Author Topic: Rigol DHO800 memory depth / bandwidth upgrades and noise level!  (Read 18038 times)

0 Members and 1 Guest are viewing this topic.

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #25 on: November 06, 2023, 06:57:04 pm »
The results are in. Here's the RMS noise level of both bandwidth options, with and without bandwidth limiter, nothing connected to the inputs.

There's surprisingly little difference, only about 5%. I repeated it three times just to be sure, I honestly expected more.



« Last Edit: November 06, 2023, 07:18:23 pm by Fungus »
 
The following users thanked this post: egonotto, Serg65536

Online voltsandjolts

  • Supporter
  • ****
  • Posts: 2549
  • Country: gb
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #26 on: November 06, 2023, 07:01:20 pm »
I see two screenshots, both at 20MHz bandwidth, showing same noise as expected.
When you say 'nothing' on the inputs, do you mean floating inputs, or 50 Ohm terminated inputs?
 

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #27 on: November 06, 2023, 07:05:33 pm »
I see two screenshots, both at 20MHz bandwidth, showing same noise as expected.

Look at channel 2.  ;)
 
The following users thanked this post: voltsandjolts

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #28 on: November 06, 2023, 07:16:03 pm »
I see two screenshots, both at 20MHz bandwidth, showing same noise as expected.
When you say 'nothing' on the inputs, do you mean floating inputs, or 50 Ohm terminated inputs?

There's two channels on each...

Floating inputs.
 
The following users thanked this post: voltsandjolts

Offline Serg65536

  • Regular Contributor
  • *
  • Posts: 134
  • Country: ua
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #29 on: November 06, 2023, 07:17:41 pm »
Yes ADC - duh.  The DST time shift always puts me off.

And I didn't mean to infer the lack of dual ADC's made it unusable, only that sampling rate seems to be an otherwise high priority spec.
I've got you wrong, sorry.
Yes, BW to SR ratio is not good here.

But it's always nice to have the ability to capture a signal with the maximum available BW and full 1.25 GS/s.

IMHO, it's more disappointing to have very high noise due to 8 GS/s ADC, and use an oscilloscope below 100 MHz on daily basics.
 

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #30 on: November 06, 2023, 07:33:40 pm »
The results are in. Here's the RMS noise level of both bandwidth options, with and without bandwidth limiter, nothing connected to the inputs.
There's surprisingly little difference, only about 5%. I repeated it three times just to be sure, I honestly expected more.

The lack of difference might be due to the fact that decimated screen data are used for the RMS calculation? (I.e. the decimation acts as another bandpass filter.)

How do the results change if you change the timebase? Also, doesn't the vertical amplification go up to 500 µV/div -- does that change the result?
 
The following users thanked this post: egonotto

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #31 on: November 06, 2023, 07:37:46 pm »
Store the data in the .csv and make the rms calc in the excel, for example.
The 40uV rms noise means aprox 270uVpp - that is what you should see on the screen then..
Also to measure noise with the open inputs is not the best approach, imho.
Readers discretion is advised..
 

Online voltsandjolts

  • Supporter
  • ****
  • Posts: 2549
  • Country: gb
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #32 on: November 06, 2023, 08:02:05 pm »
Floating inputs are random number generators, use 50 Ohm termination.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #33 on: November 06, 2023, 08:17:24 pm »
You cannot measure noise on 1µs/div. You are high-pass filtering low frequencies out. Pretty much all of the 1/f noise which is actually large part of noise in scope amps..

Set scope at 1 ms/div, 1mv/div, and NO 20 MHz BW limiter.
To see what is actual noise with 100 MHz BW.

Then repeat with 20 MHz limit.

And if you look at your stats, you have very large difference between max and min AC RMS value.
This is the point where you reset stats and see again. If you again accumulate have such a large difference between max and min AC RMS something is wrong.
Make sure to put a non shorting cap  (or wrap a piece of aluminium foil over BNC) on input so there is not induced voltage.
Shorting input with 50 Ω is not going to show what noise is with 1MΩ inputs but less, which is not what you going to see with passive probes connected.
Whenever i have shown any data from any scope that is how it's done.
50Ω measurements are separate set of data.

If you still have large excursions in AC RMS, maybe you have some popcorn noise or something.. Or EMC problem.



« Last Edit: November 06, 2023, 08:25:48 pm by 2N3055 »
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 
The following users thanked this post: egonotto

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #34 on: November 06, 2023, 08:19:27 pm »
The results are in. Here's the RMS noise level of both bandwidth options, with and without bandwidth limiter, nothing connected to the inputs.
There's surprisingly little difference, only about 5%. I repeated it three times just to be sure, I honestly expected more.

The lack of difference might be due to the fact that decimated screen data are used for the RMS calculation? (I.e. the decimation acts as another bandpass filter.)

How do the results change if you change the timebase? Also, doesn't the vertical amplification go up to 500 µV/div -- does that change the result?

[4]: 500 μV/div is a magnification of 1 mV/div setting. For vertical accuracy calculations, use full scale of 8 mV.
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #35 on: November 06, 2023, 08:34:11 pm »
How do the results change if you change the timebase?

I did it at 1us/div. because that's what Dave always does it at.  :-//

Also, doesn't the vertical amplification go up to 500 µV/div -- does that change the result?

I was told that 1mV and 500uV were just software expansions of the 2mV range. I can't find it in the dtasheets though. I'm sure  2N3055 will fill us in on the details.

 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #36 on: November 06, 2023, 08:36:37 pm »
How do the results change if you change the timebase?

I did it at 1us/div. because that's what Dave always does it at.  :-//

Also, doesn't the vertical amplification go up to 500 µV/div -- does that change the result?
I was told that 1mV and 500uV were just software expansions of the 2mV range. I can't find it in the dtasheets though. I'm sure  2N3055 will fill us in on the details.

See my post above. Copied from the datasheet. I'm too lazy to retype that much. :-DD

[4]: 500 μV/div is a magnification of 1 mV/div setting. For vertical accuracy calculations, use full scale of 8 mV.
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #37 on: November 06, 2023, 08:41:55 pm »
[4]: 500 μV/div is a magnification of 1 mV/div setting. For vertical accuracy calculations, use full scale of 8 mV.

Ok, it was the bottom two scales on the DHO900, only the bottom scale on the DHO800.

I found it at the bottom of the datasheet:
 
The following users thanked this post: 2N3055

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #38 on: November 06, 2023, 08:42:19 pm »
How do the results change if you change the timebase?

I did it at 1us/div. because that's what Dave always does it at.  :-//

Also, doesn't the vertical amplification go up to 500 µV/div -- does that change the result?

I was told that 1mV and 500uV were just software expansions of the 2mV range. I can't find it in the dtasheets though. I'm sure  2N3055 will fill us in on the details.

It was not meant to offend you.
I just pointed out the facts. Even 1ms/div is kind of short really.. I would suggest doing it on longer timebases too. Just to see how much low frequency noise is there.
As I said before, much as  I like Dave, he makes mistakes too and is not always doing lots of preparations before doing videos.



"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 
The following users thanked this post: Krisztián

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #39 on: November 06, 2023, 08:44:59 pm »
Ok, so 1 mV/div is the highest "real" amplification. I think the DSO900 has an additional magnified step at the bottom end, 200 uV/div.

If we assume that the scope does its noise calculations from the acquisition buffer (as opposed to the screen data like the DS1054Z), then using the magnified settings should neither improve nor falsify the results.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #40 on: November 06, 2023, 08:52:37 pm »
Ok, so 1 mV/div is the highest "real" amplification. I think the DSO900 has an additional magnified step at the bottom end, 200 uV/div.

If we assume that the scope does its noise calculations from the acquisition buffer (as opposed to the screen data like the DS1054Z), then using the magnified settings should neither improve nor falsify the results.

All of the DHO scopes use same front end amp. They all have best 1mV/div. 200 and 500µV/div are software magnifications. 800/900 and 1000/4000...

As as side not these are the covers I use. E-bay few bucks for a few of them..
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #41 on: November 06, 2023, 09:07:12 pm »
As as side not these are the covers I use. E-bay few bucks for a few of them..

Here's mine:
 
The following users thanked this post: 2N3055

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #42 on: November 06, 2023, 09:10:47 pm »
The results are in...

70Mhz full bw: 61.8uV
100Mhz full bw: 64.2uV

70MHz with 20Mhz limit: 56.38uV
100MHz with 20Mhz limit: 56.83uV

Let the arguments begin!
« Last Edit: November 06, 2023, 09:12:35 pm by Fungus »
 
The following users thanked this post: egonotto

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #43 on: November 06, 2023, 09:21:58 pm »
This is all channels at 100MHz bandwidth. CH1 and CH2 have tinfoil, CH3 and CH4 don't.

The winner is: CH3 - no tinfoil!  :popcorn:



I put a 50 Ohm terminator on CH3 and it went down to 38.8uV average.
« Last Edit: November 06, 2023, 09:24:16 pm by Fungus »
 
The following users thanked this post: egonotto

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #44 on: November 06, 2023, 09:25:27 pm »
Try average for acquisition...

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #45 on: November 06, 2023, 09:46:13 pm »
This is all channels at 100MHz bandwidth. CH1 and CH2 have tinfoil, CH3 and CH4 don't.
The winner is: CH3 - no tinfoil!  :popcorn:
I put a 50 Ohm terminator on CH3 and it went down to 38.8uV average.

Could you also measure the residual offset voltage (V_avg) on each channel? It will enter the V_rms calculation and might have a non-negligible effect. Some of the traces do not look perfectly centered to their respective 0V arrows.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #46 on: November 06, 2023, 09:49:06 pm »
Try average for acquisition...

I set 32768 averages and it went down into the nV range. Rigol FTW!

« Last Edit: November 06, 2023, 09:52:25 pm by Fungus »
 

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #47 on: November 06, 2023, 09:50:31 pm »
@ebastler:
All except channel 3, maybe he had forgotten to make a self-cal after his playaround with the bandwith down/upgrades.
Btw, the title of the topic is somewhat irritating...

Edit:

Quote
and it went down into the nV range.

No wonder, given the number of passes selected.
But then you also have to question the plausibility of the result.
« Last Edit: November 06, 2023, 09:52:53 pm by Martin72 »
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Rigol DHO800 - to hack or not to hack!
« Reply #48 on: November 06, 2023, 09:53:50 pm »
Btw, the title of the topic is somewhat irritating...

Yeah, I thought that, too. Does it change if somebody changes it?

Edit: Apparently not. Maybe is has to be OP...
« Last Edit: November 06, 2023, 09:57:51 pm by Fungus »
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #49 on: November 06, 2023, 09:54:58 pm »
This is all channels at 100MHz bandwidth. CH1 and CH2 have tinfoil, CH3 and CH4 don't.

The winner is: CH3 - no tinfoil!  :popcorn:



I put a 50 Ohm terminator on CH3 and it went down to 38.8uV average.

These are nicely made  measurements. Now AC RMS is stable and measurements look valid.
Now people can compare to other scopes with same settings..

FUN FACT: RMS measurements (AC+DC and DC only RMS) will work absolutely perfectly even with scope undersampling. They work by integrating (averaging) and if you just take enough samples it will converge to right result. It is valid technique used in many instruments.


"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf