Author Topic: My Rigol DS1054Z has noise and beam thickness dependency of the vertical offset  (Read 4475 times)

0 Members and 1 Guest are viewing this topic.

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
I recently purchased a Rigol DS1054Z to replace my unpretentious Hantek DSO2C10. Model and Version are at the attached screenshot (pict.1)1757861-0. When I tried to calibrate the 10X probe, my Rigol showed a very strange effect (pict.2)1757867-1. The upper and lower shelves of the test meander differ significantly in thickness. I did factory reset, reset to default, auto-calibration, but nothing helped. To exclude the influence of the built-in test signal generator, I connected Rigol to an external sawtooth wave generator and saw that the beam thickness gradually changes from the top to the bottom of the screen (pict.3)1757873-2.
Then I disconnected probe and step by step made vertical offset from top to the bottom of the screen by using Vertical Position knob. The thickness of the beam and the noise changed from min 20mV at the top to max 180-200mV at the bottom of the screen (pict.4)1757879-3.
The conditions of the measurements were:
 - Nothing was connected to the inputs;
- The First channel was ON only;
- Channel menu: Scale – 500mV/Div; Coupling – AC; BW Limit – OFF; Probe – 10X; Invert – OFF; Volts/Div Coarse; Unit – [V].
- Horizontal: Scale – 100us/Div; Dalayed – OFF; Time Base – YT.
- Acquire menu: Mode – Normal; MemDepth – Auto; Anti-aliasing - OFF.

Then I zoomed beam that was at the bottom of the screen to 50mV/Div and 5ns/Div I saw that it's not a statistic noise but continuous wave with 125Mgz frequency (pict.5)1757885-4.
All of four channels have the same effect. But it appears if only one of any channel is enabled. Switching between AC -DC in the channels menu doesn't change anything. But if choose GND in the channel menu, the beam thickness and noise voltage becomes min (20mV) and not depend on the vertical offset. I guess that the problem is in the all four input amplifiers circuit of DS1054Z.

Is this a defect of my oscilloscope only, or are all DS1054Z like this? If anyone has the same DS1054Z, could you please check how it looks on your oscilloscope?
Thanks in advance,
Boris.

PS: Link to video, but the sound is in Russian language:  https://cloud.mail.ru/public/V893/WAqkPGPP1
PSS: Sorry for mistakes. English is not my native language))
« Last Edit: April 09, 2023, 09:23:56 pm by BorisSS »
 

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Hi,
Concerns the signal comparisons between rigol and hantek:
The rigol has much more memory depth and sample rate than the hantek, as can be seen in the pictures.
If I remember correctly, you can set the memory on the rigol to manual.
Set it to the smallest value, then the signal should look "thinner".

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Quote
Concerns the signal comparisons between rigol and hantek:
The rigol has much more memory depth and sample rate than the hantek, as can be seen in the pictures.
If I remember correctly, you can set the memory on the rigol to manual.
Set it to the smallest value, then the signal should look "thinner".

Thank you Martin72. You are exactly right. But the goal of my topic was not to compare Rigol vs Hantek)) I want to understand, is it specifically my Rigol that has such a defect, or does DS1000 model range have beam thickness and noise dependces on the  beam position on the screen?
 
The following users thanked this post: Martin72

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7463
  • Country: hr
Quote
Concerns the signal comparisons between rigol and hantek:
The rigol has much more memory depth and sample rate than the hantek, as can be seen in the pictures.
If I remember correctly, you can set the memory on the rigol to manual.
Set it to the smallest value, then the signal should look "thinner".

Thank you Martin72. You are exactly right. But the goal of my topic was not to compare Rigol vs Hantek)) I want to understand, is it specifically my Rigol that has such a defect, or does DS1000 model range have beam thickness and noise dependces on the  beam position on the screen?

No, there should not be difference in trace "thickness" with more negative value  (difference between trace at bottom and top of the screen). If you bought it as new and if there is any chance to return it for warranty repair/replacement I would try that first.

If not, then try a factory reset and self cal again.. Just in case..
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Quote
or does DS1000 model range have beam thickness and noise dependces on the  beam position on the screen?

Ah, I overlooked this.

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Quote
No, there should not be difference in trace "thickness" with more negative value  (difference between trace at bottom and top of the screen). If you bought it as new and if there is any chance to return it for warranty repair/replacement I would try that first.
If not, then try a factory reset and self cal again.. Just in case..

Thanks 2N3055! I bought it new and can get money back. But I like Rigol more than Hantek. Even despite this defect and the fact that Rigol is much more expensive)) Of course I will take this opportunity if the other owners of the DS1054Z confirm that this is a defect specifically of my device, and not of all model row of DS1000.
« Last Edit: April 09, 2023, 10:13:18 pm by BorisSS »
 

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
On work we have three ds1054Z, I can have a look on tuesday... ;)

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Quote
On work we have three ds1054Z, I can have a look on tuesday...
Thanks a lot Martin72! I appreciate your assistance very much!
Pease be sure to make the same settings as mine. Because you were exactly right. If to set less mem depth or less sample rate by adding another channels, the beam will be thinner (see attachments). May be the correct way is: to run oscilloscope, then to enable First channel only, then to push Storage button, then select Default, then set X10 and BW Limit to OFF in Channel menu, then switch to X10 on the probe, then connect First input to the built-in test signal generator, then press AUTO button.
PS: Please do not be offended by such detailed instructions, there are just some nuances that may affect the accuracy of measurements. 
 

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Quote
PS: Please do not be offended by such detailed instructions

Never, because:

A comparison only makes sense if you have or create equal conditions. ;)

Offline Someone

  • Super Contributor
  • ***
  • Posts: 5155
  • Country: au
    • send complaints here
Run a self-calibration, fixes many problems like this.
 

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Quote
Someone
Run a self-calibration, fixes many problems like this.
My thanks to your advise Someone. But I already wrote that I did it twice before I turned to you for help))
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 5155
  • Country: au
    • send complaints here
Quote
Someone
Run a self-calibration, fixes many problems like this.
My thanks to your advise Someone. But I already wrote that I did it twice before I turned to you for help))
You said lots of things, was the self calibration run after warm up? has the problem changed over temperature?
 

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Quote
You said lots of things, was the self calibration run after warm up? has the problem changed over temperature?
How long do I have to wait to be sure that the problem does not disappear after warming up? Twenty minutes or more have passed before factory reset, calibration and measurements. Is that enough or not?
 

Offline Miti

  • Super Contributor
  • ***
  • Posts: 1405
  • Country: ca
Yeap, three at work and one at home, they all do that. maybe someone smarter than I can explain this behavior.

Miti
Fear does not stop death, it stops life.
 

Offline noisyee

  • Contributor
  • Posts: 45
  • Country: cn
Then I zoomed beam that was at the bottom of the screen to 50mV/Div and 5ns/Div I saw that it's not a statistic noise but continuous wave with 125Mgz frequency (pict.5) (Attachment Link) .

Exact 125MHz indicates it may be the interleave spur of the ADC. The DS1000Z, if I remember correct, use HMCAD1511, which use 8 small ADCs internally interleaved to act as single 1GSPS, dual 500MSPS, or quad 250MSPS ADCs.
Maybe try different vertical scale to see if the noise is scale dependent.
 
The following users thanked this post: Someone, 2N3055

Offline Someone

  • Super Contributor
  • ***
  • Posts: 5155
  • Country: au
    • send complaints here
Yeap, three at work and one at home, they all do that. maybe someone smarter than I can explain this behavior.
As above reply by noisyee its pretty clear the noise is from poorly calibrated ADCs being interleaved. Does that model have a noise specification (or even typical values from the manufacturer) ?
 
The following users thanked this post: BorisSS

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Quote
Miti
Yeap, three at work and one at home, they all do that. maybe someone smarter than I can explain this behavior.
Thanks a lot! Are you sure that you see the same on your 4 Rigols and at the same conditions as me? Do you have DS1054z? Ask you because I found a picture here in this forum, there top and bottom shelfs of the internal calibration gen are equal by thickness (see attached pict).
« Last Edit: April 10, 2023, 02:22:07 am by BorisSS »
 

Offline Anthocyanina

  • Frequent Contributor
  • **
  • Posts: 395
  • Country: 00
  • The Sara
mine doesn't do that. I don't have a 20MHz square wave, but the ramp is consistent on the negative and positive side, as is the open input on all divisions, at around 100mvpp, and the square wave at 1khz also looks symmetrical. after some persuasion, my generator was gracious enough to give me something close to a 20MHz square wave, and it looks symmetrical as well.
« Last Edit: April 10, 2023, 03:18:33 am by Anthocyanina »
 
The following users thanked this post: BorisSS

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Anthocyanina thank you so much for the great work! Although this is bad news for me and I will have to return device to the seller((
I would like to clarify, did you definitely set Normal Mode in the Acquire menu? I get about the same results in High Res mode.
 

Offline Anthocyanina

  • Frequent Contributor
  • **
  • Posts: 395
  • Country: 00
  • The Sara
Anthocyanina thank you so much for the great work! Although this is bad news for me and I will have to return device to the seller((
I would like to clarify, did you definitely set Normal Mode in the Acquire menu? I get about the same results in High Res mode.
hi, yes, i always keep it normal mode, everything looks symmetric in any of the modes for me. i would say it really must be a defect, or, i don't know, but this thought popped in my mind, are you using different probes on the hantek and the rigol? maybe defective probe has some wridness when the signal goes negative? no idea what it could be if it's the probe, or if a probe can fail in that manner, just a thought. if not the probe, then for sure something wrong with the scope. good luck getting a replacement if so!  :)
 

Offline Miti

  • Super Contributor
  • ***
  • Posts: 1405
  • Country: ca
A don’t know exact conditions it does that, what I know is that the trace goes thinner and thicker as I move it up and down.

Edit: Attaching some screen shots.
« Last Edit: April 10, 2023, 11:53:57 am by Miti »
Fear does not stop death, it stops life.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Doesn't seem normal to me. Send it back.
 

Offline Njk

  • Frequent Contributor
  • **
  • Posts: 348
  • Country: ru
Did not notice the phenomenon with mine
 
The following users thanked this post: BorisSS

Offline Miti

  • Super Contributor
  • ***
  • Posts: 1405
  • Country: ca
And the triangle wave. I think yours is an extreme case of something they all do. I agree, you should return it.
Fear does not stop death, it stops life.
 
The following users thanked this post: BorisSS

Offline BorisSSTopic starter

  • Contributor
  • Posts: 12
  • Country: ru
Exact 125MHz indicates it may be the interleave spur of the ADC. The DS1000Z, if I remember correct, use HMCAD1511, which use 8 small ADCs internally interleaved to act as single 1GSPS, dual 500MSPS, or quad 250MSPS ADCs.
Maybe try different vertical scale to see if the noise is scale dependent.
When I set Coupling – GND in the channel menu, the beam becomes thin and does not depend on the position on the screen. So I assumed the problem is in the input amplifier. But may be you're right, not me. I don't have a schematic diagram so I don't know where the input closes to GND. May be at the ADC input.

Thanks again to everyone who is trying to help me figure out the problem. It is very important to set exactly my measurement conditions when you simulating my case on your devices. It is important that:
- Only one channel is involved.
- Acquire menu: Mode – Normal; MemDepth – Auto or 24.0M pts; Anti-aliasing - OFF.
- Channel menu: Scale – 500mV/Div; BW Limit – OFF; Probe – 10X.
- Slider on the probe must necessarily be set to 10X.
Some of you have attached screenshots in which the probe was set to 1X. But noise is lower in such case and test meander and ramp signals looks more symmetrical on my Rigol (see my screenshots). Please see how test meander will look if the probe is switched to 10X.
« Last Edit: April 10, 2023, 08:23:59 pm by BorisSS »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf