[time-nuts] another Ebay mixup, 5370
Henk ten Pierick
henk at deriesp.demon.nl
Thu Jun 14 16:13:19 EDT 2007
For jitter measurements the bandwidth, and thus the rise time, is not
important. Only the timing which is not affected by the rise time.
The filter has to settle enough so bandwidth must be high enough. A
low pass which increases the rise time is oke. For digital scopes,
and thus also for the Amhurst software, it is required to have enough
samples on the edge. Often this is done with a lowpass.
For jitter measurements, noise bandwidth is awful and should be as
low as possible.
For a spectrum analyzer noise density is important, not noise bandwidth.
On Jun 14, 2007, at 21:31, SAIDJACK at aol.com wrote:
> In a message dated 6/14/2007 11:34:04 Pacific Daylight Time,
> henk at deriesp.demon.nl writes:
>> If realy a fet probe is needed, than use a lowpass filter with
>> bandwidth after the probe. Most of the time the majority of the
>> jitter is still seen and the large noise bandwidth of the probe is
> Hi Henk,
> that's a good data point! I have the HP differential probe and will
> try it
> as well.
> BTW: I really like the 54701A probes for Spectrum Analyzer work,
> they are
> pretty flat all the way out to above 3GHz, and virtually
> If you use the special HP Spectrum Analyzer probe itself (822xx
> part number
> or similar I think) then you have to be extremely carefull not to
> touch 5V or
> 12V etc on your PCB, otherwise the probe will blow and it's over
> $1000 to
> repair I hear. Friends tell me that's why never to buy those on
> Ebay: they are
> likely all blown out.
> ************************************** See what's free at http://
> time-nuts mailing list
> time-nuts at febo.com
More information about the time-nuts