Help choosing DSO :) MSO5000 ?

[nctnico]

In the end neither are 'perfect' allround. AFAIK both Rigol and Siglent only decode what is on screen. If you want to get something more polished then look at the R&S RTB2004 for example.

The RTB2004.. Isn't that a 2500 USD scope ? Or did I find the wrong model ?  If yes... Let me remind you that I am trying to justify a jump from 1300 USD to 1850 USD (or avoid the jump) 

I know. I just wanted to point out that there are other choices around. Car analogy: You are now looking at Hyundai and KIA but they aren't at the luxury and comfort levels BMW and Mercedes offer.

[nubinstanley]

First image some signal. It looks pretty much very noisy. MSO5000 user will say: "you see, I see same noise.."


Then you zoom in and realize all of that is a signal:


There are no noise reduction techniques that preserve full bandwidth AND decrease noise on a signal that is not strictly repetitive. You either give up bandwidth (i.e you get a 1Mhz scope) or noise is there.

A side note: this is with a scope on 1mv/div natively (10mv with 10x probe) riding on top of 11,5V signal. DC coupled...
Ruminate on that. That is what a low noise scope front end can do...
With a 500MHz+ full bandwidth... 

Again, forgive my ignorance.. I do understand the advantages of having a low noise scope.

But, could you give an example situation where I might come across such a signal ? Maybe then I'll be able to make a judgement - whether or not such a use-cases would ever be one that I'll need to consider. Also, are you implying that such a measurement would have been impossible on the Rigol ?

[nubinstanley]

Thank you for taking the time to share all this information! My confusions is also between the MSO5000 and the SDS2104X+. I can justify a jump from the SDS1104X-E to the MS5000 but was confused about a jump to the SDS2104X+.

From a usability perspective, how is the scope ? UI, protocol decoding... ?

Yes, it can be confusing and everyone has their perspective.  The scope discussion comes up regularly, the usual suspects reply (including me) and no conclusions are reached.  All as it should be!

I have the SDS1104X-E and SDS2104X+, I don't have any MSO5000 but I did have the earlier DS1054Z.  You mentioned not having used an oscilloscope in a while.  The entry-level-and-just-above scope space is an extremely competitive area right now and significant advances have happened recently and continue.  The result is that you can get a lot of bang-per-buck now, but perhaps with some occasional rough edges.  You mentioned you will work with microcontrollers and want decoding and such.  If that is all or the vast majority of what you intend to do with it, then I'm probably not the best person to help you pick a scope.  Otherwise, I'd tell you that the less-expensive SDS1104X-E is probably a pretty clear choice--it is much cheaper, it works very well and the bugs are mostly worked out and it really has a quite nice feature set when hacked.  What it does not have is any sort of signal generator, a large screen (it's OK, just not huge) or a good LA option.  It does have an LA option and you can hack the license, but the hardware is a bit expensive and can't easily be DIY-ed.  The price point allows you to upgrade later with fewer regrets once you spend some time with a modern DSO and figure out all the things that it can do.  And in today's rapidly advancing market, I don't think it makes sense to buy capability today that you might not be able to use for 3-5 years--or never.  And you might miss out on something you didn't know you needed until you try and do it and it isn't there.

 

[nubinstanley]

Just another day at the office...

@nubinstanley,

Buy the X-E. That's all you ever need in the near future and these things aren't cheap in your region.

If you find that you would need the MSO5000 for something, it will be a good exercise to find a way to overcome its inexistence.

 

Yeah... maybe its more of a "want" than a "need". Anyhow, thanks for the advice. 

[nubinstanley]

In the end neither are 'perfect' allround. AFAIK both Rigol and Siglent only decode what is on screen. If you want to get something more polished then look at the R&S RTB2004 for example.

The RTB2004.. Isn't that a 2500 USD scope ? Or did I find the wrong model ?  If yes... Let me remind you that I am trying to justify a jump from 1300 USD to 1850 USD (or avoid the jump) 

I know. I just wanted to point out that there are other choices around. Car analogy: You are now looking at Hyundai and KIA but they aren't at the luxury and comfort levels BMW and Mercedes offer.

 

[nubinstanley]

Anyhow, first of all, let me apologize for triggering another "noise war". 

I am happy to have received varying perspectives and now I think that the ball is in my court. Thank you all, for all the inputs/advice. Maybe I'll drop a message here after I finally buy a scope (only if I buy something that I actually need and not what I want  )

[nctnico]

Anyhow, first of all, let me apologize for triggering another "noise war". 

I am happy to have received varying perspectives and now I think that the ball is in my court. Thank you all, for all the inputs/advice. Maybe I'll drop a message here after I finally buy a scope (only if I buy something that I actually need and not what I want  )

In the end it is up to you. A good way is to try before you buy. If you have the chance to try the oscilloscopes you are interested in, then go that route. Make some kind of test plan & prepare some circuits you want to do measurements on (analog signals, protocol decoding, some math, etc) otherwise you'll be just twiddling some knobs. Maybe this thread https://www.eevblog.com/forum/testgear/functional-comparison-of-rs-rtb2000-siglent-sds2000x-and-keysight-dsox1000/ made by  forum member RBBVNL9 can serve as input for what kind of tests to do.

Short of that you can try to find videos with people actually using the oscilloscope to make measurements. Not so much reviews, but actual use. From there you can try to imagine how a particular oscilloscope would work for you. But still, some things are very much depending on what kind of circuits you are working on an what kind of workflow you like.

edit: typo

[2N3055]

First image some signal. It looks pretty much very noisy. MSO5000 user will say: "you see, I see same noise.."


Then you zoom in and realize all of that is a signal:


There are no noise reduction techniques that preserve full bandwidth AND decrease noise on a signal that is not strictly repetitive. You either give up bandwidth (i.e you get a 1Mhz scope) or noise is there.

A side note: this is with a scope on 1mv/div natively (10mv with 10x probe) riding on top of 11,5V signal. DC coupled...
Ruminate on that. That is what a low noise scope front end can do...
With a 500MHz+ full bandwidth... 

Again, forgive my ignorance.. I do understand the advantages of having a low noise scope.

But, could you give an example situation where I might come across such a signal ? Maybe then I'll be able to make a judgement - whether or not such a use-cases would ever be one that I'll need to consider. Also, are you implying that such a measurement would have been impossible on the Rigol ?

That is a ripple on a power supply.
I don't have a MSO5000 here to try. Someone who has it can try to recreate scenario.

[2N3055]

In the end it is up to you. A good way is to try before you buy. If you have the chance to try the oscilloscopes you are interested in, then go that route. Make some kind of test plan & prepare some circuits you want to do measurements on (analog signals, protocol decoding, some math, etc) otherwise you'll be use twiddling some knobs. Maybe this thread https://www.eevblog.com/forum/testgear/functional-comparison-of-rs-rtb2000-siglent-sds2000x-and-keysight-dsox1000/ made by  forum member RBBVNL9 can serve as input for what kind of tests to do.

Short of that you can try to find videos with people actually using the oscilloscope to make measurements. Not so much reviews, but actual use. From there you can try to imagine how a particular oscilloscope would work for you. But still, some things are very much depending on what kind of circuits you are working on an what kind of workflow you like.

This is good advice!

[bdunham7]

It absolutely belongs in a discussion where people are asking questions about UI, workflow, etc. The only people wishing it wasn't here are the Siglent fanboys.

What it boils down to is a slightly different way of doing things that happens to annoy certain people.  I see the point, but have no real issues adapting to using zoom if I want to scroll horizontally.  Really the only thing that Siglent gets unforgivably wrong here is not allowing you to condense the whole-record display into a bar indicator (showing only what part of the record is being displayed in the zoom window) at the top like you see on Tek scopes.  This would put the matter to bed entirely as far as I'm concerned.  Instead we get jokes about curtains. 

[Fungus]

So the question is this : "Is the Rigol's thick traces (assuming that I'd be working on low level voltages only rarely) bad enough to justify me spending around 570 USD more ?

From what you've stated as your use case? I vote no. "Thick traces" don't have any affect at all on the type of measurements you'll be making (ie. signal timing, signal rise times, signal shape+amplitude, timing between multiple channels, serial decoding...etc).

What you need is bandwidth, sample rate, lots of memory. The Rigol wins out there.

There's also workaround for measuring low level voltages.

Really though: Nobody can possibly answer that question for you. It's your money, it's your decision. The information has been posted.

[2N3055]

A paralel topic:

https://www.eevblog.com/forum/testgear/new-electricalcomputer-engineer-student-test-bench-equipment-recommendations/msg4093312/#msg4093312

From what you've stated as your use case? I vote no. "Thick traces" don't have any affect at all on the type of measurements you'll be making (ie. signal timing, signal rise times, signal shape+amplitude, timing between multiple channels, serial decoding...etc).

What you need is bandwidth, sample rate, lots of memory. The Rigol wins out there.

There's also workaround for measuring low level voltages.

Really though: Nobody can possibly answer that question for you. It's your money, it's your decision. The information has been posted.

Trace noise DO have (smaller or larger) effect on signal rise times, signal shape+amplitude. Some measurements will average out data in statistics.. Some won't.  Decoding won't care, or anything that you do on a scope that is just checking if signal is there.

SDS2000X+ has MORE bandwidth (both in "non hacked" basic version and in full blown "unlocked" version). It has sufficent sample rate (but MSO5000 does have faster sample rate). Siglent has lots more of memory (4x100 or 2x200 MPoints). Even more in segmented mode.  Actuall, on your own merits Siglent wins here.

And again there is no workaround for measuring low level voltages that is equivalent to proper front end sensitivity.
There are some specialized preamps that could be used to measure in some cases but they are not workarounds that can replace native scope sensitivity.
It is a different question if OP will ever need to measure down to microvolts. OP might not need that and in that case it is not important.

I agree with your last sentence completely.

[nctnico]

And again there is no workaround for measuring low level voltages that is equivalent to proper front end sensitivity.

I think you have to rephrase this: there is no workaround for getting rid of noise at all signal levels. After all, a relatively large amount of noise is affecting all signal levels.

Having owned a rather noisy DSO myself where the noise got in the way of using it, I'm wary of dismissing the downsides of a noisy DSO so easely. But this argument is running in circles now. The fact is that at this price level there will be compromises.

[kcbrown]

But, could you give an example situation where I might come across such a signal ? Maybe then I'll be able to make a judgement - whether or not such a use-cases would ever be one that I'll need to consider. Also, are you implying that such a measurement would have been impossible on the Rigol ?

As mentioned, power supply ripple.  Whether it'll be visible on the Rigol or not depends on the intensity of the ripple.  I'm sure there are many other things that might qualify, but suspect that most of them will be of interest only to people who are doing analog work.

If all you're going to do is digital work, and you know this, then a scope with a relatively noisy frontend isn't going to give you problems, and other things are going to matter a lot more.

Honestly, as a general purpose tool, something like the Rigol is probably going to be just fine.  The problem is that you won't necessarily know in advance when you'll need to see signals from your DUT that fall within the noise range of the frontend.  Sometimes you're trying to figure out why your circuit is failing and it's only after you see the full characteristics of the signal that you realize what's really going on.  If the signal includes swings that fall below the noise floor, you won't see them.  Maybe that will matter and maybe it won't, but it may be that you won't realize it until you see it.  As a general rule, the quietness of your frontend is like horsepower in a car.  Chances are you're not going to need a lot of it but there are some situations in which there just isn't a substitute.

One question I have to ask is: in your region, what are the prices of:

• The Rigol MSO5000 (probably the 5074)
• The Siglent SDS-2104X+
• The Siglent SDS-1104X-E
• The Siglent SDS-1104X-U
• The Instek GDS-1054B

Prices tend to vary a lot across regions, and I think it would be interesting to see what the various options would actually cost you.

[kcbrown]

In the end neither are 'perfect' allround. AFAIK both Rigol and Siglent only decode what is on screen.

Yep.

I don't know how good zoom mode is on the Rigol.  On the Siglent, you can use that to simultaneously get full memory decoding and display of decoding of any particular section of the signal.

Note that the Siglent's acquisition method is to acquire only to the time boundaries of the screen anyway, so the main situation you'll run into limitations with respect to on-screen decoding are when you zoom into a section of the acquired signal without using zoom mode.  As with any tool, you'll be fine once you learn how to use it properly, and in this case, using zoom mode is the proper way to deal with the situation.

If you want to get something more polished then look at the R&S RTB2004 for example. Maybe it makes more sense to buy something cheap right now, see where the roads leads you to and upgrade later.

I'd generally agree with this, though that will depend on why the OP is looking for what amounts to a long-term purchase and what his minimum requirements are, as well as the likely amount of time before he outgrows his initial purchase. 

Depending on price and availability in his region, and depending on his needs, the OP should probably take a hard look at the Instek GDS-1054B.

[nubinstanley]

Anyhow, first of all, let me apologize for triggering another "noise war". 

I am happy to have received varying perspectives and now I think that the ball is in my court. Thank you all, for all the inputs/advice. Maybe I'll drop a message here after I finally buy a scope (only if I buy something that I actually need and not what I want  )

In the end it is up to you. A good way is to try before you buy. If you have the chance to try the oscilloscopes you are interested in, then go that route. Make some kind of test plan & prepare some circuits you want to do measurements on (analog signals, protocol decoding, some math, etc) otherwise you'll be just twiddling some knobs. Maybe this thread https://www.eevblog.com/forum/testgear/functional-comparison-of-rs-rtb2000-siglent-sds2000x-and-keysight-dsox1000/ made by  forum member RBBVNL9 can serve as input for what kind of tests to do.

Short of that you can try to find videos with people actually using the oscilloscope to make measurements. Not so much reviews, but actual use. From there you can try to imagine how a particular oscilloscope would work for you. But still, some things are very much depending on what kind of circuits you are working on an what kind of workflow you like.

edit: typo

 

I am trying to get a demo of the MSO5000. Will go through the thread.

[nubinstanley]

That is a ripple on a power supply.
I don't have a MSO5000 here to try. Someone who has it can try to recreate scenario.

 

[nubinstanley]

So the question is this : "Is the Rigol's thick traces (assuming that I'd be working on low level voltages only rarely) bad enough to justify me spending around 570 USD more ?

From what you've stated as your use case? I vote no. "Thick traces" don't have any affect at all on the type of measurements you'll be making (ie. signal timing, signal rise times, signal shape+amplitude, timing between multiple channels, serial decoding...etc).

What you need is bandwidth, sample rate, lots of memory. The Rigol wins out there.

There's also workaround for measuring low level voltages.

Really though: Nobody can possibly answer that question for you. It's your money, it's your decision. The information has been posted.

Yes! Decision time.

[nubinstanley]

But, could you give an example situation where I might come across such a signal ? Maybe then I'll be able to make a judgement - whether or not such a use-cases would ever be one that I'll need to consider. Also, are you implying that such a measurement would have been impossible on the Rigol ?

As mentioned, power supply ripple.  Whether it'll be visible on the Rigol or not depends on the intensity of the ripple.  I'm sure there are many other things that might qualify, but suspect that most of them will be of interest only to people who are doing analog work.

If all you're going to do is digital work, and you know this, then a scope with a relatively noisy frontend isn't going to give you problems, and other things are going to matter a lot more.

Honestly, as a general purpose tool, something like the Rigol is probably going to be just fine.  The problem is that you won't necessarily know in advance when you'll need to see signals from your DUT that fall within the noise range of the frontend.  Sometimes you're trying to figure out why your circuit is failing and it's only after you see the full characteristics of the signal that you realize what's really going on.  If the signal includes swings that fall below the noise floor, you won't see them.  Maybe that will matter and maybe it won't, but it may be that you won't realize it until you see it.  As a general rule, the quietness of your frontend is like horsepower in a car.  Chances are you're not going to need a lot of it but there are some situations in which there just isn't a substitute.

One question I have to ask is: in your region, what are the prices of:

• The Rigol MSO5000 (probably the 5074)
• The Siglent SDS-2104X+
• The Siglent SDS-1104X-E
• The Siglent SDS-1104X-U
• The Instek GDS-1054B

Prices tend to vary a lot across regions, and I think it would be interesting to see what the various options would actually cost you.

Prices converted to USD below :

• The Rigol MSO5000 (probably the 5074) : 1280 USD
• The Siglent SDS-2104X+ : 1862 USD
• The Siglent SDS-1104X-E : 670 USD
• The Siglent SDS-1104X-U : 525 USD
• The Instek GDS-1054B : 592 USD (from Amazon -> not confirmed from an authorized reseller)

At the lower end, I did some research and finalized on the SDS1104X-E.  I could justify spending double that amount (to my wife and myself  ) and so I continued with my research. I landed on the MSO5000, which seemed like the perfect one for me and then I came across the front-end noise discussion. This in turn lead me to the SDS2104X Plus and the question I had was -> Should I spend the extra ~580 USD for it.

Anyhow, I think I have received all the information I can possible understand. Need to take a decision....

[nubinstanley]

A paralel topic:

https://www.eevblog.com/forum/testgear/new-electricalcomputer-engineer-student-test-bench-equipment-recommendations/msg4093312/#msg4093312

From what you've stated as your use case? I vote no. "Thick traces" don't have any affect at all on the type of measurements you'll be making (ie. signal timing, signal rise times, signal shape+amplitude, timing between multiple channels, serial decoding...etc).

What you need is bandwidth, sample rate, lots of memory. The Rigol wins out there.

There's also workaround for measuring low level voltages.

Really though: Nobody can possibly answer that question for you. It's your money, it's your decision. The information has been posted.

Trace noise DO have (smaller or larger) effect on signal rise times, signal shape+amplitude. Some measurements will average out data in statistics.. Some won't.  Decoding won't care, or anything that you do on a scope that is just checking if signal is there.

SDS2000X+ has MORE bandwidth (both in "non hacked" basic version and in full blown "unlocked" version). It has sufficent sample rate (but MSO5000 does have faster sample rate). Siglent has lots more of memory (4x100 or 2x200 MPoints). Even more in segmented mode.  Actuall, on your own merits Siglent wins here.

And again there is no workaround for measuring low level voltages that is equivalent to proper front end sensitivity.
There are some specialized preamps that could be used to measure in some cases but they are not workarounds that can replace native scope sensitivity.
It is a different question if OP will ever need to measure down to microvolts. OP might not need that and in that case it is not important.

I agree with your last sentence completely.

I forgot to ask one question : If I go with the SDS2104X+, in your opinion, what is something from the MSO5000 that I would absolutely miss ?

[2N3055]

I forgot to ask one question : If I go with the SDS2104X+, in your opinion, what is something from the MSO5000 that I would absolutely miss ?

There was enough details said about what is different.
What would you miss I cannot say.
It highly depends on what you use it for.

At my desk I have 7 scopes now. Each one of them is better in something than others and worse at something than others.

[tv84]

I forgot to ask one question : If I go with the SDS2104X+, in your opinion, what is something from the MSO5000 that I would absolutely miss ?

Nothing! Only the 500 bucks in the wallet.

So, if you can afford the SDS2kX+ go with it!

[bdunham7]

I forgot to ask one question : If I go with the SDS2104X+, in your opinion, what is something from the MSO5000 that I would absolutely miss ?

That is the sort of question that results in "paralysis by analysis".  I can't say what you would miss, but one feature I'd like to have that the SDS2104X+ doesn't is the HDMI output to directly connect a monitor.  16 LA channels and 4 analog channels on the screen at once is very crowded, especially when you're stuck losing 25% of your screen in the zoom mode.

[Fungus]

Think about what else you could have on your bench with $500 to spend.

Maybe a nice bench multimeter?

A good power supply and soldering iron?

Either would be a much better way to spend that money IMHO.

[Fungus]

... especially when you're stuck losing 25% of your screen in the zoom mode.

Is working with half a screen the latest Siglent owner's workaround to not being able to zoom out?

It still doesn't address the problem of zooming out on a one-off event that you just managed to capture.