On May 17, 22:10, Louis Schulman wrote:
> OK, the power supply works! I have power on the motherboard. Now on
> to the next problem.
>
> When attached to a composite monitor, I get video, but it is a
> rolling/pulsing screen. Yes, I have the sync set to 60 Hz.
I've seen that happen on Mark 1's where the sync isn't well matched to the
monitor's needs and it couldn't lock. That seemd to happen with 50Hz
versions and certain monitors, due to the timing of the sync pulse relative
to the start of frame. Since this is a Mark 2 and I assume you're using it
at 60Hz, I doubt that's the problem. Does switching between the 50Hz and
60Hz settings make a difference? If so, there's probably something wrong
with the timing.
Looking at the schematic, VSYNC is generated by the counters and
multiplexers at 13A, 19B, 20B and 21B, and the 74LS74 flipflops at 18A and
18B. The 50Hz/60Hz switch controls the multiplexers at 13A and 19B, and
they choose the reload values for the counters at 20B and 21B. You should
see a high multiple (I'm too lazy to work out the frequency :-)) of 60Hz at
21B pin 15 and a lower multiple at 20B-15. That in turn feeds the LS74 at
18B, via the gate at 16A, and the VSYNC signal comes out of 18B-5. If you
have a scope, that will tell you if the signal looks about right, if not,
but you have a logic probe, at least you can look for 60Hz pulses.
If it's extreme, perhaps the vsync isn't getting to the video socket. It's
mixed by ORing it [1] with the HSYNC in the LS32 at 9B (pin 9 is VSYNC, 10
is HSYNC, 11 is composite). The composite sync is buffered by by an LS04
at 10H (in pin 10, out pin 11) and then via R59 is mixed with the video
data (via R58) to the base of the video transistor.
[1] which is what causes the problem for some 50HZ monitors; the HSYNC is
effectively lost during the VSYNC pulse, which is very close to the first
used scan line, so the top few lines of the screen tend to tear.
> So, can you supply a troubleshooting procedure to find where the
> problem lies? Again, keep in mind that this is a Sorcerer II (with
> apparently 32K in Rows 2 and 3, is this the proper configuration)?
Two rows is quite normal -- if the ICs are 4116 or equivalent you have 32K;
if they're 4104 or equivalent you have 8K. I think you mean rows B and C
(look for the numbers at the edge of the board). Physical row C is
actually "row 1" or bank 1 as far as the RAM decode is concerned, and row B
is bank 2. Row A is bank 3. The banks have to be filled in order 1,2,3
(rows C,B,A). So, yes, your setup is normal.
If you can read the screen display, BASIC will tell you how much RAM is
free after it initialises. If you don't have BASIC, the monitor ROM will
tell you the address of the top of RAM after it initialises.
--
Pete Peter Turnbull
Network Manager
University of York
Received on Fri May 18 2001 - 17:19:44 BST