8 Jan
2013
8 Jan
'13
5:43 p.m.
I had often run D.E. WREGE's version of Spacewar! on my PDP-12. But
when I tried to run it on my PDP-8/E with a VC8/E and DEC VT01 (Tek
611), I would get flashing artifacts on the screen in addition to the
normal display.
I initially thought I had a hardware problem - but I could only get my
VC8/E and VT01 combination to fail while running DEC's VC8/E diagnostics
by setting the timing switch on the VC8/E incorrectly.
When I initially looked over the VC8/E code in SPCWAR, it looked O.K.
and I thought the problem was likely in the clock or interrupt code.
After reviewing it extensively and not finding an issue, I tried
multiple hardware clocks (EP, EC) and various VC8/E settings without
success.
Finally, I revisited the VC8/E code in SPCWAR - and the problem turned
out to be a subtle bug in the VC8/E code.
According to DEC's VC8/E writeup in the LAB 8/E manual, a "wait for
done" should be given AFTER the last LOAD command (X or Y) is given and
BEFORE the intensify command. In other words the code should be:
TAD X /Get X coordinate
DILX /Send to VC8/E
TAD Y /Get Y coordinate
DILY /Send to VC8/E
DISD /Skip on done
JMP .-1 /Wait until done
DIXY /Display (Intensify) XY point
...
The original SPCWAR code it was:
Get X
DILX /Send to VC8/E
Get Y
DILY /Send to VC8/E
DIXY /Display (Intensify) XY point
DISD /Skip on done
JMP .-1 /Wait until done
Because of the above coding error, the VC8/E analog output would not
"settle" the correct amount of time (which for a VT01 is about 80
microseconds) because a DIXY command was given immediately after the
DILY. This error created artifacts on the screen because the intensify
command was in effect while the VT01 was still in the process of moving
it's "beam" to the correct location.
BTW: Slower displays, such as the VT01/Tek611/Tek613, do NOT need a
clock to slow down Spacewar to an appropriate speed.
If you would like a copy of the corrected SPCWAR.PA or a listing
(SPCWAR.LS), you can get them via anonymous FTP to bickleywest.com or
via your browser at ftp://bickleywest.com/pdp8_spcwar/
While I was at it, I also added an assembly option to start SPCWAR in
Spaceship mode (per the original Spacewar!) as opposed to UFO mode.
Cheers,
Lyle
--
Bickley Consulting West Inc.
http://bickleywest.com
"Black holes are where God is dividing by zero"
11 Jan
11 Jan
1:55 a.m.
Lyle Bickley wrote:
I had often run D.E. WREGE's version of Spacewar!
on my PDP-12. But
when I tried to run it on my PDP-8/E with a VC8/E and DEC VT01 (Tek
611), I would get flashing artifacts on the screen in addition to the
normal display.
[...]
BTW: Slower displays, such as the VT01/Tek611/Tek613, do NOT need a
clock to slow down Spacewar to an appropriate speed.
You've got me confused. I thought you said that the VT01/Tek611 was
where you had the problem.
I used to run it on a LAB-8/e using a VR14 monitor, and didn't observe
the problem, but I think the VR14 is pretty slow.
Eric
12:25 p.m.
On Thu, 10 Jan 2013 23:55:47 -0700
Eric Smith <eric at brouhaha.com> wrote:
Lyle Bickley wrote:
The problem was with incorrect coding of the VC8/E driver in Wrege's
Spacewar!
[Early on in my attempts to diagnose of the VT01 artifact problem, I
thought the issue was with hardware (VC8/E or VT01), but subsequently
found it to be software.]
I had often run D.E. WREGE's version of
Spacewar! on my PDP-12. But
when I tried to run it on my PDP-8/E with a VC8/E and DEC VT01 (Tek
611), I would get flashing artifacts on the screen in addition to
the normal display.
[...]
BTW: Slower displays, such as the VT01/Tek611/Tek613, do NOT need a
clock to slow down Spacewar to an appropriate speed.
You've got me confused. I thought you said that the VT01/Tek611 was
where you had the problem. I used to run it on a LAB-8/e using a VR14 monitor,
and didn't
observe the problem, but I think the VR14 is pretty slow.
The VR14 is much faster that the VT01 (which is probably why it
worked). Here are the VC8/E's DEC defined load delay specs
(us=microseconds):
Tek602: 6us
VR14: 21us
VR20: 16us red, 21us green
VT01/Tek611/Tek613: 100us
The above are the standard configurations available for the VC8/E by
changing switches and/or jumpers. Other delay time configurations are
possible by changing various capacitor/resistor combinations...
The incorrect code in Wrege's original driver intensified the beam
before the VT01's X/Y amps had settled, hence the artifacts. I fixed
that by re-writing the VC8/E driver to DEC's specs - so it will work
with any device supported by a correctly configured VC8/E.
configured
Cheers,
Lyle
--
Bickley Consulting West Inc.
http://bickleywest.com
"Black holes are where God is dividing by zero"
12 Jan
12 Jan
2:28 p.m.
On Thu, Jan 10, 2013 at 11:55:47PM -0700, Eric Smith wrote:
I used to run it on a LAB-8/e using a VR14 monitor, and didn't
observe the problem, but I think the VR14 is pretty slow.
This version online looks like it was fixed. Were you using it?
It stuck a clear flags 6050 in also.
http://www.chdickman.com/pdp8/VC8/SPCWR3_vc8i.PA
Their is one difference between them that may be wrong. It looks like
the new version may have an unwanted line wrap.
Above version diff output
< -6;-10; -2;-10; 0;-10; 2;-10; 4;-10; 5;-10; 7;-10; -5;-11; -3;-11; -1;-11
---
New version has the second digit of the -11 on the next line.
-6;-10; -2;-10; 0;-10; 2;-10; 4;-10; 5;-10; 7;-10;
-5;-11; -3;-11; -1;-1
1
4189
days inactive
4193
days old
3 comments
3 participants
participants (3)
-
djg＠pdp8online.com -
eric＠brouhaha.com -
lbickley＠bickleywest.com