[TowerTalk] Serious RF Problem

[Jim Brown]

On Fri, 13 May 2005 05:26:44 +0000, VR2BrettGraham wrote:

>Good point - though for all the radios to have a pin 1 problem is a
>little odd (though if it is another device that's locking up, maybe it
>has a pin 1 problem itself).

Sadly, it is not odd -- the ham radio community, and the engineers who 
design for it, are mostly unaware of it. When Neil Muncy (ex-W3WJE) 
taught us about it in 1994, virtually ALL pro audio equipment was built 
with pin 1 problems, even the most expensive broadcast equipment and mix 
consoles. Today it's more like 50%. 

Today, virtually all computer audio gear has pin 1 problems at its 
inputs and outputs. AND so do the serial ports!  I am convinced that pin 
1 problems are the root cause of the vast majority of "RF in the shack" 
problems. 

I use Writelog to control and key my radios for contests, including my 
K2/100. The key line on the computer locked up at 15 watts on 160 and 80 
when I loaded an antenna as a long wire on those bands. That antenna 
runs within 2 feet of the computer and the transmitter, under a wood 
desk that the gear is sitting on, so it is REAL close. AND, because of 
the proximity and the feed point, there is a lot of RF current there, so 
the coupling mode is magnetic.

To fix it, I replaced the K2 serial cable (which was foil/drain-
shielded, untwisted wires) with shielded CAT5. I used one of the pairs 
in the CAT5 for each pair in the serial cable that the radio control 
used. So one pair for pin 2, with the return of the pair going to the 
DB9 shell; another pair for pin 3, with the return of the pair going to 
the DB9 shell; another pair for pin 4 (Writelog's key line) with the 
return at the computer going to the DB9 shell, and the return at the 
radio going to the key jack (which IS tied to the radio chassis). The 
fourth pair in the CAT5 goes to the pin (don't remember which) that 
Writlog uses for PTT. I can now run 1 kW into that antenna with no 
lockups. To use the connector shell, you must use connectors at both 
ends that have metal shells, and the equipment shells must be bonded to 
the chassis. 

And consider this. When you put ferrite beads on an interface cable, you 
are choking the RF current on that cable. The reason why that choke 
WORKS, if it does, is that it is reducing the current that excites the 
pin 1 problems. Unfortunately, if you have a lot of RF in the shack, 
it's tough to get a big enough choke. 

Another point. The reason I switched to CAT5 was to get twisted pairs 
with a high twist ratio. Cable shielding provides almost NO magnetic 
shielding (see Ott, for example), especially below tens of MHz. The best 
way to reduce magnetic coupling is by using twisted pairs for signal 
circuits, and the tighter the twist ratio, the better. CAT5 has the 
highest twist ratio I know of. 

BTW -- I did some tests, and below 14 MHz, that interface cable doesn't 
even need to be shielded -- IF it's wired as I described! At 20 meters 
and above, it made a difference. 

73,

Jim Brown  K9YC