One more "MSO old vs new" discussion

[toli]

Hi All,

While I do have access to a good quality MSO at work, which I can borrow for my home use when it isn't needed in the office, I think of getting another simpler unit for home. Sometimes you need it "now", and not when it becomes free. Therefore, I've started looking at my options.

The "ideal" unit for my needs would have 200MHz BW, 4 analog channels, large color display, high waveform update rate, serial trigger and decode, intensity graded display, and <1000$. However, I'm aware of the fact that's probably not an option, so I've started looking for what I can get.
There's always the "standard" 54645D, and even better the 54622D. The 54622D does only 100MHz with 2 channels (and at 200MSPS its probably not really good enough for 100MHz, maybe more like 50-70MHz), it has a single color display which isn't as "nice to use" with multiple channels displayed, the waveform update rate isn't high in modern terms, and it doesn't do serial decode.
On the other hand there are more modern chinese units such as Rigol and seems like Siglent have something to offer too with the latest 1000X series. While I'm not much familiar with Siglents newest models (and will appreciate some input from these who can comment on these), the Rigol seems to be a more versatile unit than the 54622D.
The MSO1074Z, limited to 70MHz, has 4 channels (although when MSO is active limited to 2/3 depending on number of active digital channels), does have the color display (7" isn't large though), has higher sample rate, with 500MSPS when 2 analog + digital channels are active, slightly higher update rate, has serial decode option. Another huge benefit of the modern units is the smaller size which is always a great thing as my work bench tends to become "not so organized" while working, and more free space is always welcome.

So on paper at least the MSO1074Z is a much more capable tool, and would be the right choice despite going for ~2-3X more. The serial decode + deep memory seem to be free upgrade at this time which is also great. However, I was unable to find much practical info about user reviews or an in depth review of this unit (MSO, not DS). Even more interesting is what other people here in the forum have to say about the comparison of these 2 (and a comparable siglent too). Is the "dry" info in the datasheet really telling the whole story and I should get the MSO1074Z and be done with it, or is there something else to consider.

[JPortici]

-Rigol doesn't decode from memory, but only from what's currently on screen, so if you zoom out too much the decode doesn't work, data is heavily undersampled (you can see at most 6 UART packets). if you zoom too much in the decode doesn't work because you are not seeing at least one packet
-do you really need a scope with MSO? or can you live with a logic analyzer?
for example the dsLogic pro costs about 100 euros (depending on which ebay seller you go to) but i find it to be a very capable tool if i need to dump a lot of data or see logic states. however i don't remember if you can trigger on bus errors..
I personally use one or the other for different occasions.
there are times where i need to catch bus errors and usually the scope i have at work is better at that, or i have to see the correlation between the serial bus and some analog parameters, or if i wand to rule out any problems in the analog side (waveform shape, level) altough in that case i wouldn't use the MSO channels, of course

[toli]

Thank you for the reply. I do need (or should I say very much prefer since I can get an MSO from work if I have no other option) an MSO, as from my experience in the past, this is the right tool for my needs. I don't really needs to dump lots of data and decode it (when I do I can do this with a USB LA - I've never needed to decode huge chunks of data in parallel to analog signals), I mainly need the ability to do a correlation of digital edges/commands to the analog signals (in mixed signal system where I need to see that some controls are arriving at the right time relative to other signals in the system, or measure the response of a device to digital programming etc).
While I can get away with a separate LA by using the sync out of the scope for trigger, it is typically not as intuitive and quick for debugging. Also, this can be a more limiting approach in the triggering options compared with a real MSO.
I am not completely against it though, and will consider it if I can get a much better scope + LA for the price - since I can borrow the MSO from work if I'm in some real trouble and must use a very capable MSO
But if the performance difference isn't that big on the analog channels for the same price, an MSO would be preferred.

[nctnico]

Look at the GW Instek MSO2000E series. Not within your budget but it has less compromises and bugs.

[Fungus]

I mainly need the ability to do a correlation of digital edges/commands to the analog signals (in mixed signal system where I need to see that some controls are arriving at the right time relative to other signals in the system, or measure the response of a device to digital programming etc).

You don't need masses of analog bandwidth for that, it doesn't matter if the rise time is 1ns or 3ns so long as it triggers on the rising edge.

It also doesn't matter much if the LA only decodes what's on screen or not.

[tautech]

The "ideal" unit for my needs would have 200MHz BW, 4 analog channels, large color display, high waveform update rate, serial trigger and decode, intensity graded display, and <1000$. However, I'm aware of the fact that's probably not an option, so I've started looking for what I can get.......

and seems like Siglent have something to offer too with the latest 1000X series. While I'm not much familiar with Siglents newest models (and will appreciate some input from these who can comment on these),

So let's expand on what you've looked at thus far and give offer some info so you can maybe make an 'informed' decision.

Latest Siglent 4ch MSO capable models are SDS1004X-E not just X as they're a previous and only 2ch series.
With the requirements you've listed a SDS1204X-E is a nearly your 'ideal' fit and depending on your final choice you can add the 16ch MSO option later or not.
For a comparison of how a non-MSO version might stack up check this chart prepared by rf-loop:
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/

What you need be aware that is easily missed is these units have dual 1Gsa/s ADC's and dual 14Mpts of memory supporting each. There's a ton of factual info later in the above thread from early Dec when they were released.
An excellent review thread is here:
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/

Most features, functionality, GUI and UI are based on the existing Siglent models SDS1202X-E, SDS1000X and SDS2000X and further features such as Search, Bode plot, Webserver and WiFi support have all been added.
Other than what's in the above threads, further info can be found or downloaded here:
https://www.siglenteu.com/digital-oscilloscopes/sds1000x-e-series-super-phosphor-oscilloscopes/

Anything further about them you want to know please ask.

[toli]

Fungus, you are obviously correct, this can be done with a low BW unit. However, this isn't the only thing I do with the scope, this is what I need the MSO capability for. For some of the things a higher BW will be needed. 200MHz looked like a good compromise between what I need, and what I could hope to get on a budget. When I need much faster scopes I take it to work/ the lab at the university. The "requirements" I've listed are what I would like to have at home, that why I've named them "ideal" in quote marks

tautech, thank you for the detailed comment and info. I will have a closer look at these models. I might end up with more questions, but I think I should first review the info you've linked.

[nctnico]

It also doesn't matter much if the LA only decodes what's on screen or not.

That matters a lot because if a scope decodes only what is on screen then you cannot relate the messages in time (in respect to the trigger point or to eachother) and you cannot zoom into a message because as soon as the beginning is off-screen the decoding will dissapear. Been there, done that and using a 'decode only what is on screen' scope sucks badly for any real work.