[sdiy] OT: Somebody may know this...

[Magnus Danielson]

From: Gene Stopp <gene at ixiacom.com>
Subject: [sdiy] OT: Somebody may know this...
Date: Tue, 14 Sep 2004 11:32:04 -0700
Message-ID: <BCD0C6AAF20CEA41903A1284F549D6170B5318 at ixca-ex1.ixiacom.com>

Hi Gene,

I haven't the faintest if you already got a good reply for this (came back
from Nice just 15 min ago and had 947 emails waiting for me...) but I'll give
you my direct thought on this.

> Hello you group of multi-disciplinary gurus, I have a tech problem that is
> outside my realm of knowledge...
> 
> <Sorry for the non-DIY question, but it's music-gear-related :) >
> 
> I rearranged my setup and I put the TV set next to my sampler rack, which
> causes audio crap. My sampler rack has some Roland 760's in it with a
> mouse/video switcher. Composite video goes into a RatShack modulator which
> puts the composite up on channel 3, into the TV.
> 
> So I says to myself, let's get that TV out of there and join the 21st
> century with a flat panel (our house is gradually going flat panel). I went
> to Fry's and flat panels with multiple inputs are kind of pricey still, so I
> looked at the TV's. Got a good price on a return that had a connector
> problem (which I promptly fixed). Set it up, moved the coax over and tried
> it in TV mode. The picture is pretty fuzzy, the text graphics are doubled,
> kind of like ghosts that you'd see on an old TV.
> 
> Tried composite video instead - same thing. Moved composite cable directly
> to sampler (bypassed switch box), same thing. Tried S-Video directly to
> sampler, same thing. I guess there is some scan-rate thing going on. There's
> no adjustment for this on either end. So I guess I'm hosed, but it would be
> good to know what's going on.

Actually, I think you have a real low-tech problem. I think you have an
impedance-matching problem somewhere, such that you get double reflections in
both end of the cable. This cause a constant shadow, which is shifted to the
left of the real text. What happends is that as you go from high intensity to
low intensity, the falling slope is so sharp that this edge mirrors in the
impedance-missallignment into the modulator or TV and then travels back to the
source where it mirrors again and now follows along added on top of the lower
intensity signal. I would check if you really has a 75 Ohm coax in place.

Hint: By doing some minor research, measurement and math you can actually
figure out how long the piece of cable you have between the mirrors. You can do
this by finding out how long the active part of a TV line is (in PAL a line is
64 us and the active part is slightly shorter). By measuring how long the
active part is on your screen and how long the mirror effect is you can get the
ratio of the active part and thus the time for the reflection. Then assuming
that the speed of light in the coax is 0.67 * c we have that the single-fly
path is about 20 cm per nanosecond or double-fly path is about 10 cm per
nanosecond you can quickly figure out the cable-length and thus which cable is
in error.

This is nothing else but a poor mans Time Domain Reflectometer (TDR for short).
TDR is a really neat tool which complement network analyzers very well. Modern
TDRs have risetimes down to 7 ps, but you come a long way on 35 ps. Wisely used
will TDR/TDT be a real life-saver for tricky design-flaws.

I'm lacking a good TDR at home, but I have a decent at work.

Cheers,
Magnus