Re: Tek 2445 for TDS 1002?


Jim Yanik wrote:
nico@xxxxxxxxxxx (Nico Coesel) wrote in
news:4458fdac.940802522@xxxxxxxxxxxxxxxxx:

"Phil Allison" <philallison@xxxxxxxxxx> wrote:


"Paul Mathews"


A lot of the opinions you'll read are based on experiences with
sampling rates of a few 10s of MHz, where aliasing can be a real
problem.


** Really ?

The TDS 1002 scope is virtually the same as the TDS 210 that I "road
tested".

Ailiasing is a major problem despite the advertised 1 GS/s spec.

The reason is simple and given in the operator's manual.

When set for 5 milliseconds per division, the actual sampling rate is
only 50kS/s.

So ailiasing starts at 25 kHz.

Big problem.

Yes indeed, it sounds plain stupid to build an oscilloscope that way.
Is Tektronix cutting corners? Nowadays you can expect a digital
oscilloscope to have enough memory to store at least 1M points so
lowering the sampling rate only occurs at very low time/div. settings
(which should change the anti-aliasing filter accordingly). IIRC the
54 series HP / Agilent oscilloscopes work that way.


I suspect that the HP/Agilent 54 series isn't priced as low as the TEK
TDS1002,nor are they in the same general class of instrument.

--
Jim Yanik
jyanik
at
kua.net

Actually, it has more to do with the number of pixels available in the
LCD than with the amount of available memory. I stand by my comments,
and I have access to a great variety of scope models from Tek, Agilent,
and LeCroy. I also have written many DSP applications. Most modern DSOs
save so much time, that it's no real inconvenience to double-check for
possible aliasing in new situations.
Paul Mathews

.

Relevant Pages

  • Re: Tek 2445 for TDS 1002?
    ... sampling rates of a few 10s of MHz, where aliasing can be a real ... Ailiasing is a major problem despite the advertised 1 GS/s spec. ... When set for 5 milliseconds per division, the actual sampling rate is only ... it sounds plain stupid to build an oscilloscope that way. ...
    (sci.electronics.design)
  • Re: Linux, PC-based, PCI or USB oscilloscopes?
    ... Especially for a beginner I recommend a analog oscilloscope, ... purposes on _slow_ signals. ... sampling rate of at least 2*n, which is described by the Nyquist ...
    (comp.os.linux.hardware)
  • Re: Tek 2445 for TDS 1002?
    ... Ailiasing is a major problem despite the advertised 1 GS/s spec. ... When set for 5 milliseconds per division, the actual sampling rate is ... it sounds plain stupid to build an oscilloscope that way. ... lowering the sampling rate only occurs at very low time/div. ...
    (sci.electronics.design)
  • Re: samplig rate conversion
    ... "twain" schrieb ... professionally & accurately (i.e. no aliasing etc.) ... Could anyone recommend me a tool for doing that (free ...
    (comp.dsp)
  • Re: AM demodulation-envelope detection method
    ... harmonis, in this case I thought the mathematical shown at the end i ... Here there is no aliasing. ... The sampling rate is more than 4X. ...
    (comp.dsp)