Re: Second Sight "jumping" picture

[rubywand@aol.com]

In article <32e6ab32.323013@PersonalNews.Germany.EU.net>,
sbehrens@arrowweb.com (S�nke Behrens) writes:

>
>I wonder whether anyone else has seen this behaviour. My SS (1.4BEEF,
>is there newer firmware, BTW?) will periodically shift the picture to
>the left by a few pixels, which then "come in" at the right side.
>After a while, the picture will jump back to normal, and after a
>further while, shift left again.
>
>This is not dependant on mode, it happens in 320 as well as 640, and I
>think also in text mode (at least I saw it in 40-column mode during
>RamFAST startup).
>
>I tried removing all cards, took the ZIP out, the RF, everything but
>memory expansion and a 3.5" floppy. Didn't change a bit ... well, the
>frequency of jumps/shifts got lower when I took the ZIP away I think,
>but I'm not sure.
>
>Maybe someone on this group has experienced this or a similar problem
>and can suggest a fix. Oh yeah, for what it's worth, this is a fairly
>old ROM01 in a PC tower case using the tower's power supply.
>
>

     When a display shifts back and forth a few pixels, the usual
explanation is that something is messing with the Horizontal Sync pulse
and/or with dot loading and/or with display memory loading.  (This is
speaking as an experimenter with TV Typewriters and assorted home-brew
video displays, _not_ as an SS owner.)

     On an old ROM-01, you are virtually certain to have notable noise on
the +5 and +12 lines. (The Slot plane traces are too narrow for the
currents '90's boards demand.) This could be causing one or more IC's to
miss-trigger and/or a register may be getting a wrong value. Since the
shift seems to stabilize for a time, it sounds like some register
controlling H Sync and/or Mem Load may be getting set to a wrong value
(e.g. one of the low-order bits is wrong).

     Adding #16 leads from the power connector +5 and +12 solder points to
the SS Slot +5 and +12 points will reduce noise. A couple .45uF 20V disc
caps (one each from the Slot +5 and +12 points to ground) will add
bypassing near the SS and further reduce noise. These steps may eliminate
the shifting.

     Something else to check are socketed IC's and a socketed clock
oscillator module-- are there any on the SS?-- which may be making a poor
connection at one or more pins. (Scootch-up and re-socket any socketed
IC's and modules.)

     Some plug-in cards do not run the connector 'fingers' all the way to
the bottom of the board. This is a special problem on the GS because the
Slot socket sprongers contact the board fingers fairly close to the bottom
edge of the card. Pulling and re-socketing the card may eliminate a poor
connection.

     If the above do not get the job done, you may have a defective
component.


Rubywand