better. The thing is: if you have 12bits, that gives 4096 different states as opposed to 256. But the screen only has a vertical resolution of what, 800 pixel?

I would say that the effective, usable benefit of 12bits vs 8bits in a low spec oscilloscope should be something like "twice the resolution" on screen.

Definitely better, but not something that will allow you to measure a signal which you wouldn't be able to measure with 8bit.

[Fungus]

The latter is the benefit of 24bit audio, especially for classical music (if done right). At the same time it's not hard fit less demanding music into 16 bits without clipping, but some studios manage to screw up even that (loudness wars). More bits can help if used properly.

Please watch this three times before posting anything like that again:

[RAPo]

In that case/my case with a 924S on its way (so need for hacking higher specs to a lower specced model), I propose two hacks:
Hack 1: a button that closes all surrounding menus/info panels around the main display and enlarges the display to the full 7-inch,
Hack 2: the possibility to adjust the timebase in a VCO-like manner, just like in an analog scope like the PM3230.

It's also obvious that the system is wide open to hacking. It will probably be done via USB within 30 minutes of the first wave of deliveries at the end of this month.

(And before you ask: No, Rigol is NOT going to do anything about this ... hacking is their bread and butter)

[Fungus]

In that case/my case with a 924S on its way (so need for hacking higher specs to a lower specced model), I propose two hacks:
Hack 1: a button that closes all surrounding menus/info panels around the main display and enlarges the display to the full 7-inch,
Hack 2: the possibility to adjust the timebase in a VCO-like manner, just like in an analog scope like the PM3230.

I'm sure you'll be very popular if you do those two things for us.

[dmulligan]

In that case/my case with a 924S on its way (so need for hacking higher specs to a lower specced model), I propose two hacks:
Hack 1: a button that closes all surrounding menus/info panels around the main display and enlarges the display to the full 7-inch,
Hack 2: the possibility to adjust the timebase in a VCO-like manner, just like in an analog scope like the PM3230.

I'm sure you'll be very popular if you do those two things for us.

Previous posts in this thread have indicated that the code is not obfuscated and therefore decompiled easily.  This might be the most modable scope ever.   If Rigol continues to allow the community to "hack" these scopes or better yet embraces the community we may have the futures most popular scope family available here.

Rigol: Do you remember what opening up iPhone to 3rd party app development did for iPhone sales?

[RAPo]

Hack 2 needs both hardware and software adaptions.
I wonder if we can just do that.

Previous posts in this thread have indicated that the code is not obfuscated and therefore decompiled easily.  This might be the most modable scope ever.   If Rigol continues to allow the community to "hack" these scopes or better yet embraces the community we may have the futures most popular scope family available here.

Rigol: Do you remember what opening up iPhone to 3rd party app development did for iPhone sales?

[Fungus]

Hack 2 needs both hardware and software adaptions.

2 probably involves the FPGA.

I'm not sure it's really needed in a DSO. On a CRO it was good to be able to maximize a wave on screen while keeping the trigger point visible. On a DSO you can zoom in and pan left/right and the trigger will work just the same.