Author Topic: Rigol MSO5000 HiRes mode  (Read 10716 times)

0 Members and 1 Guest are viewing this topic.

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Rigol MSO5000 HiRes mode
« on: October 29, 2022, 10:50:16 am »
Calling Rigol MSO5000 owners...

Could some kind-hearted MSO5000 owner post screenshots of the Hires mode in use to show what effect is has on the ADC noise?

The Hires mode has various settings (from 2x to 32x?), it would be good to have a post that can serve as a reference for when the eternal "noise" question raises its head. I'm thinking of a series of screenshots at, say, 1us/div horizontal, no input connected, starting with normal sampling mode then enable Hires mode and gradually turn it up to maximum.

Ideally there should only be one more post in this thread.  :popcorn:

(or maybe a second one with more screenshots confirming the first)

 

Offline voltsandjolts

  • Supporter
  • ****
  • Posts: 2547
  • Country: gb
Re: Rigol MSO5000 HiRes mode
« Reply #1 on: October 29, 2022, 11:33:50 am »
Ideally there should only be one more post in this thread.  :popcorn:

Agreed.
 

Offline balnazzar

  • Frequent Contributor
  • **
  • Posts: 417
  • Country: it
Re: Rigol MSO5000 HiRes mode
« Reply #2 on: October 29, 2022, 11:54:24 am »
 

Offline balnazzar

  • Frequent Contributor
  • **
  • Posts: 417
  • Country: it
Re: Rigol MSO5000 HiRes mode
« Reply #3 on: October 29, 2022, 12:11:12 pm »
Come on, there is no need to get angry. It was just funny that the user above replied that way..  ;D

I'm interested in detailed Hi-Res screenshots of the MSO5000 too. You did good in asking for them, I was thinking about doing the same.
 

Offline gf

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: de
Re: Rigol MSO5000 HiRes mode
« Reply #4 on: October 29, 2022, 12:12:00 pm »
The Hires mode has various settings (from 2x to 32x?)

Is 32x the limit? If no, what is the limit?

32x is supposed to reduce the voltage amplitude of white noise by factor sqrt(32) = 5.66.
A 32-tap moving average filter has a -3dB cut-off frequency of ~110MHz, if sample rate is @8GSa/s.
[ And a 20MHz filter would require 175 taps, and could reduce white noise amplitude by factor 13.2 ]

If this numbers cannot be achieved (approximately) in practice, then something weird must going on.
I hope, hires does store the samples in memory with more than 8-bit precision, does it?
« Last Edit: October 29, 2022, 12:16:51 pm by gf »
 

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #5 on: October 29, 2022, 12:14:37 pm »
The Hires mode has various settings (from 2x to 32x?)

Is 32x the limit? If no, what is the limit?

What the manual actually says is this:


That might be 16x, but I've seen 32x mentioned in other places ...  :-//

Maybe we could wait for a MSO5000 owner to post some screenshots, then we'll all know for sure. A screenshot is worth a thousand forum posts.
« Last Edit: October 29, 2022, 12:25:35 pm by Fungus »
 

Offline gf

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: de
Re: Rigol MSO5000 HiRes mode
« Reply #6 on: October 29, 2022, 12:32:23 pm »
That sounds like 32x, but...  :-//

If I take the words literally, then "does not exceed" means "less than", so it would mean that the BW is (generally?) less than fs/32, or BW < 250MHz @8GiSa. But how much less?

[ And a -3dB BW of fs/32 would actually not imply 32x, but rather ~14 moving average taps (which is admittedly an unlikely number) ]

Furthermore, the manual does not say, that the factor can be configured.

Quote
Maybe we could wait for a MSO5000 owner to post some screenshots, then we'll all know for sure. A screenshot is worth a thousand forum posts.

And to answer the question, whether the factor is really configurable, and what's the limit.
« Last Edit: October 29, 2022, 12:40:15 pm by gf »
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Rigol MSO5000 HiRes mode
« Reply #7 on: October 29, 2022, 12:34:52 pm »
Calling Rigol MSO5000 owners...

Could some kind-hearted MSO5000 owner post screenshots of the Hires mode in use to show what effect is has on the ADC noise?

The Hires mode has various settings (from 2x to 32x?), it would be good to have a post that can serve as a reference for when the eternal "noise" question raises its head. I'm thinking of a series of screenshots at, say, 1us/div horizontal, no input connected, starting with normal sampling mode then enable Hires mode and gradually turn it up to maximum.
IMHO you should no longer pursue this quest of yours trying to find something that simply isn't there. In the end hi-res is nothing more than an adjustable low-pass filter. Because of the filtering you'll lose details just as when they are buried in noise. For seeing fine details in a signal, there is no substitute for having a low-noise signal path.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 
The following users thanked this post: tautech, Martin72

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #8 on: October 29, 2022, 12:54:22 pm »
IMHO you should no longer pursue this quest of yours...

And so it begins.  :palm:

Clue: I just started an entire thread where somebody could prove definitively that I'm wrong about something. I've completely exposed myself to that possibility, in public.

If you're so sure of the results then maybe you could just wait for the screenshots to appear then say "I told you so".

Then again, maybe it's not me that's on any sort of "quest".
 

Offline balnazzar

  • Frequent Contributor
  • **
  • Posts: 417
  • Country: it
Re: Rigol MSO5000 HiRes mode
« Reply #9 on: October 29, 2022, 01:24:22 pm »
IMHO you should no longer pursue this quest of yours...

And so it begins.  :palm:

Clue: I just started an entire thread where somebody could prove definitively that I'm wrong about something. I've completely exposed myself to that possibility, in public.

If you're so sure of the results then maybe you could just wait for the screenshots to appear then say "I told you so".

Then again, maybe it's not me that's on any sort of "quest".

I think we could just ask Martin72 to do that if and when he has time and will to spare. He has the scope & he knows how to test it against small signals.
 

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #10 on: October 29, 2022, 03:59:01 pm »
If I take the words literally, then "does not exceed" means "less than", so it would mean that the BW is (generally?) less than fs/32, or BW < 250MHz @8GiSa. But how much less?

The manual doesn't parse very well. I think it might refer to the displayed signal, ie. what you see on screen, but it's just an ignorant guess.

The Analog bandwidth will still be 350Mhz and the ADC will still be sampling at 8Ghz, so...

Edit: Maybe could be tested in part two of this thread.

« Last Edit: October 29, 2022, 04:01:15 pm by Fungus »
 

Offline Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Rigol MSO5000 HiRes mode
« Reply #11 on: October 29, 2022, 04:17:20 pm »
I don´t know why opening a new thread as we already discussed this in the huge MSO5000 thread..
Short form:

- Hi-Res was later on via firmware update somehow implemented, before MSO5000/DSO(MSO)7000 didn´t got that(Dave: "Where is the Hi-Res Mode ??!!")
- No chooseable bits like other brands got (+0.5, +1, 1.5....+3Bit)
- No further informations about from rigol
- Works somehow..
- No effects when using this as I did the 1mV ripple-measures in several conditions of the scope


Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #12 on: October 29, 2022, 05:17:38 pm »
- Hi-Res was later on via firmware update somehow implemented, before MSO5000/DSO(MSO)7000 didn´t got that(Dave: "Where is the Hi-Res Mode ??!!")

It's here now though, let's see what it does!

- No chooseable bits like other brands got (+0.5, +1, 1.5....+3Bit)
- No further informations about from rigol
- Works somehow..

So?

- No effects when using this as I did the 1mV ripple-measures in several conditions of the scope

Yeah, we all know it doesn't have sensitivity at 1mV. How about a bit more than that, or just an open input?
« Last Edit: October 29, 2022, 05:19:17 pm by Fungus »
 

Offline balnazzar

  • Frequent Contributor
  • **
  • Posts: 417
  • Country: it
Re: Rigol MSO5000 HiRes mode
« Reply #13 on: October 29, 2022, 05:51:33 pm »
- No chooseable bits like other brands got (+0.5, +1, 1.5....+3Bit)
- No effects when using this as I did the 1mV ripple-measures in several conditions of the scope

That's bad.
 

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #14 on: October 29, 2022, 06:10:26 pm »
- No chooseable bits like other brands got (+0.5, +1, 1.5....+3Bit)
- No effects when using this as I did the 1mV ripple-measures in several conditions of the scope

That's bad.

How is "x2, x4, x8, x16" less useful than "+1.5 bits"?

All that matters is what you see on screen. Where's the screenshots?
 

Offline balnazzar

  • Frequent Contributor
  • **
  • Posts: 417
  • Country: it
Re: Rigol MSO5000 HiRes mode
« Reply #15 on: October 29, 2022, 06:28:56 pm »
- No chooseable bits like other brands got (+0.5, +1, 1.5....+3Bit)
- No effects when using this as I did the 1mV ripple-measures in several conditions of the scope

That's bad.

How is "x2, x4, x8, x16" less useful than "+1.5 bits"?

All that matters is what you see on screen. Where's the screenshots?

Here is my opinion, but of course it's a noob-ish opinion:

1. Knowing xN is not less useful than knowing the number of additional "bits", but the problem is, AFAIK, that you cannot *select* different levels. Selecting x2 could be very useful, whereas if the HR mode operates always in x16 it can squash your signal or flatten it too much.

2. We don't even know if it's x16 or x32. Again, AFAIK.

3. It's not well documented and that's annoying by itself.
 

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #16 on: October 29, 2022, 07:20:53 pm »
AFAIK, that you cannot *select* different levels.

No level selection on the MSO5000? That's not ideal...

The firmware change log claims "12-bit high resolution mode".

Me? I'm not expecting 12 bits, or even 10 bits... but there should be a visible step in noise reduction.

Even if it's only 8 or 9 bits it goes from the realm of "noisy beast" into the realm of "quiet" oscilloscopes.
« Last Edit: October 29, 2022, 07:25:52 pm by Fungus »
 
The following users thanked this post: balnazzar

Offline Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Rigol MSO5000 HiRes mode
« Reply #17 on: October 29, 2022, 07:35:12 pm »
Quote
3. It's not well documented and that's annoying by itself.

"Not well" is too optimistic...
According to the data sheet and usermanual you only find two sentences with interesting informations:
12 bit max.
Bandwith up(or down..) to 1/32 of samplerate
That´s all and that is very less, should the customer do their homework and do some research...
As I´ve owned one I did some tests as the new firmware came up:
https://www.eevblog.com/forum/blog/new-rigol-scope/msg2233980/#msg2233980
When I see this today, with the knowledge of today, it annoys me not to have tested it differently. ;)
But we have enough people here with a MSO5000....

 
The following users thanked this post: Fungus, balnazzar

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #18 on: October 29, 2022, 08:09:55 pm »
As I´ve owned one I did some tests as the new firmware came up...

Thanks!

That looks to me like only one or two bits improvement, not consistent at all with the number "16" (which should give three or four bits).

 

Offline gf

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: de
Re: Rigol MSO5000 HiRes mode
« Reply #19 on: October 29, 2022, 08:20:27 pm »
Quote
3. It's not well documented and that's annoying by itself.

"Not well" is too optimistic...
According to the data sheet and usermanual you only find two sentences with interesting informations:
12 bit max.
Bandwith up(or down..) to 1/32 of samplerate
That´s all and that is very less, should the customer do their homework and do some research...
As I´ve owned one I did some tests as the new firmware came up:
https://www.eevblog.com/forum/blog/new-rigol-scope/msg2233980/#msg2233980
When I see this today, with the knowledge of today, it annoys me not to have tested it differently. ;)
But we have enough people here with a MSO5000....

The rise/falltimes and the straight line shape of the edges in the 4th screenshot https://www.directupload.net/file/d/5378/rtdacdlo_png.htm look plausible for a 16x moving average (when compared to the 3rd screenshot). It is however hard to assess w/o vertical zoom-in, whether the noise amplitude went down by a factor of 4 as well. Rather looks like less.
 
The following users thanked this post: balnazzar

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #20 on: October 29, 2022, 08:43:26 pm »
As I´ve owned one I did some tests as the new firmware came up:
https://www.eevblog.com/forum/blog/new-rigol-scope/msg2233980/#msg2233980

A few pages further down you posted some more screenshots with similar results:

https://www.eevblog.com/forum/blog/new-rigol-scope/msg2264154/#msg2264154
 

Offline FungusTopic starter

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Rigol MSO5000 HiRes mode
« Reply #21 on: October 29, 2022, 09:11:18 pm »
If they claim 12 bits then that's an improvement of 4 bits over their 8-bit ADC.

Four bits improvement indicates 16x oversampling... that gives us a source for the number "16".

I'm not seeing 4 bits of improvement in those screenshots though.
 

Offline Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Rigol MSO5000 HiRes mode
« Reply #22 on: October 29, 2022, 09:58:33 pm »
Quote
According to the data sheet and usermanual you only find two sentences with interesting informations

Searching in the data sheets/user manuals of the 7000 and 8000 for eventually more, nothing...
8000 "says" the same and in 7000 the hi-res mode is not mentioned.. ???
Checked at last the DS1000Z: Same as 5000.

Quote
A few pages further down you posted some more screenshots with similar results:

Today, with the knowledge I got now about it, I would do the tests in a different way.
The test with the squarewave is good and can repeated when it´s interesting to know where the bandwithlimit lies.
Next could be a test like I did one, two weeks ago with a very small signal to see an improvement against non-hires mode.
Assuming the same method is used, I can test it with the DS1054Z at work next week.
But best would be testing with the 5000 itself.
 
The following users thanked this post: egonotto

Offline Anding

  • Contributor
  • Posts: 46
  • Country: hk
Re: Rigol MSO5000 HiRes mode
« Reply #23 on: October 29, 2022, 10:45:30 pm »
What are the instructions to run this test?   

Assume

New MSO50104 with latest firmware and factory options installed, 100MHz bandwidth
Supplied Rigol probes
External Hantek AWG (in the handheld oscilloscope), if needed
One BNC to BNC cable
One BNC to crocodile clips
The usual breadboard, resistors, etc. if needed

Basic familiarity with operating the MSO5000 at this point.  But just tell me the circuit, waveform and parameters.  By the way, how do we judge the noise?  A fatter trace - according to Dave that may be sensitivity?  https://youtu.be/Znwp0pK8Tzk  Better to obtain some statistics if someone can kindly can suggest how
« Last Edit: October 29, 2022, 10:50:40 pm by Anding »
 
The following users thanked this post: egonotto

Offline Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Rigol MSO5000 HiRes mode
« Reply #24 on: October 29, 2022, 11:29:35 pm »
Hi,

First and very simply, you can check it´s own noise with and without hi-res mode, visually.

Then you can do it like this here:

https://www.eevblog.com/forum/testgear/is-a-rigol-mso5000-overkill-for-a-hobbyist/msg4481968/#msg4481968

And trying other waveforms like other members did in the linked thread.

Or something different - goal is to understand how the hires mode of the rigol works in real.
Rigol "told" very less about it, but main thing seems that it is depending on the samplerate.
So your tries should including a force down of the samplerate by e.g. adding more channels at the same time (reduces the max. samplerate, regardless if they are used or not).


 
The following users thanked this post: egonotto, Anding


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf