Should I buy a Rigol MSO5000?

[Fungus]

For me the aliexpress link results in an error page "Leider ist dieser Artikel an Ihrem Standort derzeit nicht verfügbar. Entdecken Sie weitere Gruppenkaufangebote."
Would you write down a sentence what you mean, like bandwidth, gain, supply options?

I get this:


There's hundreds of sellers with that amplifier though, just search for "40db broadband amplifier"

Specs:

[dietert1]

Thanks. So you were using that one?
It is spec'd for 60 dB, that is G=1000 and bandwidth 30 .. 4000 MHz. In your scope screen you were using it at 8 MHz (out of bandwidth) and at a gain of about 70. Maybe not the best solution for a scope, where we want to preserve the input signal shape.
Also one might want a preamplifier that can be used with the usual scope probes. These amplifiers are 50 Ohm RF systems. Does the Rigol support 50 Ohm input termination?

Regards, Dieter

[Fungus]

Thanks. So you were using that one?
It is spec'd for 60 dB, that is G=1000 and bandwidth 30 .. 4000 MHz. In your scope screen you were using it at 8 MHz (out of bandwidth) and at a gain of about 70. Maybe not the best solution for a scope, where we want to preserve the input signal shape.
Also one might want a preamplifier that can be used with the usual scope probes. These amplifiers are 50 Ohm RF systems. Does the Rigol support 50 Ohm input termination?

Looking at millivolt signals with ordinary probes probably won't work well on any oscilloscope. It's usually done in 50-Ohm terminated systems.

FWIW: The MSO5000 doesn't have internal 50Ohm termination.

The proposed solution would only be for occasional use. If looking at mV signals is a central part of what you're planning to use your oscilloscope for then maybe you need to look at a 12-bit "HD" oscilloscope.

[dietert1]

We have a Rigol DS2202E and it has the 50 Ohm input termination. Anyway that's half the story. The preamplifier input should also be switchable between high impedance and 50 Ohm.
Certainly if a scope doesn't pick a signal because it is to small, a preamplfier will have a similar problem.
Sometimes i found snap-on ferrites on scope probes useful when using digital scopes. Or feed the probe cable through a ferrite ring core multiple times in order to reduce common mode noise into the ground clip impedance.

Regards, Dieter

[tautech]

Whether or not it's a problem depends on how often you need to look at this sort of signal...

nb. Siglents don't go down to microvolts either. If low very level signals are what you do all day long then you're going to need the amplifier no matter what 'scope you own. 

Best you refresh your thoughts on this matter here:
https://www.eevblog.com/forum/testgear/is-a-rigol-mso5000-overkill-for-a-hobbyist/msg4481968/#msg4481968

[Martin72]

Looking at millivolt signals with ordinary probes probably won't work well on any oscilloscope.

So for what purposes you are using this amp you´ve linked ?
Please show pics from the amp and your scope how they are connected together and with the signal source and of course screenshots where you can see it´s behaviour concerning additional noise, phase-shifting and so on.
Thankyou!

[Anding]

Also, I don't want to curb your enthusiasm, but read along the tail of the other thread. If you want to find its limits, it will suffice that you do stuff below (say) 50 mV.

There has been a lot of discussion about the ADC noise of the Rigol MSO5000 compared with the Siglent SDS2000X Plus, which is more expensive (arguably enough of a price difference to notice if you are hobbyist, perhaps not if you will use it as tool of your trade). 

Instead of this comparison, I'm curious how is the Rigol noise compared to all of other those 'scopes costing up to and around $1,000?

A) Is the Rigol 'noisy' for a 'scope costing up to $1,000, or
B) Is the Rigol noise about the same as other 'scopes costing up to $1,000? 

Does the fact that the Rigol is being compared with the more-expensive Siglent mean it is already "punching above it's weight", but too much to expect a "knockout"?

[tautech]

Also, I don't want to curb your enthusiasm, but read along the tail of the other thread. If you want to find its limits, it will suffice that you do stuff below (say) 50 mV.

There has been a lot of discussion about the ADC noise of the Rigol MSO5000 compared with the Siglent SDS2000X Plus, which is more expensive (arguably enough of a price difference to notice if you are hobbyist, perhaps not if you will use it as tool of your trade). 

Instead of this comparison, I'm curious how is the Rigol noise compared to all of other those 'scopes costing up to and around $1,000?

A) Is the Rigol 'noisy' for a 'scope costing up to $1,000, or
B) Is the Rigol noise about the same as other 'scopes costing up to $1,000? 

Does the fact that the Rigol is being compared with the more-expensive Siglent mean it is already "punching above it's weight", but too much to expect a "knockout"?

See the capture Martin took where a $500 DSO was noticeably better than the 5k.
I linked it a couple posts back.

[Anding]

See the capture Martin took where a $500 DSO was noticeably better than the 5k.
I linked it a couple posts back.

I saw Martin had kindly shared images from a Silgent, the Rigol and a Teledyne LeCroy.    Perhaps there are two different Silgents here, the first a $500 DSO and the latter ones the SDS2000X?

[tautech]

See the capture Martin took where a $500 DSO was noticeably better than the 5k.
I linked it a couple posts back.

I saw Martin had kindly shared images from a Silgent, the Rigol and a Teledyne LeCroy.    Perhaps there are two different Silgents here, the first a $500 DSO and the latter ones the SDS2000X Plus ?

Exactly correct except for missing the Plus.  The great little and very popular SDS1104X-E represents itself very well against many much more expensive.

[Fungus]

Sometimes i found snap-on ferrites on scope probes useful when using digital scopes. Or feed the probe cable through a ferrite ring core multiple times in order to reduce common mode noise into the ground clip impedance.

That won't help on the MSO5000 because most of the noise is in the ADC.

[Martin72]

Looking at millivolt signals with ordinary probes probably won't work well on any oscilloscope.

So for what purposes you are using this amp you´ve linked ?
Please show pics from the amp and your scope how they are connected together and with the signal source and of course screenshots where you can see it´s behaviour concerning additional noise, phase-shifting and so on.
Thankyou!

Reminder.....

[Fungus]

Reminder.....

Here's a picture of what your sock puppet says is too complicated/fiddly/generally impossible to do:

All the guy did was insert the cheapest, cheap-ass $3(plus shipping) Aliexpress amplifier into his signal path and the pink trace turned into the yellow trace.



There was no fiddling or messing around, just connect it up and it works. In this case it was inserted in the path of an AWG.

I don't know what more "proof" anybody could ask for.

[Anding]

I wish to offer my point of view as a newbie who has just taken delivery of an MSO5104.

It has taken me a lot of time and scrutiny of the forum posts (plus time spent on the scope itself) to honestly conclude that I have not wasted my money on a flawed instrument.

The noise in the ADC is undisputed.  You can buy a much less expensive scope that has less noise.  If you spend more money you can buy a scope with less noise and still all of the MSO features.  Martin72’s pictures make this clear.

All I want to say is that I found the relentless presentation / implication of the ADC noise as a “fatal flaw” to be unhelpful - because nobody says “the noise is likely a problem if you doing X but will be no issue if you want to do mainly Y and occasionally Z”

I appeal to the experts here to define the relevant X, Y, and Z in the sentence above and thus make the discussion informative to newbies such as myself.

For me,
X: hifi audio work? Not my interest as it happens
Y: microcontroller and FPGA embedded projects
Z: VLF antenna work, I can see the noise getting in the way unless I amplify

[dietert1]

When we use a scope, we want to measure something, like signal shape, amplitude, noise..
Nothing of this will be possible with that cheap amplifier in the signal chain, using it out of bandwidth and with its low input impedance. That recommendation was misleading to say the least.

Regards, Dieter

[Anding]

For making VLF antennas it's basically just a case of adding or removing capacitors until resonance is found at the desired frequency, typically 20-100 kHz.  And you only do it once.  As long as an amplifier is stable then relative readings are enough.  Granted it's a different amplifier module though. 

[2N3055]

Reminder.....

Here's a picture of what your sock puppet says is too complicated/fiddly/generally impossible to do:

All the guy did was insert the cheapest, cheap-ass $3(plus shipping) Aliexpress amplifier into his signal path and the pink trace turned into the yellow trace.

There was no fiddling or messing around, just connect it up and it works. In this case it was inserted in the path of an AWG.

I don't know what more "proof" anybody could ask for.

Dear Fungus,

Now I see you don't understand what we are talking about. I apologize for not being clear enough.
Yes we can see some scope screen and some waveforms.

But that is exactly the point: you see something on a screen.
And if all you need is to see if, for instance, you have clock on some pin it is fine like that.
But in that case even the original signal representation by noisy MSO5000  would be enough for GO/NoGO check.

But scope is not just used for scope art. We need to measure things. We need to see if signal is distorted.
We need to verify amplitude.
For that amplifier needs to be calibrated, it needs to have standardized input and output impedance. We need to have it
characterized for bandwidth, distortion, phase distortions..
When that "amplifier" is part of scope itself all that stuff is already done by scope manufacturer and included in scope specifications and calibrated.
Also since when you have a scope that has minimum 5mV/div, from that to 500uV there are 3 amplification steps missing (2mv, 1mv, and 500uV div). Are you going to have 3 different amplifiers? For what bandwidth? What input imepedance? Those 5 USD amplifiers are RF amps. They are 50Ω (they should be but nobody know if they are and what it's real impedance is, you need a VNA to measure it). They don't go to DC.  You might find an amplifier that is high impedance (1MΩ//some pF) but with what BW? Is that one DC too?.
So, if you have a lab with several thousands worth of other equipment you might characterize and use your own preamps for specific purpose. And that is actually done all the time. People make custom preamps for specific purpose too.
But none of that is "replacement of fix" for scope limited range. It is an application specific thing that takes a lot effort and  knowledge to be done right. None of it is "just throw in a preamp". And you might need dozens of different preamps and all the stuff that goes with it.

Fact is, and that is what I said before but it was not registered, there are oscilloscope preamplifiers. Basically those are scope front ends (the high sensitivity low noise ones, some even differential) with all the needed parameters already sorted out.  Those can be (and are) used to purpose you are advocating. Problem with those is they cost so much, it is cheaper to buy SDS2354X HD (12 bit low noise one) than one of these fancy preamps.

[Fungus]

And if all you need is to see if, for instance, you have clock on some pin it is fine like that.

Are you seriously saying I'm telling people to use an amplifier to see if there's a clock on a pin?

All I want to say is that I found the relentless presentation / implication of the ADC noise as a “fatal flaw” to be unhelpful

Yep. The Siglent fanboys are a constant nuisance.

nobody says “the noise is likely a problem if you doing X but will be no issue if you want to do mainly Y and occasionally Z”

I say things like that all the time.

I say things like "Digital use: Not a problem." and "Looking at mV signals? Problem."

When we use a scope, we want to measure something, like signal shape, amplitude, noise..
Nothing of this will be possible with that cheap amplifier in the signal chain, using it out of bandwidth and with its low input impedance. That recommendation was misleading to say the least.

You're all batshit insane. Seriously.

Every single time I mention an amplifier I make a point to qualify it with "for occasional use" or "in a pinch".

I even tell people not to use them, eg. here's a post on the previous page of this thread where where I say: "The proposed solution would only be for occasional use. If looking at mV signals is a central part of what you're planning to use your oscilloscope for then maybe you need to look at a 12-bit "HD" oscilloscope."

Link: https://www.eevblog.com/forum/testgear/should-i-buy-a-rigol-mso5000/msg4481770/#msg4481770

[gf]

I wonder whether the mV ranges really need to be used at full bandwidth, or whether a lower bandwidth (say 20MHz or even less) would suffice for many low-voltage use cases?

Of course, if the ADC noise is the dominant contributor to noise, then it won't suffice to limit the bandwidth in the frontend, but the bandwith reduction from the 4GHz noise bandwidth of the ADC to (say) 20MHz needs to be done by applying a low-pass filter in the digital domain, after sampling at full 8GSa/s. The seems to exist a lowpass filter math function.

[Fungus]

I wonder whether the mV ranges really need to be used at full bandwidth, or whether a lower bandwidth (say 20MHz or even less) would suffice for many low-voltage use cases?

The depends on whether you're using a probe or 50 Ohm coax.

If you're using a probe then it will be switched to 1x mode and it's almost compulsory to use the 20MHz bandwidth limiter.

20MHz needs to be done by applying a low-pass filter in the digital domain, after sampling at full 8GSa/s. The seems to exist a lowpass filter math function.

Aka "hires" mode.

[gf]

If you're using a probe then it will be switched to 1x mode and it's almost compulsory to use the 20MHz bandwidth limiter.

In fact, I hadn't even 1x probes in my mind at that moment. But you are of course right, In this case, bandwidth must be renounced anyway.

Aka "hires" mode.

How many adjacent samples does the MSO5000 average in hires mode?
The handbook obviously does not tell. I only find

    In "High Res"mode, the signal bandwidth does not exceed 1/32 of the sampling rate.

which would imply <= 250MHz bandwidth @8GSa/s (or correspondingly >= 14 moving average filter taps).

EDIT: I just saw in a different thread your statement

You have to turn it on and set the number of oversamples (8x, 16x, 32x, etc).

What is the maximum value that can be set?

[Martin72]

Reminder.....

Here's a picture of what your sock puppet says is too complicated/fiddly/generally impossible to do:

I do not got a sock puppet and this is not what I´m asking you for, I already know this pic.
So I´m still waiting.

[tv84]

Taken from Rigol site.

[2N3055]

Taken from Rigol site.

LOL they are comparing to old SDS2000X that is not made anymore....

[nctnico]

We have a Rigol DS2202E and it has the 50 Ohm input termination. Anyway that's half the story. The preamplifier input should also be switchable between high impedance and 50 Ohm.
Certainly if a scope doesn't pick a signal because it is to small, a preamplfier will have a similar problem.

Well, there isn't a single, universal pre-amplifier suitable for all situations so you can't go from generic requirements. If your use cases vary widely, you'll need a wide variety of pre-amplifiers. An oscilloscope input is pretty allround already but don't expect something similar from a pre-amplifier that is tailored to a specific use case (like amplifying RF signals, picking up low-level signals for audio / sensors, differential input, etc).