Rigol's New DHO800 Oscilloscope unbox & teardown

[Martin72]

Has the user interface responsiveness improved over the DS1054Z? I owned a DS1054Z and really liked it... aside from the sluggish UI, which made me sell it for a Siglent.

No one can give you a reliable answer at the moment, because hardly anyone has the new model.
In that case, they would also have to have the DS1054Z for comparison purposes.
But I know someone who can do that when the new scope is not only available but also in his hands.

[Fungus]

No one can give you a reliable answer at the moment, because hardly anyone has the new model.

Dave's unboxing video has a few seconds of him twisting the knob and saying "It's responsive!"

(and it can zoom out!)

[ptluis]

Thanks. Sorry I was not at my computer - indeed then all anybody needs to do is flash their 800 series to 900 series firmwares. To then get the CAN bus decoding. So i don't see this as an issue in that case... sorry for any confusion here. (at least I am fine with that then, not so sure what the other guy was turning his nose up about).

Now you catch my attention, I seem to miss that DHO900 has CAN decoding, I want to buy a tablet scope with CAN decoding, micsig STO1004 was on my list, but this Rigol DHO800 hackable seems very interesting indeed.

[NE666]

this Rigol DHO800 hackable seems very interesting indeed.

Not to single out this particular member, as the same is being repeated by others as "fact".  However, perhaps a reminder is due that it is by no means *certain* yet that this (DHO800->900) can be done, at least not without negative effects.

It feels as if people have let the potential significance of this post slip past unheeded: https://www.eevblog.com/forum/testgear/rigols-new-dho800-oscilloscope-unbox-teardown/msg5062843/#msg5062843

Until enough people actually have scopes in hand and can have tested it thoroughly,  I'd advise being very conservative about purchasing an 800 series now if you must have 900 series functionality.

[Fungus]

this Rigol DHO800 hackable seems very interesting indeed.

Not to single out this particular member, as the same is being repeated by others as "fact".  However, perhaps a reminder is due that it is by no means *certain* yet that this (DHO800->900) can be done, at least not without negative effects.

We're not talking about hardware hacking here. CAN bus decoding is in the firmware somewhere, it's a question of enabling it on the DHO800. I predict the chances are very good that it can be done.

[NE666]

I predict the chances are very good that it can be done.

You mean, you're guessing / wishing.  Meaningful prediction requires data.  There's a difference.

[Martin72]

Dave's unboxing video has a few seconds of him twisting the knob and saying "It's responsive!"

Someone asks if the UI has become better/faster than the old rigol and you say "yes" - Based on a scene in a video ?
Not your serious or...

(and it can zoom out!)

Ehh....Wow ?

[Warhawk]

I am not able to read every single post regarding the new Rigols but let me share my thoughts on the bandwidth. I see a bit obsession with the sampling rate and the bandwidth. The Nyquist theorem (Shannon-Kotelnikov for folks growing up behind the iron curtain) is valid only in situations when the signal is restored using the sinc filter. Good that scopes have it by default. However, we also need to consider that the frequency response of the filter is not "brick-wall". The input stage will pass frequencies way above the corner frequency with just reducing the magnitude and increasing the group delay. The content after the fs/2 will be aliased. This happens always. The key is to keep levels of this aliased content below the resolution (simply said).
For the standard low-pass filter, we can already see the phase distortion kicking-in one decade below the cut-off. Oscilloscopes frequently use tricks such as equivalent sampling. This works, of course, only for repetitive signals that are stable over time.
Long story short, it is generally accepted that the sampling should be at least 10 times higher than the signal of interest. What I am trying to say is take it easy when you make your statements on sampling and the bandwidth. I just look at my old Infinium scope that is 500MHz/1GSa/s. There use to be even 1GHz scopes with only 100Msa/s, etc (Lecroy). Sometimes it really makes sense to have high bandwidth AFE even for relatively slow sampling ADC (1.25G/4 in case of DHOxxx). Especially if you work with single tone signals (but that's what spectrum analyzer is for).

PS: I am willing to discuss any error I made in my conclusions. I am just a human 

[tautech]

(and it can zoom out!)

Ehh....Wow ?

How many orders of magnitude ?
Not gunna be pretty with such limited memory when all channels are active.

[Fungus]

I predict the chances are very good that it can be done.

You mean, you're guessing / wishing.

Not at all.

Meaningful prediction requires data.  There's a difference.

The data is that it's not a hardware limitation and the firmware is wide open.

[Fungus]

Someone asks if the UI has become better/faster than the old rigol and you say "yes" - Based on a scene in a video ?
Not your serious or...

Yep. 

[Fungus]

Not gunna be pretty with such limited memory when all channels are active.

Oh, glub... tautech's found a horse and he's going to ride it into the ground.

5MPts / channel with all channels on isn't too shabby.

[nctnico]

Long story short, it is generally accepted that the sampling should be at least 10 times higher than the signal of interest.

Again, this statement has been generally un-accepted a long time ago. It stems from a time where sin x/x was hard to implement in a DSO and people thought a DSO should mimic an analog scope as much as possible. Nowadays users have a choice between less bandwidth + soft roll-off or maximum bandwidth nearly up to Nyquist with a sharp filter. And for sure the latter will have some aliasing but don't forget the probing solution typically has a filtering effect as well.

[2N3055]

I am not able to read every single post regarding the new Rigols but let me share my thoughts on the bandwidth. I see a bit obsession with the sampling rate and the bandwidth. The Nyquist theorem (Shannon-Kotelnikov for folks growing up behind the iron curtain) is valid only in situations when the signal is restored using the sinc filter. Good that scopes have it by default. However, we also need to consider that the frequency response of the filter is not "brick-wall". The input stage will pass frequencies way above the corner frequency with just reducing the magnitude and increasing the group delay. The content after the fs/2 will be aliased. This happens always. The key is to keep levels of this aliased content below the resolution (simply said).
For the standard low-pass filter, we can already see the phase distortion kicking-in one decade below the cut-off. Oscilloscopes frequently use tricks such as equivalent sampling. This works, of course, only for repetitive signals that are stable over time.
Long story short, it is generally accepted that the sampling should be at least 10 times higher than the signal of interest. What I am trying to say is take it easy when you make your statements on sampling and the bandwidth. I just look at my old Infinium scope that is 500MHz/1GSa/s. There use to be even 1GHz scopes with only 100Msa/s, etc (Lecroy). Sometimes it really makes sense to have high bandwidth AFE even for relatively slow sampling ADC (1.25G/4 in case of DHOxxx). Especially if you work with single tone signals (but that's what spectrum analyzer is for).

PS: I am willing to discuss any error I made in my conclusions. I am just a human 

ETS means that sampling was made at higher equivalent rate so it is not 100MSp/s sampling anymore.. So Nyquist (Shannon) criteria holds.
Downsampling signals with digital down converting IS domain of SA or SDR....
You can use it on scope as a deliberate technique. I think Performa01 did show some examples here on forum..
But scope should be able to handle it's own BW in full.
250MHz BW with 312,5MSp/s  is just no...
DSO800 is officially 100MHz BW and that is fine.. DHO900 claims 250 MHz BW... That is either a big problem or they enable AA filtering to 100MHz, but they need to explain and document that.

[dreamcat4]

PS: I am willing to discuss any error I made in my conclusions. I am just a human 

no i think as you correctly point out - the relationship between the quoted nominal inpuut (analog) bandwidth versus the sampling rate is indeed somewhat fungible matter. not least because on the analog side the nominal bandwidth is the -3db or half power point (or 1/rt2 whatever it is). so a higher analog bandwidth would reduce that attenuation level at the upper end of the total usable frequency range.

and also as you point out - that other factors such as the resolution, or the accuracy (of the adc), or the time wise jitter. things like this can differently impact the resultant noise or confidence onto specific waveform types, or specific measurement types. whether it is some fft analysis (or bode), or the whole multi sampling / repeating waveform analysis.

so that all makes sense. but what i am seeing here is also that there is this camp of people making the 10x generalization (broadly across purposes). and theres the people quoting the nyquist 2x upper aliasing limit (for some sine wave or whatever). then i suppose the 6x is meant to be a sorta 'in-between' those 2 extremes perhaps?

so looking here at the 312ms/s for the industry std analog -3db attenuation bandwidth... that is how much? ---> about 50mhz for a 6x factor? is that a reasonable working generalization for 4 channels?

ah but if we need only 2 ch then its 100mhz. that is what another guy has already said here earlier on in the thread. (to maintain that 6x factor). i believe these are fair assessments.

but a nicer assessment will be after proper reviews with the supplied probes, and a costing breakdowns, for the 800 versus 900 choosing. and those other included features too. because that also feeds into the overall value proposition. for the total sticker price.

as a false advertising - yes i agree its disingenuous claims. but also that it 'doesnt hurt' so long as you understand the truth. rather that instead the 900 seems worse value doe to being priced too much higher than it idealls should be nearer to 800 series somewhere inbetween.

[tautech]

5MPts / channel with all channels on isn't too shabby.

5yr old SDS1104X-E offers 7Mpts with all channels active or 14 with just 2 active.

[Fungus]

5MPts / channel with all channels on isn't too shabby.

5yr old SDS1104X-E offers 7Mpts with all channels active or 14 with just 2 active.

Actually I'm wrong... these go 50Mpts/25MPts/10Mpts with 1/2/3+4 channels enabled.

[iMo]

..
so that all makes sense. but what i am seeing here is also that there is this camp of people making the 10x generalization (broadly across purposes). and theres the people quoting the nyquist 2x upper aliasing limit (for some sine wave or whatever). then i suppose the 6x is meant to be a sorta 'in-between' those 2 extremes perhaps?..

Nope.. There are two arguments with 2 numbers basically:
a) the modern scopes require only 2.5xBW sample rate -  because the newest scopes have well made design
b) the sample rate should be 4-5xBW - because that has been a traditional recommendation

The "..the sampling should be at least 10 times higher than the signal of interest.." above is not related to the BW, but to the "signal of interest" - that is something different (for example your signal of interest could be 10kHz sine wave).

Simply the claimed 250MHz BW and the 312.5MSa/s is on first glance a suspicious combo, and as 2N3055 wrote above we need some more detailed clarification in that regard (it could be the scope switches the BW down to 100MHz automatically when operating with all 4 channels enabled, or something like that).

[Warhawk]

Long story short, it is generally accepted that the sampling should be at least 10 times higher than the signal of interest.

Again, this statement has been generally un-accepted a long time ago. It stems from a time where sin x/x was hard to implement in a DSO and people thought a DSO should mimic an analog scope as much as possible. Nowadays users have a choice between less bandwidth + soft roll-off or maximum bandwidth nearly up to Nyquist with a sharp filter. And for sure the latter will have some aliasing but don't forget the probing solution typically has a filtering effect as well.

Let's talk about the scope only. Probing is a completely different story that is likely less understood (I have the joy to work with IsoVu, occasionally).
What do you exactly mean that the user has a choice? How is the sharp filter implemented? What is generally believed to be sharp enough?
I have bit problems believing that scopes have a brick-wall filter as a part of the front-end. However, I have never analyzed any. What happens in software is irrelevant.

 

[Warhawk]

I am not able to read every single post regarding the new Rigols but let me share my thoughts on the bandwidth. I see a bit obsession with the sampling rate and the bandwidth. The Nyquist theorem (Shannon-Kotelnikov for folks growing up behind the iron curtain) is valid only in situations when the signal is restored using the sinc filter. Good that scopes have it by default. However, we also need to consider that the frequency response of the filter is not "brick-wall". The input stage will pass frequencies way above the corner frequency with just reducing the magnitude and increasing the group delay. The content after the fs/2 will be aliased. This happens always. The key is to keep levels of this aliased content below the resolution (simply said).
For the standard low-pass filter, we can already see the phase distortion kicking-in one decade below the cut-off. Oscilloscopes frequently use tricks such as equivalent sampling. This works, of course, only for repetitive signals that are stable over time.
Long story short, it is generally accepted that the sampling should be at least 10 times higher than the signal of interest. What I am trying to say is take it easy when you make your statements on sampling and the bandwidth. I just look at my old Infinium scope that is 500MHz/1GSa/s. There use to be even 1GHz scopes with only 100Msa/s, etc (Lecroy). Sometimes it really makes sense to have high bandwidth AFE even for relatively slow sampling ADC (1.25G/4 in case of DHOxxx). Especially if you work with single tone signals (but that's what spectrum analyzer is for).

PS: I am willing to discuss any error I made in my conclusions. I am just a human 

ETS means that sampling was made at higher equivalent rate so it is not 100MSp/s sampling anymore.. So Nyquist (Shannon) criteria holds.
Downsampling signals with digital down converting IS domain of SA or SDR....
You can use it on scope as a deliberate technique. I think Performa01 did show some examples here on forum..
But scope should be able to handle it's own BW in full.
250MHz BW with 312,5MSp/s  is just no...
DSO800 is officially 100MHz BW and that is fine.. DHO900 claims 250 MHz BW... That is either a big problem or they enable AA filtering to 100MHz, but they need to explain and document that.

I believe we both know how ETS works. Does the new DHO series support it?

250MHz BW with 312,5MSp/s  is just no...

I am on the same page but I wonder how many customers for this product need this bandwidth.

[Martin72]

Simply the claimed 250MHz BW and the 312.5MSa/s is on first glance a suspicious combo, and as 2N3055 wrote above we need some more detailed clarification in that regard.

Plus why 250Mhz ?
Only for marketing arguments ?
Top model 4000 it´s bandwith drops down to 400Mhz when all channels are active, so it´s logical that the one adc model DHO1000 got max 200Mhz bandwith.
I wish the specs of the centaurus chipset were avaible.

[TomKatt]

I'm wondering what the applications are for people in the target market of scopes like the DHO800 that even 100MHz+ BW would be a deciding factor?  The more I look at these new scopes I think it's a miracle to see such a feature set (plus the possibility of battery power) available at under $400...  I don't think things like 12bit in itself make this scope vastly superior to existing models, but when you wrap the whole package up it's certainly very attractive, especially for enthusiasts looking for their first 'real' scope.  That the sampling rate drops down to 312M on 3 channels doesn't reduce it's value in  it's target market - so long as that limitation is clearly communicated.

EDIT - consider that I think the SDS1104X-E is still $100 more, and the SDS1104X-U with a single ADC setup similar to how the DHO800 is designed is the same price.  So it seems like a good comparison for the DHO800 would be the SDS1104X-U, and I'd have to say that if I was looking to purchase a scope and my budget was $400, this DHO800 would be a real contender.

[nctnico]

Long story short, it is generally accepted that the sampling should be at least 10 times higher than the signal of interest.

Again, this statement has been generally un-accepted a long time ago. It stems from a time where sin x/x was hard to implement in a DSO and people thought a DSO should mimic an analog scope as much as possible. Nowadays users have a choice between less bandwidth + soft roll-off or maximum bandwidth nearly up to Nyquist with a sharp filter. And for sure the latter will have some aliasing but don't forget the probing solution typically has a filtering effect as well.

Let's talk about the scope only. Probing is a completely different story that is likely less understood (I have the joy to work with IsoVu, occasionally).
What do you exactly mean that the user has a choice? How is the sharp filter implemented? What is generally believed to be sharp enough?
I have bit problems believing that scopes have a brick-wall filter as a part of the front-end. However, I have never analyzed any. What happens in software is irrelevant.

Just look at some of Dave's teardowns. In some you'll see an LC filter between the front-end and ADC with at least a 3rd order response (and addition to the filtering that already happens inside the frontend).  Remember that the filter response also aliases beyond the Nyquist frequency. What happens in software is very relevant; in the end the purpose of an oscilloscope is to visualise a signal so our brains can understand what it looks like in the circuit. A DSO brings a lot to the table in that respect.

[iMo]

..
I wish the specs of the centaurus chipset were avaible.

The Centaurus chipset should be 2GSa/s, afaik.. Why 1.25GSa/s in 800/900 then? Perhaps the lower grade FPGA in 800/900 limits the sample rate??

[Martin72]

  So it seems like a good comparison for the DHO800 would be the SDS1104X-U, and I'd have to say that if I was looking to purchase a scope and my budget was $400, this DHO800 would be a real contender.

Absolutely and therefore I´ll compare the DHO804 not only to the DS1054Z, but also with a SDS1104X-E.
Having 12 bit is not really a buying argument in this pricerange, it just makes things more complicated to handle.