A few times I've been asked about how patch-cable lengths ( of CV and/or Gate signals ) can affect signal strength.
I found a few interesting posts in the Yahoo Doepfer Usergroup that might make things a bit clear to you.
First of all;
Gate cables even longer than 10 meters usually are no problem.
Although a gate signal might get slightly weaker when you use extremely long cables, the (simple) gate signal will often stay strong enough to trigger your modules.
CV-Cables of this length may have a slight loss depending on the electrical characteristics of the input and the output.
There are ways to measure it, but it is very difficult to judge whether a not completely clean octave tracking is caused by long cables.
Real loss of signal quality starts with asymmetric audio cables at such a length, and you have always to keep in mind that electromagnetic and electrostatic influences (hum and sizzle) can affect longer cables more than shorter cables.
If you take a relatively standard low-cost coax cable of say 380pF/m and 128 Ohms/km this will not cause any noticeable loss in audio top-end from a 1k Ohm output impedance until you exceed 20m or so but you are more likely to get increased noise and interference.
From a CV point of view the cable resistance is a more important parameter (if your CV is controlling VCOs that is) but even 20m is only 2.6 Ohms so this can be ignored compared to the high input impedance of most VCO CV inputs; again long lengths are more prone to pickup so 50Hz mains can modulate the CV.
( you may also experience hum loops by just connecting gear together that is powered from different power outlets across the room ). " *
( * thanks to Tony Steventon from Synovatron for the 2nd half of this post )