Cure RFI in your computer room

 Edited to 64 characters per line,

and reposted with the permission of the author. 

      

      

---------------------------------------

Date: Thu Mar 23 09:16:13 1989

From: JERRY MURPHY (aa300)

      TANDY SIG SYSOP, CLEVELAND FREE-NET

Subj: Cure RFI in your computer room

      

Here is an article I wrote, which was published in the March

1989 issue of Coco Clipboard Magazine.  Not everyone gets that

magazine, but here is that article on how I reduced Radio

Frequency Interference from my computers.  I hope you find it

of some use.

      

In my amateur radio room, more commonly - and appropriately -

called "the shack", there are several transmitters of radio

frequencies. Most of them were designed to do just that, and

they do it very well. But a few other pieces also generate

radio frequencies, and it's been driving me buggy; they're

supposed to be computers, not radio transmitters! The current

crop of computers includes 1 Model 16B, 1 Model 4, 1 Coco 2, a

pair of Coco3's, and a Model 100 laptop.  My Model 1 is

somewhere out in the garage, along with the other computers

from yesteryear.  I'm having to write this on the Model 4,

because the pair of Coco3's are busy operating my ham radio

station, and I'm not too thrilled with the word processor in

the 16B.  I keep the Coco2 here for the grandkids to play with

when they come over.  But I suspect Ted wanted me to write

about the Coco3's; I better get back to that subject...  

      

The radios here operate throughout the amateur radio spectrum,

from 3 MHz to 449 MHz, inclusive.  I relax by using good old

fashioned CW on HF, but also spend considerable time monitoring

UHF/VHF Packet and FM voice, and keep in touch with the Public

Service communications that I am a part of, using HF RTTY and

AMTOR.  I selected the Coco3 to handle the digital modes for

the unique and marvelous capabilities of OS-9, Level 2. One of

the Coco3's is equipped with a pair of 80-track floppies and a

40-Meg hard drive; the other has a pair of 40-track floppies

and a pair of hard drives, one 20-Meg and one 40-Meg.  Burke &

Burke interfaces are in both units, and both use WizPro

software for the terminal chores. Both generate a remarkable

amount of RFI (Radio Frequency Interference), as will *ANY*

device which depends on digital switching at such high

frequencies; the crystal in the Coco is 28+ MHz.  Harmonics and

subharmonics of this and other products are all over the

spectrum anywhere close to the Coco3. It's easy to find the

things which radiate by attaching a length of coaxial cable to

the antenna input on the HF receiver, and move the shorted end

of it close to various things in the shack, and at the same

time observing the S meter, and listening on a number of

frequencies of your choice.  I found some hot spots to be the

cables between the various peripherals.  The original TV I used

as a monitor was completely unreliable due to interference.

During disk I/O, the RFI was intolerable. Enough was enough; I

removed all the RFI, and now enjoy a peaceful and reliable

hobby in both areas of interest.  How, you ask?  Read on.

      

The TVI was easiest to cure.  I placed an aluminum cookie sheet

under the TV and above the Coco3, and then grounded it.  This

is called a Faraday shield, but removed only some of the

TVI.  Ferrite beads, size FB-8, mix 43, were then installed on

each of the AC leads connected to the transformer in the Coco3,

both 120 and 12 VAC, total of five beads. Goodbye to TVI.  I

have to confess I've since said goodbye to the TV as well;

CM-8's had to be procured to properly display the high class

graphics I have in my terminal programs!

      

Next was the attack on the RFI from the cables.  You probably

have ribbon cables like I used to have, going to the disk

drives, the printer, etc.  They are traditionally flat, with

parallel conductors.  Bad news! Change them to the kind of

cables which are either round and shielded (and grounded), or

get the cable with twisted pairs.  Individual pairs are

twisted, but cabled in a flat configuration, with a flat place

to attach connectors every couple of feet.  The Transpost-

ioning of the conductors reduces the induced currents in ad-

jacent wires, thereby reducing energy in them which might be

radiated.  Shielded cables are more expensive, and don't always

fit through the rectangular slots in our equipment, but they

are slightly more effective in reducing interference.

      

If you still have radiated energy causing problems, you can

once again turn to ferrite.  Both round and flat ferrite forms

are available to impede the energy floating on the outsides of

your cables, including shielded coaxial cables and telephone

wires.  Palomar Engineers markets a wide variety of ferrite

devices.  I bought a number of their split-core devices, and

applied them to nearly every cable in the shack, right at the

source of the energy.  The drive cables have a 2 1/2" flat

ferrite device right next to the connector at the controller

pack; there's another inside the drive cabinet.  There's

another pair of ferrite devices at each end of the printer

cables.  The cables to the monitors each have a pair of ferrite

devices at the ends.  Cables to the hard drives each have

ferrite devices.  All metal cabinets are bonded together and to

ground. AC line cords make several turns through a ferrite

toroid.  Don't forget the telephone modem and cables; they

might also be hot. 

      

If I listen very carefully, I can almost hear the birdies on

some of the frequencies where they used to be loud and wide.

      

If you have difficulty finding Palomar ferrite devices, or

similar, try my local source which is adequately stocked in

most sizes and mixes:  

      

North Olmsted Amateur Radio Depot

16460 Lorain Road

North Olmsted, OH 44070

Phone: (216) 777-9469

      

Tell Rick, K8SCI, I said "HI!".

      

73, de Jerry Murphy, K8YUW

      

Copyrighted 1989

---------------------------------------