Re: Removing lid shielding (Re: 8 MB GS RAM cards with FOCUS& Zip GS)

[Rubywand <rubywand@swbell.net>]

Supertimer writes ...
> 
> Rubywand <rubywand@swbell.net> wrote:
> 
> >Supertimer writes ...
> >>
> >> The metal shield...the thing about it is that when you push a card
> >> like the TransWarp GS to high speeds, the shield sometimes
> >> bounces RF radiation that destabilizes it, or so they say.
> >
> >     It was kind of you to not say who "they" are. Bouncing RF radiation
> >may well be the most 'creative' explanation yet concocted for a
> >glitching Transwarp.
> 
> It was in an issue of Genie Lamp.  I forgot which one...if you have
> them all archived, you can check.  It referred to the Zip GS, I think,
> but I figured it must also apply to the TWGS.  ;-)


     Logical. The problem is with the idea of RF radiation generated
inside the GS bouncing around and disrupting any GS circuit. One might
just as well dig up the old John Campbell article (in Analog?-- not
sure) and consider the impact of Eloptic Rays.

     Noise generated across skimpy power supply and Slot bus power lines
is, certainly, a problem. The undersized power leads act like resistors
across which noise voltages appear during switching transistions. The
noise spikes can propagate along power lines throughout the system and
cause false triggering, incorrect addressing, and errors in memory data
R/W.

     There is good reason to believe major noise spiking also damages
the power supply. How is unclear-- maybe a feedback control circuit is
damaged; maybe the spikes give false feedback and lead to failure of the
main oscillator circuit when it tries to deliver more current than it
safely can. For sure, there are piles of dead GS standard and "heavy
duty" power supplies all of which had outputs rated well above the
current demands which led to failure. 

     Meanwhile, our original (1986) standard GS power supply continues
to deliver the juice for a ROM-01 with 10MHz/64k ZipGSx accelerator, 4MB
mem Expansion, Stereo and Sound input boards, RamFAST interface, and
Focus Hard Card. The difference between our PS and the ones which
cratered is that, years ago, ours had the skimpy leads supplied by Apple
swapped out for fatter leads.

    (Just so that no one worries without reason, we are talking about
GS's which are fairly 'loaded' with power-sucking plug-in boards. A
system with no accelerator, PC-Transporter, or Second Sight board
probably is not especially loaded in terms of power requirements.)


     Anyway, if a user wants to cure noise spike problems, the way to do
it is to get rid of the 'resistors' across which the noise is
generated-- i.e. fatten the +5V, +12V, and GND power supply leads and
add a couple +5V and +12V Slot bus jumpers (like from the motherboard PS
connector to Slots 3 and 7). The mod is described in more detail in the
newsgroup FAQs on Ground. (See the F00xPOWER.TXT file in the Faqs folder
at  ftp://ground.isca.uiowa.edu/2/apple2/Faqs ).

     The Genie Lamp has published plenty of useful information. Like
every other computer information resource, it is bound to screw up from
time to time. The reason for slamming the bouncing radiation business is
that it is a false explanation which leads to a bogus cure. Removing the
lid shielding to eliminate noise glitches is roughly equivalent to
having your local witch cast a No Glitch spell.



Rubywand