Author Topic: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown  (Read 93338 times)

0 Members and 1 Guest are viewing this topic.

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #75 on: November 17, 2012, 11:55:09 pm »
Quote
I wonder if there a digital low pass filter implemented in the fpga..  seems like that would be an easy thing (edit: easy for Rigol) to turn on/off in the firmware.

Could be a possible way, something like an FIR in particular a polyphase filter or IIR filter is typical.

But right now I'm a little bit confused. It seems, that my 70MHz device delivers the full 200MHz bandwidth and the real limiting factor is the delivered probe (1x 8MHz, 10x 350MHz) itself.
I today connected a 50MHz XO91 quartz oszillator via BNC connector. The signal I can see on the screen is not what I would expect. Instead of the fundamental 50MHz with trend to a triangle waveform, I can see a nice square-wave (see 50MHz Oscillator 1.png).
Probing the same source with the probe the signal looks different (see 50MHz Oscillator 2.png).
From this point of view it makes completly no sense to pay USD 1789.- for the DS2202 instead of USD 925.- for the DS2072 and the search for a bandwidth limiting circuit was wasted time.

branadic
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #76 on: November 18, 2012, 12:04:58 am »
But right now I'm a little bit confused. It seems, that my 70MHz device delivers the full 200MHz bandwidth and the real limiting factor is the delivered probe (1x 8MHz, 10x 350MHz) itself.

Isn't the LMH6518 programmed to at least limit the bandwidth to 100MHz in the 70MHz model?
 

Offline zibadun

  • Regular Contributor
  • *
  • Posts: 112
  • Country: us
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #77 on: November 18, 2012, 05:22:34 am »

But right now I'm a little bit confused. It seems, that my 70MHz device delivers the full 200MHz bandwidth and the real limiting factor is the delivered probe (1x 8MHz, 10x 350MHz) itself.


This is interesting.  I just did an FFT on 19 mhz square wave with a sharp rise and it shows harmonics well into 200 MHz...    This is with the stock probe at 10x.    So I'm not sure either where does the 70 MHz limit come into play.. :-//

« Last Edit: November 18, 2012, 05:24:35 am by zibadun »
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #78 on: November 18, 2012, 06:10:02 am »
Someone really needs to input a 200MHz sine wave into a 70MHz input and confirm the attenuation.
I can't believe the 70Mz or 100Mhz would be -3dB down at 200MHz  ???
I'm also assuming the 200MHz version would have some faster timebase settings too?

Dave.
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #79 on: November 18, 2012, 07:08:30 am »
Taken direct from the datasheet
DS2202: 2 ns/div to 1000 s/div
DS2102/DS2072: 5 ns/div to 1000 s/div
So ...  :-// Your mileage may vary
But 1000s is sheer performance ... you'll probably be sitting there like a corpse waiting for a refresh
« Last Edit: November 18, 2012, 07:10:52 am by T4P »
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #80 on: November 18, 2012, 08:01:59 am »
I'm also assuming the 200MHz version would have some faster timebase settings too?

I believe the 2ns timebase setting and the 100MHz BW limiter are the only two firmware differences between the DS2202 and the DS2102/2072.
« Last Edit: November 18, 2012, 08:03:53 am by marmad »
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #81 on: November 18, 2012, 12:04:20 pm »
I today supplied my 3-channel programmable PLL board based on the CDCE913 to the DS2072. Channel 1 is programmed for 100MHz and channel 2 for 200MHz.
Attached is the behavior of the scope today. I explicit say today because  I found  a curios behavior. I remember that I yesterday were able to choose 56Mpoints with one channel active for a short time and were able to sample 28ms with the full 2GS, what I found amazing. Today this is again limited to 14Mpoints for one channel active.
I also could have sweared that the frequency response yesterday was different than today. Seems that the devil is inside the scope.  >:D
Today I will modify one of my DIY 500MHz 10:1 active FET probes into an AC probe and follow the signals inside. Maybe I can find the devil.
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #82 on: November 18, 2012, 02:31:57 pm »
56Mpts at full speed?! Many people here thought it was a gimmick ... nope it's not. Seems like DDR2 was a good idea, instead of using slow SDRAM and then there's DDR3 ...
 

Online tom66

  • Super Contributor
  • ***
  • Posts: 6678
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #83 on: November 18, 2012, 03:48:33 pm »
Doesn't DDR2 have a 64-bit wide data bus? So they can pump data on each clock edge, 8 bytes at a time. Even using basic 125MHz DDR2 they'd reach the required capture rate. I think the limiting factor now is ADC cost and the FPGA.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #84 on: November 18, 2012, 04:12:14 pm »
I remember that I yesterday were able to choose 56Mpoints with one channel active for a short time and were able to sample 28ms with the full 2GS, what I found amazing. Today this is again limited to 14Mpoints for one channel active.

I have no problem selecting 56MPts with one channel at the moment - but I have noticed some quirkiness in the scope's behavior regarding sample depth choices when switching between the AUTO memory selection (or features which automatically cause  memory reconfiguring - such as Record Open - or - turning on and off channels) and manual selection. Sometimes the available memory choices seem 'stuck' based on the last group of known parameters - rather than the current. I haven't pinned this down yet to a demonstrable repeatable bug.
« Last Edit: November 18, 2012, 04:17:57 pm by marmad »
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #85 on: November 18, 2012, 04:22:47 pm »
Quote
56Mpts at full speed?! Many people here thought it was a gimmick ... nope it's not.

DS4000 and DS6000 series comes with 140Mpts in standard configuration, while DS2000 series has 14Mpts in standard configuration and 56Mpts optional.
Would be interesting to teardown the 4000 series.  The 2000 series is very nice, it's sad that this is just a 2 channel scope. The screen is also a perfect choice in size. To say it with the words of Dave: "I love it."
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #86 on: November 18, 2012, 04:34:40 pm »
To say it with the words of Dave: "I love it."

Me too. My biggest complaint with it is the idiotic bit of code which clears the waveform from the display prior to executing commands which resize/move it. I just can't figure out if this is really just a (bad) choice by Rigol programmers to do it this way (as Rigol claims) - or is somehow dictated by the hardware configuration.
 

Offline zibadun

  • Regular Contributor
  • *
  • Posts: 112
  • Country: us
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #87 on: November 18, 2012, 07:10:27 pm »
I have no problem selecting 56MPts with one channel at the moment

that option is due to expire on mine in 20 hours.  oh well...
 

Online tom66

  • Super Contributor
  • ***
  • Posts: 6678
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #88 on: November 18, 2012, 07:14:59 pm »
If you take out the clock battery, does it forget that it's due to expire?

Perhaps if one were to blob a little solder across the 32kHz xtal, the xtal would stop, and the clock wouldn't tick? You'd get inaccurate time/date elsewhere of course, but I don't think that's vitally important.

Dave, do you still have the DS2000 and the Marconi signal generator? Time to verify the actual bandwidth limits like you did with the Agilent, and found the 500MHz to have decent response up to 640MHz?
 

Offline Astroplio

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gr
    • Welcome to myrobots
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #89 on: November 18, 2012, 07:25:04 pm »
I remember that I yesterday were able to choose 56Mpoints with one channel active for a short time and were able to sample 28ms with the full 2GS, what I found amazing. Today this is again limited to 14Mpoints for one channel active.
I also could have sweared that the frequency response yesterday was different than today. Seems that the devil is inside the scope.  >:D

Did you run a self-cal in the meantime? Someone else reported on an other thread that if you do that the trial version for the options expire! :-(

 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #90 on: November 18, 2012, 07:39:40 pm »
Did you run a self-cal in the meantime? Someone else reported on an other thread that if you do that the trial version for the options expire! :-(
Good point - I didn't think to wonder if he'd done a self-cal since I had assumed he knew about the bug; it's mentioned in at least 3 separate threads here:

https://www.eevblog.com/forum/product-reviews-photos-and-discussion/first-impressions-and-review-of-the-rigol-ds2072-ds2000-series-dso/msg161118/#msg161118

https://www.eevblog.com/forum/blog-specific/eevblog-369-rigol-ds2000-oscilloscope-playing-around/msg155862/#msg155862

https://www.eevblog.com/forum/product-reviews-photos-and-discussion/rigol-ds-2072-noise-problem/msg141050/#msg141050
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #91 on: November 18, 2012, 09:46:45 pm »
Quote
Did you run a self-cal in the meantime?

Yes I did when the scopes received and after warm-up. The 56Mpoint option returns from time to time and disappears again.
From time to time I also have the problem, that after power on the whole screen is full of noisy signal. When I than change vertical division so that one of the relays of the input attenuator clicks everythings working fine upon here.
I wasn't able to measure the real bandwidth incircuit at all, also with my active probe modified to an 1x ac active probe.
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #92 on: November 18, 2012, 09:58:23 pm »
From time to time I also have the problem, that after power on the whole screen is full of noisy signal. When I than change vertical division so that one of the relays of the input attenuator clicks everythings working fine upon here.

I haven't seen my DS2072 ever do this - perhaps one of your relays (or something else) is slightly malfunctioning.

The 56Mpoint option returns from time to time and disappears again.

This sounds random - or is there a pattern?
« Last Edit: November 18, 2012, 10:00:32 pm by marmad »
 

Offline A Hellene

  • Frequent Contributor
  • **
  • Posts: 602
  • Country: gr
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #93 on: November 18, 2012, 10:02:54 pm »
Noisy signal? Ouch...


-George
Hi! This is George; and I am three and a half years old!
(This was one of my latest realisations, now in my early fifties!...)
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #94 on: November 18, 2012, 10:29:55 pm »
Quote
I haven't seen my DS2072 ever do this - perhaps one of your relays (or something else) is slightly malfunctioning.

I don't think it's the relay. The mailfunction is preindicated while booting, the buzzer makes no sound while booting in this case.
By the way, the software version is 00.00.01, what is your's?.
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #95 on: November 18, 2012, 10:39:54 pm »
By the way, the software version is 00.00.01, what is your's?.

Software version: 00.00.01.00.05
Hardware version:  1.0.1.0.0
FPGA version:
SPU 03.01.02
WPU 00.06.00
CCU 12.29.00
MCU 00.05

To get the full version information, go to the trigger menu and set Edge, then press F7-F6-F7-Utility button combination quickly. Then check additional info under System > System Info. To escape from this "special" mode, do again  F7-F6-F7-Utility while in the trigger menu.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #96 on: November 18, 2012, 10:58:31 pm »
Okay, interesting. I have the following:

Software version: 00.00.01.00.02
Hardware version:  1.1.0.0
FPGA version:
SPU 03.01.02
WPU 00.06.00
CCU 12.29.00
MCU 00.05

Seems your software is somewhat newer. Is there a way to get new software into the device and if so, where can I get the latest version?
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #97 on: November 18, 2012, 11:06:05 pm »
Seems your software is somewhat newer. Is there a way to get new software into the device and if so, where can I get the latest version?
You should definitely upgrade your software - there is a bug in 01.00.02 which can cause the scope to become bricked (useless) due to memory writing errors (read this message: https://www.eevblog.com/forum/blog-specific/eevblog-369-rigol-ds2000-oscilloscope-playing-around/msg153700/#msg153700)

You need to contact your dealer (or Rigol directly) for a copy of the newest firmware.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #98 on: November 19, 2012, 01:01:39 pm »
Thanks, I today received the latest firmware (00.00.01.00.05) and will update tonight.
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #360 - Rigol DS2000 Oscilloscope Teardown
« Reply #99 on: November 19, 2012, 01:03:27 pm »
I heard the new official firmware is another 3 weeks away.

Dave.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf