No Script, No Fear, All Opinion
RSS icon Home icon
  • EEVblog #360 – Rigol DS2000 Oscilloscope Teardown

    Posted on September 26th, 2012 EEVblog 14 comments


    Forum Topic http://www.eevblog.com/forum/blog-specific/eevblog-360-rigol-ds2000-oscilloscope-teardown/
    Teardown Tuesday.
    What’s inside Rigol’s new DS2000 series entry level oscilloscope?

    HiRes Teardown photos HERE.

    Be Sociable, Share!
    • http://www.siliconchip.com.au Nicholas

      Looks very nice but they don’t seen to make any MSOs this nice (yet). Based on that cut-out and moulding you found, I assume it’s just a matter of time.

    • http://www.dgkelectronics.com Kalle Hyvönen

      That one obscured chip near the big Rigol branded chip looks like it has been “wiped” clean off the info with a laser? Or is it some kind of a substance?

      • http://www.eevblog.com EEVblog

        It appears to be some kind of substance added on top.

    • William Andrew

      Tried briefly looking for an ADC that would fit the big QFP Rigol part.

      Analog Devices doesn’t make ADCs faster than 1GSa off the shelf, and TI/National’s ADC08D1020 doesn’t really fit the pinout.

      The ADC08D1020 is also $300+ in 1k quantities over the counter, which is practically half the price of the (baseline) scope, so I’d say its a custom part, possibly not by a US manufacturer.

      As such I’d be interested to see the noise floor of this thing, they claim down to 500uV/div and 12 bit resolution (oversampled), but I found the DS1000 series to be pretty noisy. To fit such a high specced ADC in such a cheap scope they must have made some compromises.

      Would the DSP do the FFT? It could be done in an FPGA, but I don’t know if a Spartan 6 really has the performance? Otherwise there wouldn’t be much point using a fancy DSP if it was just doing UI and measurements. On the other hand, its placement far out of the way of the display FPGA and memory is a bit odd for that theory.

      The “tape” you discovered appeared to me like the face had been lasered off. Are you sure it was actually protruding from the surface?

      • http://www.eevblog.com EEVblog

        It certainly appeared to be “protruding” from the surface, and looked all the world like tape of some sort. On one ship it even seems to hanging over the edge of the chip.

    • Ronald Lijs

      Hi Dave,

      You would know about this already I guess, but nice Rigol have added your tear down to their official website of the product at the bottom!

      http://www.rigolna.com/products/digital-oscilloscopes/ds2000/

      Can’t wait for the review!!

    • Mooie

      If I’m not completely blind, the power supply seems to be an ALPS one. Or at least the design is ALPS, which might have been licensed to Rigol. In the video around 5:45-5:50ish, there’s a bit of text on the board saying something like;

      LPS_SMPS80_V01.02
      2012/01/16

      I’d guess there’s a hidden “A” under the gunk, but I can of course be wrong.

      • http://www.eevblog.com EEVblog

        It’s certainly possible, and quite likely.

    • branadic

      The pinout of the PGA looks like it could be the LMH6518 from National.

      Kind regards, branadic

    • branadic

      I’m now very sure that the rigol labeled PGA/VGA is a LMH6518.
      The input circuit is clearly a copy and paste version as show in the datasheet of the LMH6518. The JFET is a MMBFJ309 (6U), protected with BAV199 (JY) and followed by a npn bipolar transistor, very common and nothing very special.
      The LMH6518, including bandwith limitation to serveral selectable values (20/100/200/350/650/750MHz and full) is followed by a fixed gain differential amp to drive the ADC converters.
      If all the other parts don’t limit bandwidth (I can’t see any bandwidth limiting alaising filter between differential amp and ADC and the JFET is fast enough), the scope could be extended to 350MHz (-3dB corner frequency), only by selecting another bandwidth limitation inside the LMH6518.

      • http://www.eevblog.com EEVblog

        If this is the case, how does it get the 70MHz limit on the low end model?

    • Pingback: scope | I,ROBOT