Oscilloscope help

[ultrasmurf]

Hi guys,

I need help in getting a new oscilloscope. I own 2 oscilloscope, 100mhz LEADER LS8106 CRT oscilloscope, and Hantek DSO5102P. I used mostly the Leader on my bench, and Hantek is for traveling to customer place. My Leader is getting a bit too old with a little bit of problem here and there. I have the service manual courtesy of leader. I tore it apart couple of years back to replace some cap and a little bit of calibration. Recently it become laggy, as in when I turn the knob or press some, it will take a few sec to respond, if it's warmed up it might become responsive again. Since I used this to generate money, and generally I just want some new shiny things on my desk I decide to try to upgrade it to newer DSO.

My use case is very simple, I use it mostly to look at a very low level sinewave signal circa 2-4mVpp, at frequency of 2-12MHz. I used the DSO mainly to look at x-y trace of linear encoder and some other normal analog signal. Now, when I tried to use the Hantek DSO to replace my Leader as my main bench oscilloscope, I can't get it to show me a stable signal (4mvpp 4MHz).  I'm not well versed in DSO, but my guess is that I can't get a stable trigger, which is funny since it should have much more triggering capability compared to my cro, and no matter what kind of available triggering that I tried it will still not be able to show a stable signal. My CRO just need a twist of the trigger level and a bit of hold here and there and it will be able to show the signal. Attached a pic to show how the DSO display my signal and how my CRO display my signal. Both are connected to the same point at the same time, also imagine the DSO signal jumping around all over the screen...

One nice sales person came to my place and showed me the TEK MDO32. It still unable to show me a my 4mvpp sine signal, except when i hit the fast acquisition mode, and that is also only through the coloured intensity grading ( with big fat shadow on either side of the signal). I'm not sure I can or want to afford the MDO32. My budget is ard usd1000.

So the main question is, what kind of trigger or function do I need in the DSO to be able to sort of mimic how my CRO behave ? What model DSO do you guys recommend ? Things I tried with the Hantek that does not work : changing the persistence setting, trying different type of triggering and their parameter, averaging (kills my ability to tune the signal), BW limit to 20MHz, AC coupled trigger, HF Reject, noise reject, etc. Only if my signal goes above 20mvpp I can get the DSO to trigger properly, but still waveform is less well defined compared to my CRO.

Thanks for all your input.

[Performa01]

You want a DSO with high sensitivity and low noise together with an excellent trigger system. So:

You do not want a scope with faked high sensitivities that are just software zoom instead of true signal amplification.

My recommendation would be the cheap Siglent SDS1104X-E which gives you:

• 100MHz bandwidth
• True full resolution vertical sensitivity of 500µV/div
• very low noise
• deep memory and deep measurements
• intensity or color grading without special mode and any limitations
• 1Mpts FFT for useful spectrum analysis
• 3-Channel high dynamic Bode Plotter for frequency response analysis in conjunction with any Siglent AWG
• Webserver for remote observation and control including screenshots and waveform download

Attached is an example for a 5MHz <2mVpp sine signal captured without any special setting except for the 20MHz bandwidth limit.

[nctnico]

I'd get a low noise pre-amplifier (preferably with differential inputs) instead of a different oscilloscope. At low signal levels a large part of the signal is likely noise coupled into the input.

There are probably some interesting pointers in this thread:
https://www.eevblog.com/forum/testgear/measuring-ultra-low-noise-powersupply-on-budget/

[Fungus]

Only if my signal goes above 20mvpp I can get the DSO to trigger properly, but still waveform is less well defined compared to my CRO.

Get an op-amp. They cost less than $1.

You can also try setting the probe to 1x mode and enabling the 20MHz filter.

[ultrasmurf]

My workbench is a party zone for noises. I did try the op amp method to try to boost up the signal further for my cro to let it see down to 1mvpp level but the outside noise just swamp it. And beside that, I want a shiny new toy... My hantek will always be a portable oscilloscope, since it is cheap and it's relatively portable. So what I want is really to replace the CRO, not because it is incapable, but more because it's getting problematic. If I have time in the future I will definitely open it up and try to fix it again.

I did some reading on the superb review of the siglent by Performa01, and also some other siglent model. Looks like they do have pretty interesting triggering capability (or I am just too old and never really update my knowledge of DSO). One question to the pros, you see my DSO screenshot, I can somewhat see the 4mvpp signal but it is modulated by a slow moving noise, and also affected by some other high freq noise. How does my CRO able to sort of filter out the noise and show my the signal while the DSO keep showing me all the extra noises ? I did try to limit the DSO bandwidth while the CRO run at its full bandwidth except for the 1x probe. I believe with good triggering I can make the DSO signal stable (i.e. not jumping left right), but it will still be affected by the modulating signal, so it will still jump up and down ?

[Fungus]

How does my CRO able to sort of filter out the noise and show my the signal while the DSO keep showing me all the extra noises ?

Because the DSO is better:

[tautech]

One question to the pros, you see my DSO screenshot, I can somewhat see the 4mvpp signal but it is modulated by a slow moving noise, and also affected by some other high freq noise. How does my CRO able to sort of filter out the noise and show my the signal while the DSO keep showing me all the extra noises ? I did try to limit the DSO bandwidth while the CRO run at its full bandwidth except for the 1x probe. I believe with good triggering I can make the DSO signal stable (i.e. not jumping left right), but it will still be affected by the modulating signal, so it will still jump up and down ?

My guess is mains and SMPS noise.

Last week I had a customer quite concerned he had a problem with his new SDS1104X-E so we went through a few scenarios and checks together and finally narrowed it down to environmental noise as by using the probe spring reference attachment instead of the reference lead the interference to his signal disappeared.
Despite checking all nearby appliances at the time some days later he contacted me to say he had found the source of the interference. 
It turned out to be a SMPS charger on the other side of the room !
Such is the sensitivity of the modern DSO.

[vk6zgo]

Hi guys,

I need help in getting a new oscilloscope. I own 2 oscilloscope, 100mhz LEADER LS8106 CRT oscilloscope, and Hantek DSO5102P. I used mostly the Leader on my bench, and Hantek is for traveling to customer place. My Leader is getting a bit too old with a little bit of problem here and there. I have the service manual courtesy of leader. I tore it apart couple of years back to replace some cap and a little bit of calibration. Recently it become laggy, as in when I turn the knob or press some, it will take a few sec to respond, if it's warmed up it might become responsive again. Since I used this to generate money, and generally I just want some new shiny things on my desk I decide to try to upgrade it to newer DSO.

My use case is very simple, I use it mostly to look at a very low level sinewave signal circa 2-4mVpp, at frequency of 2-12MHz. I used the DSO mainly to look at x-y trace of linear encoder and some other normal analog signal. Now, when I tried to use the Hantek DSO to replace my Leader as my main bench oscilloscope, I can't get it to show me a stable signal (4mvpp 4MHz).  I'm not well versed in DSO, but my guess is that I can't get a stable trigger, which is funny since it should have much more triggering capability compared to my cro, and no matter what kind of available triggering that I tried it will still not be able to show a stable signal. My CRO just need a twist of the trigger level and a bit of hold here and there and it will be able to show the signal. Attached a pic to show how the DSO display my signal and how my CRO display my signal. Both are connected to the same point at the same time, also imagine the DSO signal jumping around all over the screen...

One nice sales person came to my place and showed me the TEK MDO32. It still unable to show me a my 4mvpp sine signal, except when i hit the fast acquisition mode, and that is also only through the coloured intensity grading ( with big fat shadow on either side of the signal). I'm not sure I can or want to afford the MDO32. My budget is ard usd1000.

So the main question is, what kind of trigger or function do I need in the DSO to be able to sort of mimic how my CRO behave ? What model DSO do you guys recommend ? Things I tried with the Hantek that does not work : changing the persistence setting, trying different type of triggering and their parameter, averaging (kills my ability to tune the signal), BW limit to 20MHz, AC coupled trigger, HF Reject, noise reject, etc. Only if my signal goes above 20mvpp I can get the DSO to trigger properly, but still waveform is less well defined compared to my CRO.

Thanks for all your input.

Analog Oscilloscopes can sometimes confuse a higher (desired) frequency "riding" on a lower frequency with a "broadish" bandwidth higher frequency signal "riding" on the desired signal.
Both can appear like the first picture..

In the first case, the zero crossings of successive sweeps at the desired rate are displaced vertically on the screen, so look like "fuzzy noise".
If you increase the time/div  you should be able to see the interfering signal.

In the second case, if you have a delayed time base, you should be able to display the higher frequency noise.

From the Hantek display, the "noise" would appear to mainly be a signal of around 500 kHz.
All provided I have deciphered your display correctly!

[Performa01]

How does my CRO able to sort of filter out the noise and show my the signal while the DSO keep showing me all the extra noises ? I did try to limit the DSO bandwidth while the CRO run at its full bandwidth except for the 1x probe. I believe with good triggering I can make the DSO signal stable (i.e. not jumping left right), but it will still be affected by the modulating signal, so it will still jump up and down ?

Basically, a proper DSO gives you a lot more information and analysis capabilities than any analog scope. Within its bandwidth limits (determined by the effective sample rate), a DSO will never hide anything and advanced triggers allow stable display of even complex signals.

That's an important difference - the analog scope has a constant bandwidth regardless of the timebase, whereas the effective DSO bandwidth starts dropping above a certain time/div setting. What is even worse, a limited effective bandwidth does not mean that frequencies above are suppressed, they will still be displayed as aliased signals with incorrect frequency.

So the effective samplerate is a key parameter and we need long memory if we want to keep it high even at slower timebase settings. That's why deep memory is required especially if an analog scope that has a delayed 2nd timebase (equivalent to the zoom feature on a DSO) is to be replaced by a DSO.

In a modern DSO, you should be able to find some suitable trigger for most complex signals, but in your case it appears to be a mixture of the real signal and some unrelated interference signal components. You would be able to get a stable trigger on each of them, but the remaining signals will still not be stable on the display as long as there is no precise integer relationship between the individual signal frequencies.

[vk6zgo]

I realise that I left my comment a bit "up in the air".
My point was that, what the Hantek was showing may have been correct, with your 3.5MHz signal riding upon a 500kHz signal.
As I said, the analog 'scope, when displaying the 3.5MHz signal would just show a thickening of the trace,
looking like higher frequency noise, because successive sweeps are superimposed upon each other.

The obvious way to see if it agrees with the DSO is to try to display the lower frequency signal.

[ultrasmurf]

Yes I get it that DSO is more sensitive, I know that CRO sometimes hide the signal in a thickening of the trace, which is ok to me, because when I see thick trace I know there is some high freq noise there and I can tweak the CRO to display it if I want to look at it. Attached is the slow freq noise and the high freq noise as seen by the DSO and CRO. They sort of agree with each other. Just that I can tune my CRO to display the 4MHz signal clearer than I can tune my DSO (but it's easier to tune the DSO to look at the noise...  ). To note on the CRO I need to change the voltage/div to higher value to show the high freq noise signal, otherwise it is too dim to take a pic.

BTW this is not about which one is better CRO vs DSO... I need help to get a new DSO to replace my aging CRO, but I need one that can let me show my signal, which I thought is a simple signal to catch... no digital stuff, no difficult triggering requirement (from CRO point of view), low enough frequency that does not really require those GHz bandwidth. I've seen the TEK MDO32 and it can show my signal clear enough for me (even with all the other noises). This was done through fast acq mode together with its color intensity grading (my signal shows a nice red trace among other i think green or yellow thick furry line). It's a demo unit that has sinced been returned, so I cant get a pic showing what it looks like. I cant justify the USD2K+ for it.
 
Do I need to build a bandpass filter to be used together with a new DSO to be able to analyze the signal that I want to see ? Is there any DSO out there within the 1K budget that has this built in ? Seems like something that should be relatively easy to implement in the DSP of an oscilloscope. The siglent seems to have the color intensity gradient function, which I think is similar to the TEK that I saw.. I've read about the rigol MSO5000 series as well, it has lots of bang and whistle, but doesn't seem to have the color gradient stuff. And then there is the keysight EDUX/DSOX11xx series. So far based on the paper spec (no, im not looking for the highest spec, I just read the spec to make sure it is not a 1Mhz stuff...) and video review,  I'm leaning towards the siglent (color intensity gradient). But please please if anyone have been in the same situation as me, trying to analyze small signal level amidst noises, what DSO can you recommend.

[tautech]

In many environments probing technique is of very high importance to get true believable results.
The reference (ground) lead on a probe can be a wonderful wideband EMI antenna hence the incorporation of the short springy wire reference link in your probe pack.
While channel input filters can be of assistance only experience can tell you where interference is coming from and most often it's not from the DUT.

You'd be amazed at the capabilities of the 200 MHz $379 SDS1202X-E.
If looking for something a bit more capable, as in better/more features SDS2202X-E at $619 is also a good choice.
1 GSa/s vs 2 GSa/s.

[tggzzz]

The reference (ground) lead on a probe can be a wonderful wideband EMI antenna hence the incorporation of the short springy wire reference link in your probe pack.

It can act as an antenna, but the principal reason for "short springy wire" ground is to minimise inductance.

A 6 inch lead has ~150nH inductance. When coupled with the probe's tip capactance it forms a resonant circuit at around 100MHz. That resonance can be very visible when looking at modern jellybean logic signals.

[seronday]

Do I need to build a bandpass filter to be used together with a new DSO to be able to analyze the signal that I want to see ? Is there any DSO out there within the 1K budget that has this built in ? Seems like something that should be relatively easy to implement in the DSP of an oscilloscope. The siglent seems to have the color intensity gradient function, which I think is similar to the TEK that I saw.. I've read about the rigol MSO5000 series as well, it has lots of bang and whistle, but doesn't seem to have the color gradient stuff. And then there is the keysight EDUX/DSOX11xx series. So far based on the paper spec (no, im not looking for the highest spec, I just read the spec to make sure it is not a 1Mhz stuff...) and video review,  I'm leaning towards the siglent (color intensity gradient). But please please if anyone have been in the same situation as me, trying to analyze small signal level amidst noises, what DSO can you recommend.

The GW-Instek GDS-2000E series DSO offers independant Low Pass, High Pass, and Band Pass filters on all channels.
The operating frequency of these filters can be configured as required.
This may be what you are looking for.

Regards.

[nctnico]

I second the suggestion for the GW Instek GDS2000E series. I have used the input filtering feature to look at small signals in a circuit which emits lot's of higher frequencies by nature and the input filtering turned out the be invaluable. I'm keeping mine just for this reason. The MicSig TO1000 series also has signal filtering. In general there aren't many oscilloscopes which have this feature though.

[GerryR]

How does my CRO able to sort of filter out the noise and show my the signal while the DSO keep showing me all the extra noises ?

Because the DSO is better:

So, let me understand this: the DSO is so much better that it picks up everything and makes everything unintelligible so you can't troubleshoot your problem, whereas the not-so-good CRO misses information, so you can get the job done.  I'll have to dwell on that for awhile.
I think the issue is you need a very expensive DSO to get the necessary filters, samples, memory depth, etc. so you can make sense of what's going on, or you need two 'scopes like the OP has, a CRO and a DSO.  The thing I mainly use my DSO for is FFT.  A good CRO and a spectrum analyzer takes care of that issue.  I will admit, though, that I would rather haul a DSO to the job site than a CRO.
Just my $.02.

[Fungus]

How does my CRO able to sort of filter out the noise and show my the signal while the DSO keep showing me all the extra noises ?

Because the DSO is better:

So, let me understand this: the DSO is so much better that it picks up everything

No, it displays everything. A subtle difference, but important. Your CRO is picking up the junk but it's simply incapable of displaying it.

and makes everything unintelligible so you can't troubleshoot your problem, whereas the not-so-good CRO misses information, so you can get the job done.  I'll have to dwell on that for awhile.
I think the issue is you need a very expensive DSO to get the necessary filters, samples, memory depth, etc. so you can make sense of what's going on, or you need two 'scopes like the OP has, a CRO and a DSO.  The thing I mainly use my DSO for is FFT.  A good CRO and a spectrum analyzer takes care of that issue.  I will admit, though, that I would rather haul a DSO to the job site than a CRO.
Just my $.02.

Once you have all the information you can start to work on it. Even a low end DSO will have features to eliminate all the external stuff, especially on periodic signals.

Look into waveform averaging (combining multiple captures of the same wave into a single displayed weave), usually labelled "average" in the waveform menu. It works really well for periodic (repeating) signals. Even on my lowly Rigol DS1054Z I can combine up to 1024 captured waves into a single display.

There should also be a "HiRes" mode in there (software low pass filter to eliminate fine noise).

Then there's the 20MHz hardware filter to make it act a bit more like your CRO.

etc.

[ultrasmurf]

Been reading the GW instek, seems like it will fit my need. Also digging through Siglent ( I want that color intensity gradient... it looks pretty) some of their model range also have this filtering, the SDS1000CML, CFL range seems to have this, but it's absent in the 1000XE, 2000XE series. Is this just a accidental drop out of the manual, or really their newer scope has none of the filtering capability ?

Also looks like I really need to stick to a certain product range in GW Instek ? In the manual only the GDS2000E, MDO2000E, and MSO2000E series have this function, while the similar looking 1000B series, and 3000 series does not have this function ? Or is it something that the manual missed out ?

Oh, and those filtering function, are they applied before the triggering logic, or are they applied after the signal is captured ? Either way I think it would be able to show a stable signal, just that if it is before triggering mechanism, then the signal will be even more stable....

[ultrasmurf]

No, it displays everything. A subtle difference, but important. Your CRO is picking up the junk but it's simply incapable of displaying it.

[/quote]

Not CRO vs DSO... butttt... the CRO is capable of displaying the noise if I choose to display the noise. Thick trace means noise somewhere, I know this and I can display it if needed. I guess it really is that the CRO is able to get a better triggering for the type of signal that I get. Doing the same in my DSO adjusting lots of trigger parameter is still not making the signal that I want to see become stable. Just to add, I totally forgot there is hold off adjustment in the DSO (hidden in menu instead of knob...). Adjusting this after looking at the noise somehow gets my signal to a tolerable viewing (still jump up and down but much better than before), it still is not where I can do my adjustment reliably, doable if im forced to...

Yes, the averaging and stuff give me a nice sinewave of my signal in DSO, but in this case as I need to do adjustment, the delay between adjustment and visual feedback makes this not an option for me...

I really need to find demo unit for those GW Instek....

[tggzzz]

No, it displays everything. A subtle difference, but important. Your CRO is picking up the junk but it's simply incapable of displaying it.

Not CRO vs DSO... butttt... the CRO is capable of displaying the noise if I choose to display the noise. Thick trace means noise somewhere, I know this and I can display it if needed.

Of course an analogue scope is capable of displaying the noise!

You can also use it to measure the RMS value of the noise:

I guess it really is that the CRO is able to get a better triggering for the type of signal that I get. Doing the same in my DSO adjusting lots of trigger parameter is still not making the signal that I want to see become stable. Just to add, I totally forgot there is hold off adjustment in the DSO (hidden in menu instead of knob...). Adjusting this after looking at the noise somehow gets my signal to a tolerable viewing (still jump up and down but much better than before), it still is not where I can do my adjustment reliably, doable if im forced to...

Triggering is one of those things that distinguishes a good scope from a less good scope.

The "hidden in the menuing system" is undesirable; analogue scopes are better than digitising scopes in that respect

Yes, the averaging and stuff give me a nice sinewave of my signal in DSO, but in this case as I need to do adjustment, the delay between adjustment and visual feedback makes this not an option for me...

Digitising and analogue scope are different tools with different capabilities, advantages and disadvantages.

Unqualified statements such as "X is better" and "ignore Y and get a Z" are unbalanced and should be disregarded. Unfortunately people, particularly beginners, like to latch onto overly simplistic statements without understanding and thinking.

BTW, DSO vs CRO is misleading. I have digitising scopes that use a CRT.

[tautech]

Been reading the GW instek, seems like it will fit my need. Also digging through Siglent ( I want that color intensity gradient... it looks pretty) some of their model range also have this filtering, the SDS1000CML, CFL range seems to have this, but it's absent in the 1000XE, 2000XE series. Is this just a accidental drop out of the manual, or really their newer scope has none of the filtering capability ?

Well yes the power of a greater selection of triggers sort of negates the need for something that hides signals.
P11 & 12/21
http://siglentna.com/wp-content/uploads/dlm_uploads/2018/02/SDS1000X-E_DataSheet_DS0101E-E03C.pdf

[StillTrying]

( I want that color intensity gradient... it looks pretty) some of their model range also have this filtering, the SDS1000CML, CFL range seems to have this, but it's absent in the 1000XE, 2000XE series. Is this just a accidental drop out of the manual, or really their newer scope has none of the filtering capability ?

The CMLs, CFLs and the "+" versions don't have any color intensity gradient, and just 1 intensity of persistence, they're definitely not today's fancy displays.
I don't think the digital filters are all that useful, limited range which changes when you change the X speed. They seem to work on curvy waveforms but if I low pass or high pass a simple square wave the resulting wave shapes seem well off from what real RC or bandwidth limiting would do.

Oh, and those filtering function, are they applied before the triggering logic, or are they applied after the signal is captured ? stable....

The CML+'s 20MHz bandwidth limit is before the trigger, the digital filters are on the already captured data, you can adjust the digital filter on a STOPed waveform.

[tudou2048]

This DSO does not seem to trigger effectively under small signals. You can try to adjust the trigger level in a wider range. For ease of observation, the "nomal" mode can be temporarily adjusted.

[ultrasmurf]

I narrow my selection down to GW Instek GDS-1054B (digital filter) and Siglent SDS1104X-E (pretty color).

I am not selecting the GW Instek 2000 series, as most likely this is a purchase based on internet data rather than actual demo of the unit, and I'd rather be out of 500USD than 1000USD, and found out I need to somehow repair my CRO just to be able to do my work.

My previous statement about GW Instek model supporting digital filter is limited to the 2000 series is wrong. The newer 1000B series as well as the 3000 series support this as well according to the manual as this is an app to be installed in the scope.

My heart wants the siglent, but my brain says Instek... I can't get any demo unit so I can't really compare their performance for my application.

Any kind soul willing to put a 4mvpp 4MHz sine wave signal amplitude modulated by 10mvpp 50KHz triangle signal, with a random 20Mhz 50mvpp noise to their GW instek 1054B or Siglent 1104X-E ? I really want to see whether the GW instek can do better with its digital filter, or Siglent can trigger reliably showing me the 4MHz signal. Also I need it to be as close to realtime as well, so vary the 4MHz signal and see if the digital filter or the triggering setting actually cause a massive delay.... a delay of about 200-300ms should be tolerable, 1s delay is bad...

[Fungus]

My previous statement about GW Instek model supporting digital filter is limited to the 2000 series is wrong. The newer 1000B series as well as the 3000 series support this as well according to the manual as this is an app to be installed in the scope.

Yep. With hacks and plugins the 1000B can have pretty much all the features of the higher-up models.