[TriEmbed] Oscilloscope Question

Sat Apr 11 00:04:57 CDT 2015

Hi Dwight,
How much speed do you need.  :-)

The primary purpose of an oscilloscope is to capture / show a signal.  The scope must be fast enough to capture the highest frequency of interest in the signal.  Advanced triggering modes and deep buffers are great but they are lipstick on a pig unless the basic mechanism is up to the task.

If you are looking at a pure sinusoid, the needed -3dB Bandwidth is the frequency of the sinusoid, but most interesting signals consist of a fundamental and an assortment of harmonics.  The highest frequency harmonic of interest is the driver.  In analog applications, the fastest frequency is often implied by the application or governing body.  Digital applications are a bit different.  The key parameter is the clock edge's rise time (time required for the driver to charge the line from 10% of the rail to 90% of the rail).  Ideal clocks (zero rise time) have inifinite bandwidth, but they don't exist outside text books.  -3 dB Bandwidth of periodic pulse trains (clocks) is often estimated by 0.35 / rise time.
(- 3dB Bandwidth means a signal of that frequency is attenuated by -3dB (voltage is reduced by about 30%, power by 50%) before it is digitized.  Attenuation increases rapidly with frequency)

Slow Ex.
5 ns edge (10% to 90% rise time)Bandwidth = 0.35 / 5 ns = 70 MHz
Fast Ex.
200 ps edgeBandwidth = 0.35 / 200 ps = 1.75 GHz!(This scope costs more than my car)  :-)

There is a close corollary to Bandwidth which is Sample Rate.  It should be at least 2 times the Bandwidth to meet Nyquist but 4 x Bandwidth gives more comfortable margin.  Sample rate can be enhanced in post processing (after the AtoD converter) with repetitive signals so it is probably not going to be the limiting factor for you.  Bandwidth will trip you up first because fast probes, preamps and converters are expensive and can't be replaced after the fact with Digital magic.  This is really the heart of the instrument.

For further information, spend some time at tektronix.com or agilent.com.  National Instruments (ni.com) also has some good whitepapers with further information.
Good luck,Kevin SchilfDigital Telesis, Inc.

      From: Dwight Morgan <dwight.w.morgan at gmail.com>
 To: Triangle Embedded Computing Discussion <triembed at triembed.org> 
 Sent: Friday, April 10, 2015 9:56 PM
 Subject: [TriEmbed] Oscilloscope Question

I’m thinking of getting an oscilloscope or a pc-based device that does that function like the NI MYDAQ. I’m not an electrical engineer, I just want to be able to look at what kind of wave form I’m getting for the small circuits I’m using or planning to use with Arduino and Raspberry Pi. Something with good instructions, ease of use, and inexpensive would be nice.  Any suggestions/comments are appreciated.  Thanks!  Dwight Morgan
_______________________________________________
Triangle, NC Embedded Computing mailing list
TriEmbed at triembed.org
http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.triembed.org/pipermail/triembed_triembed.org/attachments/20150411/8b004b76/attachment.htm>