[Diy_efi] Re: diy_efi Digest, Vol 17, Issue 45

[Bill Washington]

Marcello,
    A DSO is a fundamentally different beast from an analog CRO.
    A DSO works by taking periodic instantaneous snapshots (Samples) of 
the signal and extrapolating between these samples.
The critical specification is the maximum number of samples per second - 
the more the better (but also the more expensive).
The theoretical maximum frequency it can display is half the maximum 
sample rate, therefore a DSO which has a max sample rate of 100Ms/s can 
display up to a 50 MHz PERIODIC signal.
When the signal is not periodic the timing of the captured waveform 
edges can be shifted by up to the maximum gap between samples, in the 
example above this would be (1/100M) s = 10ns.
Note also that the sample rate changes with the timebase setting 
selected - the DSO does not have unlimited memory, and can therefore 
store only a limited maximum number of samples - when the timebase gets 
slowed down the sample rate has to be reduced so that it does not run 
out of memory - therfore also resolution reduces.

    One further feature which you may find useful is "Glitch Capture" - 
this feature (if it is included on the specific DSO) may be turned ON or 
OFF manually - When it is (installed and) turned ON, the DSO has the 
capability of taking one or more additional samples when it detects a 
glitch occur within a pre-determined (by the user) window 
(amplitude,width).
    These Glitches would otherwise often be missed by the periodic 
sampling nature of the DSO.

    One other thing to be VERY wary of with a DSO - they can easily fool 
you with a seemingly correct looking wavefprm which has been sampled at 
too slow a timebase - IE too slow a sampling rate - when it gets one 
sample every 2, 3, 4, or more cycles of a periodic or aperiodic but 
pulsing waveform - In this case you may see an 'accurate-looking' 
representation, but it is not a true picture of the waveform you are 
trying to view - essential information has been lost and cannot be 
recovered!

    My recomendation would therefore be to go for a Sample/second rate 2 
or 3 times higher than you need (ie 4 to 6 times greater than the 
fastest signal you want to capture) with a Glitch capture facility ----- 
Hip Pocket permitting, of course!!!

Best Regards
Bill

>Message: 5
>Date: Fri, 25 Feb 2005 22:22:27 -0800 (PST)
>From: "Marcello A. Belloli" <mbelloli at speedymotorsports.com>
>Subject: [Diy_efi] Choice of labscope
>To: "diy efi" <diy_efi at diy-efi.org>
>Message-ID:
>	<1744.24.7.39.231.1109398947.squirrel at www.speedymotorsports.com>
>Content-Type: text/plain;charset=iso-8859-1
>
>Hello Everyone,
>     AFter my fun with the eprom switcher project, I've decide to get a
>new DSO.  I'm currently looking at a Tektronix 2232 DSO plus.  I have
>no idea what the plus is for.  But it is a 100mhz 2 channel DSO with
>RS232.  I asked the seller what the smallest time division was and he
>said .05 Micro Seconds.  My current unit has 5 micro seconds as its
>lowest setting.  And apparently cannot display the signals I was
>looking at its lowest time division.  Does everyone think this will
>be enough?  I'm looking for any comments.  I'm not sure how far I
>need to go to get it done.
>      My next project is using a pic18f542 at 40mhz.  Would this scope be
>able to look at the signals created by this without problem?
>      I've been thinking I should get a logic analyzer instead.  If it
>bases its change in logic state the same as my hardware, I should be
>able to see problems.  And they usually have many more I/O inputs
>than a DSO.  I could look at all 18 pins on an 18pin pic at the same
>time.  My big question is am I really going to need to do this, or
>I'm just looking at signals once in a while when something isn't
>going right?
>     What does everyone think?
>
>- Thanks,  Marcello
>
>  
>
>  
>
_______________________________________________
diy_efi mailing list
diy_efi at diy-efi.org
http://lists.diy-efi.org/mailman/listinfo/diy_efi