REVIEW - Owon SDS7102 - A look at the SDS series from Owon

[marmad]

I know that FFT suffer issue with limited by memory and wish to check if anyone see the difference and see if FFT take advantage of full memory or just partial (ie screen).

Neither the Owon or the Rigol or the Hantek scopes are very good at FFT. I don't know what your BW and budget requirements are, but if a good FFT in a DSO is something that's very important to you, you might want to look at the Hameg scopes: HMO722 / HMO1022 - I believe they have the best FFT on the market in the lower end scopes. They can calculate FFTs using from 2048 up to 65536 points.

The Hameg scopes are a bit more expensive than some of the others, but they are a quality product - and you get a much better FFT, a lot more triggers, and other features which make the price difference a bit easier to take.

[aghp]

Yes, Hameg and specially Rohde@Schwarz oscilloscopes are high quality products.
HMO 1022 is 2 channel 100MHz 2GSa / 1GSa/s, up to (real) 2000wfm/s and just if FFT is very important it have max 64ksample (32k resolution)  one of most "high end" oscilloscope in this class.
It have very professional features.

Its price looks: here http://webstore.rohde-schwarz.com/mea/hamegr-hmo1022.html  1.3kEur + VAT?  What means littlebit under 1.6kEur dependent EU country VAT%. (yes, exmple in Finland I can not buy from R&S Germany web store, I need buy it from R&S Finland web store, it is R&S policy.

Owon is around under 0.5kEur dependent of also seller. Some sellers are dropshippers and some sellers are professionals who also  know really what they sell.  You get what you pay.
keep care you buy version what is originally from owon accepted and sold for overseas sale. Only these have overseas warranty.  Some cheap sellers do not never give any help to you if you meet some problems. There are lot of good trusted sellers available around of world. Just select them. Not chinese car-garage dropshipper seller who know only that "it have nice box and it is new".
after you buy he call his friend who caall his friend that some box need ship and he get maybe 5 yuan for eat.

[marmad]

Its price looks: here http://webstore.rohde-schwarz.com/mea/hamegr-hmo1022.html  1.3kEur + VAT?  What means littlebit under 1.6kEur dependent EU country VAT%. (yes, exmple in Finland I can not buy from R&S Germany web store, I need buy it from R&S Finland web store, it is R&S policy.

Owon is around under 0.5kEur dependent of also seller. Some sellers are dropshippers and some sellers are professionals who also  know really what they sell.  You get what you pay.

Yes, Kari - as I already mentioned in my message, the Hameg costs more than the Owon. But what you are willing to pay depends on what features are highly important to you - as well as how you plan to use the DSO.

Personally, I find it odd that anyone would buy an Owon or Hantek nowadays instead of the Rigol DS2072 - except for the following reasons:

1) You need a DSO immediately - and you just don't have an extra +€300 to spend now. So your plan is to replace this 'starter' DSO in the future (although it's more money in the long run).
2) You only need to use the DSO for the simplest possible repairs - and you will never make any money from using it.
3) You absolutely have to have battery portability at the expense of all other features.

Otherwise, why not invest for your future - save up the extra couple of hundred - and buy a vastly superior scope? Since the Rigol DS2000 series has been introduced, buying one of these €400-€500 DSOs is a little like paying 50-75% less for a new automobile which - while it does actually get you around - only goes 5kph.

[aghp]

Here I find Rigol DS2102 (100MHz) around  1150,- (Batronix Offer)

Owon SDS7102V I can find around ~450 Eur from reputable sellers who may have also after sales support and warranty service locally and customer have posibility to test and check in semi professional lab if it pass for his needs.

Owon have naturally pros and cons. As all...

I'm littlebit surprice that so many  professionals for real work buy these Owons today.
I know some arguments but they are these common... good support, reliable, small size, big display, high sampling rate down to slow horizontal speeds, battery of course. Also in trawel use around of worls, not big loose if damaged or loosed...
But this is NOT for deep lab work for example catch some fast clitches from digital signal streams...  but I know lot of real industrial service work where this is just enough as F....e golden priced handhelds..   

[digsys]

Personally, I find it odd that anyone would buy an Owon or Hantek nowadays instead of the Rigol DS2072 - except for the following reasons:
2) You only need to use the DSO for the simplest possible repairs - and you will never make any money from using it.
3) You absolutely have to have battery portability at the expense of all other features.

Not starting anything here, but I find these comments over-the-top.
I bought the Owon because - It has a large hires VGA screen, it has a built-in battery, and pretty good specs. YES, it has limitations.
BUT, IF it was my only scope, I could easily make a shitload of money from it ! 40+ yrs in R+D. If I had a Tech who told me he could
only use an expensive scope, I'd sack him. A good Tech can work with whatever he's got, if NEED be, and do it WELL.
I bought the Owon mainly for EV motor / control calibration / tuning. IF I blow the scope, and failures in this game are quite serious,
I don't care. I'm NOT dragging my $30K or $50K LeCroys to the job !!

Since the Rigol DS2000 series has been introduced, buying one of these €400-€500 DSOs is a little like paying 50-75% less for a new
automobile which - while it does actually get you around - only goes 5kph.

I like analogies, but this is WAY off the mark.
I can state that, based on my last 40 yrs, the Owon would have been fine for say 90-95% of my work. YES, there are situations where the
Rigol would win hands down, but generally ANYONE who'd need that capability would be SAVVY enough to know what he (she) wants.
Sorry to butt in, but mis-information irks me. IMO, YMMV

[marmad]

I bought the Owon because - It has a large hires VGA screen, it has a built-in battery, and pretty good specs. YES, it has limitations.
BUT, IF it was my only scope...

Hmm.. it seems to me the pertinent information here is that it's NOT your only scope. For those of us who can't afford the really great expensive scopes (which you obviously own many of) and have to suffice with a single DSO - the difference between paying €400 or €700 - and the added features and capabilities you get for that increase in price - might be a big issue.

I like analogies, but this is WAY off the mark.

So, you've used the two scopes we're discussing in-depth?

I can state that, based on my last 40 yrs, the Owon would have been fine for say 90-95% of my work.

That's YOUR work - it's not mine - or anyone else's necessarily.

And, BTW, you seem to fit into my category #3 from above perfectly - second cheap scope for battery portability.

[marmad]

Here I find Rigol DS2102 (100MHz) around  1150,- (Batronix Offer)  Owon SDS7102V I can find around ~450 Eur from reputable sellers who may have also after sales support and warranty service locally and customer have posibility to test and check in semi professional lab if it pass for his needs.

1) You're comparing the model prices based on bandwidth. But why not compare on sample speed? 2GSa/s: DS2072 = €710 / SDS8202V = €635 (Batronix prices excl.) - a little bit closer in price. Or why not compare model prices based on wfrm/s? Or on the width of the horizontal scale, etc?

2) Unfortunately, although people have looked into, it doesn't seem to be possible to 'increase' your bandwidth (or other features) after purchase of Owon SDS models. This is not true for the Rigol DS2000 series. Also, although people are being quiet about it, the current firmware of DS2000 appears to have a bug which allows you to keep using the Options (56MPt memory, protocol decoders, extra triggers - an added €660 value) forever - or at least until you decide to upgrade your firmware

[digsys]

So, you've used the two scopes we're discussing in-depth?

Emona is just up the road from us, I did "play" with one and was impressed. Consequently, 2 of my friends have
recently bought the 2/4 channel Rigols, based on my recommendations. In these cases,
both work on high-end stuff MOST the time (S/Modes / Motor controller designs) and both had lotsa loot. No contest.
If you read my post carefully, I did NOT trash YOUR post. I just had issues with TWO of your "absolute definitive"
statements, for WHICH I supplied alternate examples. (=examples)

[marmad]

If you read my post carefully, I did NOT trash YOUR post.

Well, I guess that's open to interpretation - using phrases like "over-the-top" and "WAY off the mark" might be considered slightly 'trashy'    But perhaps you misunderstood my meaning: I wasn't suggesting you COULDN'T make money with an Owon - I was suggesting the opposite: that IF you can make money with your DSO, why not invest a little more now (which you will earn back) and get a much better tool for doing that? Whatever it's advantages are (and it certainly has some - I've recommended it in certain circumstances to many people, in both my posts and reviews), the Owon is just not a very good DSO for certain things (such as the design and debugging of digital electronics).

That post was aimed towards people that have to make do with a single DSO for several years. If I could justify two DSOs, I'd likely still have my SDS7102 for certain battery-operated situations. But if you know that you're only going to have one DSO to use for awhile - the decision to spend less or a bit more now, IMO, got radically different once the DS2000 series was introduced. You just can't look at the available options - and value per $/€ spent - the same as you could a year ago.

... and both had lotsa loot. No contest.

Again, I don't think this is an issue about 'lotsa' loot anymore - rather a bit more loot.

I just had issues with TWO of your "absolute definitive" statements, for WHICH I supplied alternate examples. (=examples)

Well, my 5kph automobile example was certainly partially a joke - I just came up with the figure off the top my head using Rigol's ~50,000 wfrm/s vs. Owon's ~40 wfrm/s - and their associated blind times - for inspiration - I never suggested that my analogy was "absolute definitive".  But I would suggest that my statement that the Rigol DS2072 is a vastly superior DSO to the Owon is definitive. Do you disagree with that?

[aghp]

Here some tiny tests with SDS7102V  HW/FW 2.8.2.

Here some tests with new SDS7102V version 2.8.2.

Note: before I have made risetime test with Tektronix generator. In specs it have <1.3ns risetime but I have measured it... it is far more fast (perhaps under 1ns). It is clearly faster than this now used HP.  So, this affect some amount and so these results are not directly comparable (exept if you think what itr affect... also pulse shape is littlebit different)

Risetime tests made using HP8161 pulse generator  around 1.3ns specified risetime and measured around 1ns. 2meter coaxial cable and Tektronix 50 0hm feed thru terminator in scope input. (this method is not best possible but littlebit better than BNC T with terminator and this is also comparable to my other tests. Note that with this method there is not 50ohm impedance becouse oscilloscope input is not "ideal". Specially in this case there is parallel capasitance what destroy matching. And this is case also with many cheap oscilloscopes what do not have true 50ohm impedance inputs but there is just relay and 50 ohm resistor without real impedance matching for whole frequency area as is case in many expensive Tek and HP ("Agilitylend" who nearly destroy HP to level "all can do these cheap plastic boxes"  and somebody who today try destroy Tek downcounting have started.)


All images, exept last one only CH1 in use.

Sweep tests and sinewaves are from Agilent 8644B signal generatorwhat have very good flatness and extremely good signal purity in this frequency area.
Same notes and connection as risetime test exept cable is around 1m.

Note: Voltage bands are I: 2mV to 100mV, II:200mV to 1V, III: 2V to 10V



V/div 50mV (band I), sweep start green line and in end it go back to starting freq. Sweep is linear and sweep time is 100s. Oscilloscope 10s/div. One div is 25MHz. Oscilloscope is in peak mode.
-3dB point is around 175MHz  related to 1MHz signal level.




V/div 500mV (Band II), sweep start green line and in end it go back to starting freq. Sweep is linear and sweep time is 100s. Oscilloscope 10s/div. One div is 25MHz. Oscilloscope is in peak mode.
-3dB point is over 200MHz (225MHz) related to 1MHz signal level. (note shape... perhaps if better impedance matching there is less "waves" in shape and it is possible then -3dB point may move down (or up))




Same settings and source level, sinewave 300MHz and infinite persistence for looking if some trig instability and/or vertical wobbler.
(freq counter max is around 320MHz and signal trig limit (single CH in use)  is around 450MHz (but out of signal level specifications for trig of course)



Risetime Voltage band I.   Signal shape is very good. Risetime (frequency response) is clearly better than specified. (<= 3.5ns)




Risetime Voltage band II.  Signal shape is good. Risetime (frequency response) is  clearly much better than specified. (<= 3.5ns) around 2.2ns in this test system. (real risetime is littlebit better but this moment I have not available better signal source. This can see if analyze image where is persistence and visible corner wobbling - also FFT tell same story)


Here now and later some more tests.


Opinions are extremely important but real data is.... sometimes also  good to know.
-aghp

[akis]

Hello, I have bought the OWON SDS7102V and have upgraded it to version 2.8.1. I am having a great problem with low input signals : the display is noisy and does not "lock on".

I have taken a picture although you need a video to see the problem clearly. Basically the curve moves all the time, jumps around, makes it impossible to read.

The picture is with a signal of 80mV pk and sensitivity of 20mV
With 800mV pk and sensitivity of 200mV it is more stable but still jumps occasionally and there is a bit of noise
With 8V pk and sensitivity of 2V it is very stable and almost no noise

Please help.

[aghp]

First

This image can not see nearly anything due to image quality.
If you have this quality of picture why you send it with this image size?
Owon have USB and you can take display image to USB and there can see all exactly.
In very rare cases problem solving need video. And more less if vidoe image is poor shaking quality as usual and no any other documentation.

But there is some mysterious in image what can detect.
Look first these trigger levels and more... visible is CH1 but it seems you have ALT mode in trig.

Please now first. Read user manual from beginning to end.

Then do factory setup. (utility -> adjust -> default.
Then connect ALL things out from every input.
Then do selfcal after oscilloscope have been least half hour on. And do not touch anything during selfcal. Specially do not touch input connectors.
Then do NOT connect anything to CH2
Then set probe settings in scope to 1x
Only CH1 on (CH2 off)
Push CH1 vertical position, it moves to center.
Push trig level knob, it moves trig to center level zero.
Turn CH1 for 5mV/div
Turn Time/div to 500us/div
Do not change any other setting  (after default and cal)
Take 1kHz sinewave from good and clean  source using coaxial cable with BNC
Connect 8mV p-p  1kHz sinewave to CH1 input.
You see accurate and stable triggered signal what stay on its place as rock.
Drop pp level to 4mV and change CH1 to 2mV/div
Still you see well triggered signal. If not turn trig level just small amount (+-1mV) up or down until it trigs.

Take direct pictures from screen so that things can see clearly.
If there is some jumping etc... you may use persistence for collect these to picture.

If NOT then we may start diagnose suspected problem inside scope.
Now I keep also supspected that problem is outside oscilloscope until more information.

Keep in mind that also probe may be broken and even if ok but it can easy also pick up noises if there is high noise fields.

[akis]

Thanks for your reply. I have spent a lot of time trying to set up the PC software so I can take those screen shots.

The oscilloscope was left running for an hour, then calibrated, then one probe inserted, just as you asked.

I have taken some snapshots but these are "still", they do not show you how much the wave "flickers" on the screen, I mean the wave jumps about all the time plus a lot of noise. The scope cannot "lock" on properly and not even with 500mV signals. There is a "click" when you turn the sensitivity to > 1V/div and then it is better.

If you look at the pictures below you can see the noise, but you cannot see how much the wave "jumps" about.

Picture 1: 2KHz sine wave - 80mV peak to peak
Picture 2: 2KHz sine wave - 40mV peak to peak

[aghp]

Perhaps you did not read my message?
Try follow it.

And I ask Owon screen shot. It is exactly copy from TFT using simply USB stick but ok now you have communication so...   you can take directly screen capture "image" (it is just as scope TFT copy) from Owon using this communication. (or with USB stick)
(image from display show all visible settings also so I can see where is your trigger level and what are other settings  (and do these settings hand, not pushing auto button.)

Then I ask you use display persistence for collect all changes to same image. (some 1 to 5 second persistence is ok in this case.

Then I ask connect signal using coaxial cable for avoid posibility that you collect some terrible noise from enviroment with possible fault probe etc


First, try get this pure clean signal first directly to oscilloscope input without probe if possible so that there is no any posibility about external noise.

Please read this my before post and try follow it.... take Owon Image (direct copy from screen) Just as my some pictures you have seen. If you do exactly then I know how to look and think your result.
Im interest about what we can see if there is sure clean signal in scope input... and just all settings done how I have ask. (Including this "default")

After this result there is then more tests but first this. 

[akis]

OK here are some pictures.

I think I have followed your instructions as much as possible. I have pressed 50% many times and also have manually turned the trigger knob, but nothing helps. I have connected the probe straight on to the tone generator. I have used two probes to make sure.

As you can see, even at 800mV there are issues. Screen persistence set to 5 sec. The last picture showed almost OK on the screen but the image was not that OK.

[aghp]

1.
Ok, please do same things (one setup is enough) with only CH2 in use and of course trig also selected CH2. 

Your probe is 1x and oscilloscope probe setting 1x?

One or two image with CH2 (same signal etc as with CH1)

Then do "default"


2.
Then

Connect just both probes to 1kHz probe comp output. Probe 10x and setting 10x (as default also)

Press just "auto" and select more or less periods visible.

Do not adjust anything!  Only set display persistence to 5 second.

Then  just one image.


3.
Then
agen default scope.

Do not connect anything to inputs.
Turn CH1 and CH2 to 20mV  (it automatically turn BW on)  (it is most sensitive setting 2mV but there is default 10x for probe)
Then set memory to 10M
Then set acquire mode to "Peak"
Time/div to 500us/div
Keep hands away from near inputs.
Press "Stop"
Change time/div to 1ms/div it zoom out.. (keep scope acquire stopped)
(you have now 10div long traces)

take Image.

[akis]

OK. I did as you said :-) Again, many thanks for taking the time to help me.

[aghp]

You tell that you upgrade FW to 2.8.1.
What version it was before?
What is your whole serial number without last three digits?
How it works just before upgrade?
Have you find that problem start just after upgrade or just random time?

Lowest voltage band is 2mV to 100mV/div (if in scope input settings Probe:Atten:X1 (and directly input BNC or we have 1:1 probe (X1)))

(analog front end is divided to three voltage bands:
I: 2mV - 100mV
II: 200mV - 1V
III: 2V - 10V
All these have different signal pathway in analog front end.
Also  trigger signal is splitted (before ADC) there to trigger circuits. )


It looks like there is now something wrong. (perhaps in Voltage band I but still all is possible)


2.1.
Do first scope default (Utility:adjust:default) and then only next things:

Connect probe to CH1. 
Turn probe switch to X10
Connext probe tip to Probe Comp
turn Sec/Div to 100us (I want see also shape)
 Display :persist:5s
(Display):Trig Freq: ON
Push Trigger "Menu" button once.

Then save "Image" with PC (in this time not with USB stick so I see trig settings bottom of screen)


2.2.
Then only touch CH1 Volts/Div knob
Turn it one step to 2V

Then save "Image" with PC agen.

Show these images.

If both give well trigged result there need continue some more tests. but if 2.2  or 2.1 (or both) give clear bad result then game over = scope need repair. 
(of course there can do endless amount of tests but just after it is proofed there is fail do not waste time for testing.

[akis]

Version was 2.6.2 and serial is SDS7102V 1211675

It was the same before the upgrade, actually I thought the upgrade would fix it. Basically when the signal level is low (below 2V peak to peak) then the scope has trouble in displaying it, or locking on to it, or both.

Is there a way to put it back to 2.6.2 maybe so as to test again?


Following your instructions again:

2.1 pictures (both from PC and USB stick) = SN1.jpg and SN1-image.jpg

2.2 pictures = sn2.jpg and sn2-image.jpg


At the bottom of the screen, when I pressed "Trigger Menu"  it said:
Single-edge, source-ch1, coupling-dc, slope-up, auto&holdoff, 1=1.000000KHz       1=_/ 0.00mV

[aghp]

Version was 2.6.2 and serial is SDS7102V 1211675

It was the same before the upgrade, actually I thought the upgrade would fix it. Basically when the signal level is low (below 2V peak to peak) then the scope has trouble in displaying it, or locking on to it, or both.

But now we have 500mVp-p and it trigs perfect.
Becouse probe comp give 5Vp-p and probe is set 10:1 so oscilloscope BNC input level is 0.5Vp-p

With this level it trig ok with both voltage bands (I and II)

(I can see in image something what is not ok but I do not know if these noise spikes are from oscilloscope itself or picked from your enviroment. )

You have battery. Use it with battery. Go to place what is as clean as practically possible  from all (electric) noise . (no computer, no switch mode "energy save" lightning etc etc )
No need post image now, only watch if these small  noise spikes disappear (or change).

If you see these small spikes disappear or just stay, only tell it.

Then, whatever this show...
tell me if you have available reliable clean adjustable source for around 10 - 500mV p-p level  square wave around 1kHz to higher frequencies. (if it have "50ohm" output then 5-250mV)
It need be so that you really know output level and that it really is clean from high noise etc.
(example some reasonable good function generator. And do you have available coaxial cable directly from generator output BNC to scope BNC. 
If not... and still you have generator. It can connect with probe but probe GND need connect to BNC GND if generator have floating output! Not to generator ground.

(idea is that we need find reliable enough good quality signal for test trig with lower level. If we can not trust test signal it is waste of time and may lead only confusing)

Is there a way to put it back to 2.6.2 maybe so as to test again?

There not bridge left to go back with normal upgrade process.
My opinion is: it is not useful to go back in this case.

[akis]

Hello

I have disconnected the mains lead and took the scope around the house, the spikes remain. I have tried 4 different probes, CH1 and CH2, and the noise remains.

In the 1V/div setting, those spikes sometimes cover 25%-45% of the division, that is like 500mV "noise"!!!

The one thing that makes it slightly better is to switch the probe to 20MHz. Still I cannot display a 8mV pk-pk waveform or even a 80mV pk-pk because of all the noise.

I do not have a reliably clean source of low level signal to try with but I remember the old scope (20MHz analogue) was working quite well on those low level signals.

What would you suggest as next step now?

Thanks

[aghp]

Hello

I have disconnected the mains lead and took the scope around the house, the spikes remain. I have tried 4 different probes, CH1 and CH2, and the noise remains.

In the 1V/div setting, those spikes sometimes cover 25%-45% of the division, that is like 500mV "noise"!!!

The one thing that makes it slightly better is to switch the probe to 20MHz. Still I cannot display a 8mV pk-pk waveform or even a 80mV pk-pk because of all the noise.

I do not have a reliably clean source of low level signal to try with but I remember the old scope (20MHz analogue) was working quite well on those low level signals.

What would you suggest as next step now?

Thanks

"The one thing that makes it slightly better is to switch the probe to 20MHz"

Without probe we see not really bad noise.

Probe connected to input - high noise.

It looks like probe itself collect noise from somewhere.
Try short circuit probe center pin directly to probe gnd.
Like image, so that short circuit wire make reliable contact and no extra wire what make any antenna.

How is noise now?
In my lab I have one energy saving fluorescent lamp where current control is made by switch mode regulator. With open probe I can collect lot of noise example 1 meter distance from this lamp.. With shorted probe.. just nothing.

Now if shorted probe is nearly without noise then it need find this noise source. Scope itself or something in your enviroment?

[aghp]

Something about this old analog and then digital and noise.


Tektronix:

With analog oscilloscopes with CRT displays, the extreme ranges of noise are not displayed, because they occur quickly and infrequently, which means that the phosphor is lit quickly and infrequently, and thus those extremes are very dim or not drawn on the screen at all.

This means that analog instruments are not just displaying voltage against time, but in fact have a third dimension: Intensity. Intensity is related to frequency of occurrence of the signal. DSO oscilloscopes will show every hit with the same intensity, no matter how frequently a particular pixel is hit.

Now you have oscilloscope what is very sensitive and it have wide analog bandwidth before ADC.
If example in low speed (low samplerate) you see high noise there is also posibility that it is HF noise far over half of samplerate and what you see is alias produced by this noise.

Also if noise is random and high freq you really can not detect it at all from old analog oscilloscope phosphor. Simply it is far away too slow and perhaps also scope bandwidth is very limited compared to your Owon. If there is going high level 400MHz signal inside to scope you see it. (with low speed you may see its alias. And it is not limited just to 400MHz. You perhaps have seen FFT images from some tests. (example sweeped FFT freq response image. Also it is good to note that probe is good antenna specially very confusing if there is higher frequencies "on the air".

Here is this small tektronix pdf.. it may give some idea and it can think also for higher level displayed "noises".

I can make one image to you where Owon is collecting noise just with example open probe or short circuited with probe own GND wire. (it is now nice "loop" antenna.)
Probe GND is ok only for audio frequencies if you work in emi free enviroment.

Soon later I make some for just example.

[aghp]

Some fun with enviroment noise.

Around 80cm to  table  is one fluorescent light... on the table ESD srface  is open Owon probe x1
Light on and light off. NOTE: There is still one noise source lamp on around  bit over 2m  distance from table.

Then IBM laptop power supply and probe short circuited with its own GND wire.
Same with more fast speed. (and persistence 5s) It mostly somehow trig becouse this noise have so clear pulses.

Last some old Owon FFT sweep.
From Zero to 500MHz (500MHz is Nyquist "wall" with 1GSa/s) (if there is signal over 500MHz it is folded back to under 500MHz, this is also good to understand if use FFT)
Now there is 1GSa/s sampling and there can see how signal attenuate... but still not so much at 500MHz. Of course now if rise signal level it may rise to full, of course.
It is very important to realize... this analog "highway" bandwidth  from BNC to ADC is just this with all samplerates! If there is AM or FM station near, if there us UHF TV transmitter near... or some other RF sources.. it may lead what ever kind of problems and it neet take care as designing how to connect probe and how it may affect measurements.

Independent of sampling speed this analog front end have  wide BW road from BNC input to ADC input.
So, of course all noises can also produce high amount of noise and it may be direct noise or noise alias image. 
Now of course if noise is high frequency signals it can  reduce something if turn 20MHz BW but also this 20MHz BW filer on but also it only attenuate example 30MHz or 100MHz.. but it can not stop these.

It is more than one lesson to learn many kind of methods for right "probing". In many cases it is also difficult for peoples who have lot of experience with many kind of lab works in theory and in practice.

But... still we do not now know IF your scope have some real problem or what is this case.

I remember some people who have lot of experience with old 20MHz whatever name analog oscilloscope and then he get new digital oscilloscope and he see things what he have never seen before... he was so confused that he turn back to his old analog after small repair it.

[akis]

OK here is another picture.

I have used the oscilloscope's own 1KHz probe comp with a fixed resistor potentiometer with the two resistors straight on the pins, and here is the noise . The pot is 10K/470R.

This picture was taken with config/default, acquire 10M, method:peak, probe X1, horz=100us