Best 350MHz scope in a hackable world (Siglent SDS2104X Plus or Rigol MSO5072)

[Fungus]

After all that rambling here is my point. If you want to decide between the Rigol and the Siglent then do your research and figure out which one most fits your wants and needs.

Yep, the thread has actually been over for two pages now.

Conclusion: "It depends".

[tautech]

Every time I read a post and say to myself WTF is this nonsense I look at the author and it's usually Fungus. Is there anyway to filter his posts out of my view?

Profile>Account settings>Modify Profile>Buddies/Ignore list>Edit Ignore list

FYI he has been known to openly stalk and incite members for reactions for them to be banned.
Watch your arse with this fruitcake.

[Fungus]

Every time I read a post and say to myself WTF is this nonsense I look at the author and it's usually Fungus. Is there anyway to filter his posts out of my view?

Profile>Account settings>Modify Profile>Buddies/Ignore list>Edit Ignore list

But don't worry, tautech will quote them all so you can read then anyway.

FYI he has been known to openly stalk and incite members for reactions for them to be banned.

That's a lie and you know it.

[nctnico]

Anyway, back to the topic at hand. For the longest time I have been considering the Siglent 1104x-e because it fits every one of my needs, but never pulled the trigger on it because it lacked touchscreen which I don't need but really want. So I waited for something to come along in price range because if I'm going to go over my budget it needs to last me a very long time. When the Rigol MSO 5000 hit the market I did love the way it looked. But it had a very dim screen and no 50 ohm termination. (Supposedly the dim screen was fixed in a firmware update but I see conflicting information on how good this was). Then the Siglent 2000x plus came out and I really do like the scope but people on here bashing it had me a little gun shy. After I did some research some things aren't ideal such as how it handles memory. It's out of my budget but I'm saving up for it. Probably as close to my dream scope as I will get.

Now you make it sound as if there are only two choices. I agree with your assesment that the Siglent is the better choice compared to Rigol judging on how quickly Siglent fixes bugs (nowadays) and Siglent has a better analog front-end. But then again: do you really need 350MHz of bandwidth? I totally get that it is easy to get hung-up on nice features but if you are on a budget (who isn't? *) then it is a good idea to take a break and put the pros and cons of various models (not just the Rigol and Siglent mentioned in this thread) on a piece of paper. With all the pros and cons on a paper it is easy to identify the more important and less important features versus the money they cost.

* Sometimes it takes me a year or more to decide which instrument to buy.

[Fungus]

Now you make it sound as if there are only two choices. I agree with your assesment that the Siglent is the better choice compared to Rigol judging on how quickly Siglent fixes bugs (nowadays) and Siglent has a better analog front-end. But then again: do you really need 350MHz of bandwidth?

A very valid point.

The elephant in the room is that when you get to 350Mhz you're well inside the area where you need to start looking at active probes.

I totally get that it is easy to get hung-up on nice features but if you are on a budget (who isn't? *) then ...

Any idea of being "on a budget" is moot once you reach active probe territory.

[tautech]

FYI he has been known to openly stalk and incite members for reactions for them to be banned.

That's a lie and you know it.

I know the truth and so do the moderators not your version of 

Your actions were recognized and discussed by a few members so pull your head in.

[nctnico]

Now you make it sound as if there are only two choices. I agree with your assesment that the Siglent is the better choice compared to Rigol judging on how quickly Siglent fixes bugs (nowadays) and Siglent has a better analog front-end. But then again: do you really need 350MHz of bandwidth?

A very valid point.
The elephant in the room is that when you get to 350Mhz you're well inside the area where you need to start looking at active probes.

Not really. I usually use a direct cable connection as high frequency signals often have a low impedance anyway. And there are plenty of options to buy or make cheap passive probes. On top of that you can choose from several active probe designs out there which are very affordable.

@tautech: pot - kettle - black

[Fungus]

Not really. I usually use a direct cable connection as high frequency signals often have a low impedance anyway.

Yes, direct cable is fine.

[mawyatt]

On same 2mv/div scale factor the Siglent has a 0.192mvpp measured noise vs 1.18mvpp reported (don't have the Rigol to measure), so a ~6.1X advantage, and on the 500uv/div a measured 0.161mvpp or 7.3X advantage.

I have both (Rigol MSO5074@350Mhz and Siglent SDS2104X@500Mhz), and measuring the avg stddev of the noise at 1mV I get 75uV on the Siglent and 167uV on the Rigol, both on full bandwidth and default values for everything else. At 2mV and limiting both to 200Mhz bandwidth I get 283uV on the Rigol vs 51uV on the Siglent.

I think that the Siglent is better in terms of noise and UI, and I prefer it for everything except digital signals, but the Rigol is not as bad as some people are telling here. I'm pretty sure that many would have not even dream of something like it 10 years ago.

IMHO the big difference is that Siglent seems much better in terms of firmware updates.

Since you own both DSOs most of the discussions are centering around, could you make the input referred noise measurement with the bandwidth of 20MHz? This was the BW used for the various noise measurements made by Howard Long and used in the power supply measurements by Martin72 as shown in the scope plots, and what I posted that you've quoted. Please include the peak to peak and standard deviation measurements.

Thanks for helping out, most of us aren't lucky enough to own both of these nice DSOs. So much of the discussions centered around speculation (and the usual non-sense) with little supporting measurements

Best,

[Fungus]

Since you own both DSOs most of the discussions are centering around, could you make the input referred noise measurement with the bandwidth of 20MHz? This was the BW used for the various noise measurements made by Howard Long and used in the power supply measurements by Martin72 as shown in the scope plots, and what I posted that you've quoted. Please include the peak to peak and standard deviation measurements.

Thanks for helping out, most of us aren't lucky enough to own both of these nice DSOs. So much of the discussions centered around speculation (and the usual non-sense) with little supporting measurements

 

Would also be nice to know what effect "hi res" mode has on the noise levels at 2x, 4x, 8x.

PS:

[nctnico]

There is another oopsie in the video as well... the oscilloscope inputs aren't terminated and this does make a difference when doing noise comparisons.

[Fungus]

There is another oopsie in the video as well... the oscilloscope inputs aren't terminated and this does make a difference when doing noise comparisons.

Dave says not.

https://youtu.be/Y6gzYbuMjOA?t=1265

[mawyatt]

I see little (<2uv) difference on the 500uv/div scale between having the input terminated with internal 50 ohms, or not at 20MHz BW limit on the SDS2102X Plus I have. With 200MHz & 500MHz it's <1uv. Normally I would expect this to be higher at all BWs, but it is what it is.

BTW does anyone have a detailed schematics of the front end utilized on the latest DSOs? Would be interesting to study what's going on before the ADCs!

Best 

[nctnico]

There is another oopsie in the video as well... the oscilloscope inputs aren't terminated and this does make a difference when doing noise comparisons.

Dave says not.

https://youtu.be/Y6gzYbuMjOA?t=1265

And he is wrong because A) on some scopes it does matter and B) you want to make measurements under the same conditions.

An open input can pick up all kinds of noise and it is not a well defined condition.

[Fungus]

And he is wrong because A) on some scopes it does matter

To be fair, he doesn't actually say "not".

He says that your front end is badly designed if it does.

[nctnico]

And he is wrong because A) on some scopes it does matter

To be fair, he doesn't actually say "not".

He says that your front end is badly designed if it does.

Sorry but a 1 mega Ohm open circuit point is bound to pick up something. An oscilloscope is never used with an open input; the source impedance will always be magnitudes lower so in order to make an apples for apples comparison AND to measure using the same circumstances the oscilloscope input must be terminated when measuring the front-end noise. It is a simple matter of practising good measuring techniques. Which in turn lead to unambiguous results.

[Fungus]

It is a simple matter of practising good measuring techniques. Which in turn lead to unambiguous results.

OK, if you insist...   

[mawyatt]

The whole DSO "noise" concept gets very confusing very quickly when you take time to think about what's actually going on and where it's coming from. Here's a few examples.

All Input Referred Sources:

1) Thermal Voltage noise
2) Current Shot noise
3) 1/f and various flavors noise
4) Power Supply noise
5) AC Mains noise
6) Clock noise
7) ADC Sampling noise
8 ADC Bit noise (reference noise)
9) Quantization noise
10) BNC pickup noise

Where do these various noise sources originate, and where/how are they coupled into the scope electronics?  Then we are asking the DSO to measure it's own Input Referred noise, and at different scope bandwidths and scale factors!!

Some means of a standard test setup needs to be applied, including scale factor, bandwidth, input termination, temperature and so on. I would think that 3 different input terminations be considered, Open (AKA 1M), Short and 50 Ohm (like the Network Analyzers do), which may give a hint as to what types of noise are contributing.

As the linear power supply noise measurement has shown, scope input noise does matter, regardless of how you chose to define it, or foolishly ignore it.
Here's a typical example, when doing troubleshooting of a couple old Tektronix scopes (2465) the manual suggest the first thing in the troubleshooting procedure is to measure the various power supply voltages and noise levels, some from SMPS and some from linear supplies. Tek even provides all these voltages on a common 16 pin DIP socket on the main PCB Top Side. Improper voltage levels or/and excessive noise levels can point you in the proper direction right from the start.

Anyway, having a nice selection of good, low noise, wide band, high resolution DSOs is certainly good for the end users. Lots of quality instruments to select from

Best, 

[nctnico]

The problem with an open input measurement is that you have to define and control the environment. Without anything connected you basically need a shielded chamber (like they use for EMC testing) and a filtered mains supply in order to do this measurement correctly (IOW: with repeatable results for every device). A good (and way cheaper) altenative would be to put a metal cap over the BNC input (like a non-shorting BNC dust cap).

[mawyatt]

The problem with an open input measurement is that you have to define and control the environment. Without anything connected you basically need a shielded chamber (like they use for EMC testing) and a filtered mains supply in order to do this measurement correctly (IOW: with repeatable results for every device). A good (and way cheaper) altenative would be to put a metal cap over the BNC input (like a non-shorting BNC dust cap).

Just get yourself a type N or SMA Open that used for VNA calibration, these have a metal sealed end caps that terminates the electric field, so no need for a expensive chamber or other complex and expensive fixtures. Get a BNC to type N or SMA adapter as shown and you have a shielded Open, completely controlled environment to use.

BTW it makes no difference on the DSO I have whether I use the shield BNC or not. Also makes no difference if I use a BNC Open, BNC Short, BNC Termination, or the DSO 50 ohm termination, they are all ~ the same.

No need for anything expensive, complex or elaborate! Just get one of those inexpensive NanoVNA Cal kits like I show, either SMA or type N, you have a Open, Short and 50 ohm Termination all shielded to use. Simple as that

Best,

[Fungus]

BTW it makes no difference on the DSO I have whether I use the shield BNC or not. Also makes no difference if I use a BNC Open, BNC Short, BNC Termination, or the DSO 50 ohm termination, they are all ~ the same.

FWIW I tried a 50 Ohm terminator on my little Micsig and it made no noticeable difference,

[ResistorRob]

Now you make it sound as if there are only two choices. I agree with your assesment that the Siglent is the better choice compared to Rigol judging on how quickly Siglent fixes bugs (nowadays) and Siglent has a better analog front-end. But then again: do you really need 350MHz of bandwidth? I totally get that it is easy to get hung-up on nice features but if you are on a budget (who isn't? *) then it is a good idea to take a break and put the pros and cons of various models (not just the Rigol and Siglent mentioned in this thread) on a piece of paper. With all the pros and cons on a paper it is easy to identify the more important and less important features versus the money they cost.

* Sometimes it takes me a year or more to decide which instrument to buy.

Well the original post wanted a comparison between the Rigol and Siglent models only. So that is why it sounds that way :-) I personally don't need 350Mhz but in order to get the features I want then I need to get into that range. I have done what you suggested with putting the pros and cons in a list and had many more options than these 2 scopes, so I do agree with you. I also took longer than a year to decide on what to buy. I'm looking for something to use until the scope dies or I die so I'm going out of what would be a reasonable and responsible budget and just getting the scope that will make me smile every time I look at it.

[mawyatt]

Well the original post wanted a comparison between the Rigol and Siglent models only. So that is why it sounds that way :-) I personally don't need 350Mhz but in order to get the features I want then I need to get into that range. I have done what you suggested with putting the pros and cons in a list and had many more options than these 2 scopes, so I do agree with you. I also took longer than a year to decide on what to buy. I'm looking for something to use until the scope dies or I die so I'm going out of what would be a reasonable and responsible budget and just getting the scope that will make me smile every time I look at it.

Followed a similar path but didn't take a year to decide. Sifting thru all the threads on EEVblog, with lots of good information and an equal dose of BS and non-sense, the decision was distilled down to the Rigol and Siglent based upon what I needed/expected in a DSO and what my budget allowed. I decided on the Siglent, but would have been happy with the Rigol as both are excellent values and mid-range DSOs IMO. Sure there are other excellent choices for mid-range DSOs, and being accustomed to Keysight & Tektronix I looked closely at what they offered, but ended up passing on them mainly because of the higher cost and this was out-of-pocket rather than a company purchase, so the "bias" shifted towards the lower cost.

Best,

[mawyatt]

The problem with an open input measurement is that you have to define and control the environment. Without anything connected you basically need a shielded chamber (like they use for EMC testing) and a filtered mains supply in order to do this measurement correctly (IOW: with repeatable results for every device). A good (and way cheaper) altenative would be to put a metal cap over the BNC input (like a non-shorting BNC dust cap).

I wouldn't think that the DSO would normally be exposed to field strengths high enough to significantly couple to the open DSO BNC input. Careful DSO placement without any nearby LED lamps or SMPS converters should keep these unwanted fields at acceptable levels. If one is seeing a significant difference in Open Circuit and Terminated DSO noise levels, this begs the question of the source/reason of this difference? Maybe it's pointing to a coupling mechanism inside the DSO rather than thru the input BNC, or the input amplifier shot noise.

Just speculating, ignoring the input amplifier shot noise (likely FET inputs), but if the internal "noises" within the DSO are the real culprit, and they are coupling to the front end. With the external termination in place, the impedance "seen" by the internal noise sources is significantly lowered and thus the coupling is reduced which shows as a lower self measured noise.

Edits above:

Anyway, something to think about?

Comments??

Best,

[nctnico]

After some thinking IMO a test with a metal dust cover over the BNC inputs is the best way to test for noise. This results in a controlled test setup.