[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

KScope parallel clock jitter

While we're on the subject:

There is a known jitter problem* with the parallel RP-125 outputs
on KScope. The problem varies between boards, temperature, and
video content and therefore may be difficult to track down. It
usually shows up as random "sparkles" in the video when a parallel
output is "serialized". Fortunately there is an easy fix: replace
the hybrid clock-doubler module, position UP0K0 on the output 
processor card (OTP), with a newer version, GVG part number
068605-00. The part is socketed.

We, AJA Video, also manufacture a miniature serializer with a
crystal jitter filter, model C10PS-CF, which corrects this problem
on KScope and other equipment. In the case of KScope it would be
better to address the problem at the source as above.

* Detail on KScope jitter problem
KScope actually has a very stable sync generator and there is no
jitter problem with the internal clocks. The output jitter is due
to a small clock doubler module on the output processor card which
generates a 27 Mhz clock from the local 13.5 Mhz clock. This module
exists only because, starting with version -03 of the OTP card, a
new RP-125 encoder chip was used. This chip has very particular
requirements for the phase relationship between the 13.5 and 27 Mhz
clocks - hence the doubler module. Unfortunately, the original
module was quite jittery - it met the RP-125 spec (3 ns!), but it
is unsuitable for "serialization". The newer version of the module,
pn 068605-00, corrects the problem (this module came out approx.
June '94). Earlier versions of the KScope OTP card do not have this
module or the problem, but because the -03 OTP card is when the
"background wash" feature started, most have the -03 or later.

John Abt
AJA Video