Oscilloscope question, Rigol DS/MSO1000 vs DS/MSO2000

[Professzore]

Dear Forum Readers,

First of all, I bet, it is the n+1th question regarding scope selection. Sorry for the repeat. I was trying to find a relevant source, but every answer I found -- yet -- were not exactly the answer for my questions. Let me explain.

I've made huge research about the topic, so I (hopefully) clearly understand the difference between memory depth, sample rate, analogue/digital (DS/MSO), input impedance. But these are technical specs only. The question is: how do they affect the usability of the scopes for my requirements (if they actually do). Here are my most important aspects:
-- prototyping with mbed platforms (max. CPU clock rarely over 120 MHz), and sometimes with Arduino (5V/3.3V logic level converting is a usual process here with no any kind of debugging possible -- this is one of the main reasons to buy a scope),
-- heavy use of communication protocols (mostly i2c, spi, later RS486, CAN, USART/UART, others just occasionally), decoding and signal integrity inspection are all must, as well as analisys of signal correlation between analogue signal and serial data,
-- analogue circuit debugging (supply voltage problems for example, so whatever a scope is used),
-- I have only a limited space on my bench, so smaller is better,
-- I can afford both, better price/value is a higher preference than price itself, so if an MSO2072A may better meet my preferences than an MSO1054A, I can pay the difference.

Based on every second-hand experience I've collected (writen tests, youtube videos from Dave and others etc.), I'm still not sure, because it's not clear if within the above mentioned environment / user preferences...
-- the additional memory depth (54 Mpts) of 2000s is a real benefit,
-- the additional sample reate (2G Sa/s) is a real benefit,
-- the MSO versions is really needed and can not be substituted with a separate LA, where the decoding are mostly done with the LA, and signal integrity checked by the scope (analogue + decoding),
-- the 4 channel analogue capability of 1000s are a real benefit (except, for example, SPI decoding with three probes and one additional probe for analogue testing).

Now I'm at the decision point: MSO1000 / DS1000(+LA) / MSO2000 / DS1000(+LA)? (If separate, which LA?)

Any help, advice, notice are warmly welcome, as well as you have some more questions to clear your answer, please, feel free to ask.

Thanks for your help.

Regards from Hungary,


Tamás

[rstofer]

Dave has done a video on whether MSO is worthwhile compared to a more conventional logic analyzer.  You may have seen it.



I don't want to give away the surprise but I believe he comes out in favor of the conventional logic analyzer.

On the topic of Logic Analyzers, if it doesn't have an external clock input, it can't do state analysis.  To me, that is the most important feature of a LA.  If it can't do state analysis it is useless.

Now that I can decode serial streams on a scope, the requirement for a logic analyzer is reduced.

I work with mbed processors quite a bit.  Arduinos too...  Regardless of the clock frequency, nothing is going to come out of the pins at anywhere near that rate.  I2C is generally clocked at around 400 kHz, SPI might go to 10 MHz, RS232 to a mere 115200 baud, and so on.  This won't challenge a 100 MHz scope.

I am more concerned with the number of input channels than I am with super high bandwidth.  I really wanted 4 channels because that is the only appropriate way to decode SPI.  Trigger on CS', grab SCK, MISO and MOSI - decode the data packet.

There are a number of scopes that can do the job at these frequencies.  My personal choice was the Rigol DS1054Z (about $400) because I can hack the features and wind up with 100 MHz bandwidth and decoding of RS232, I2C, SPI and Parallel (which I haven't tried).

Dave has done a number of videos regarding these low end scopes and you have probably seen them:


You will find that this scope has its supporters (I am one) and its detractors.  The problem is, to get a better scope you need to spend a lot more money.

I have only had the scope for a couple of months so it hasn't had time to really irritate me.  For the previous 12 years, I used an old Tektronix 485 that I picked up on eBay for $200.  It just displays wavy lines and has no advanced features at all.  The 350 MHz bandwidth is comforting although not required.  The point is, even with this limited feature set, I was still able to build stuff.  Troubleshooting is more about technique than equipment.   Sure, it's easier with better gear but it's possible with even limited equipment given a little bit of thought.

In the end, is a 2000 series scope going to outperform a 1000 series scope?  Probably!  But don't underestimate the advantage of having 4 channels.  To me, that was the most important criteria when I decided to upgrade.

[Professzore]

Hi,

Thanks for your answer. It confirms my prelilminary idea of the simpler way: 1054 + separate LA.

I've seen both (and many more) videos of Dave recently.

On the topic of Logic Analyzers, if it doesn't have an external clock input, it can't do state analysis.  To me, that is the most important feature of a LA.  If it can't do state analysis it is useless.

May I ask you to list a few examples of this logic analyzers, please? I made a research, but found that none of the well known vendors (for example, Saleae) supports external triggering.

Thanks again.

Tamas

[rstofer]

Hi,

Thanks for your answer. It confirms my prelilminary idea of the simpler way: 1054 + separate LA.

I've seen both (and many more) videos of Dave recently.

On the topic of Logic Analyzers, if it doesn't have an external clock input, it can't do state analysis.  To me, that is the most important feature of a LA.  If it can't do state analysis it is useless.

May I ask you to list a few examples of this logic analyzers, please? I made a research, but found that none of the well known vendors (for example, Saleae) supports external triggering.

Thanks again.

Tamas

I hope one off the other folks can help.  My logic analyzer is based on an FPGA and a project
http://www.sump.org/projects/analyzer/

It certainly isn't elegant, laying all over my workspace but it works very well.  I did have some problems using the UI as Java was revised over the years.  I haven't used it in 2 or 3 years so I'm not sure how it will work out next time I need it.

Having 32 channels plus a trigger is very nice.

I also have this unit which uses the Sump UI:
https://sigrok.org/wiki/Openbench_Logic_Sniffer

It can also use the SigRok UI which might be easier to set up but I haven't tried it with the Sump style LAs.  Maybe I should do that...

You would want to add the Buffer Wing because only 16 of the 32 channels are buffered on the board itself.  This should easily handle a 50 MHz state clock but the spec reads '?'.

[Professzore]

Hi,
Thanks for the reply again.
Hard decision indeed. :-)
I was trying to find a comprehensive review of DS2000 from Dave, but I've failed. Nothing except a short (25 minutes) playaround. Any idea?
Thanks,
Tamás

[JPortici]

the project yaigol thread
i also went for the 1054z because no money for better toys for the playground at home, but i'd never ever give a cent more to rigol for their scopes. in my opinion they are not worth it. the GDS-2000E series from GWi could be a very interesting alternative