Path: num2.nntp.dca.giganews.com!number1.nntp.dca.giganews.com!border2.nntp.ams.giganews.com!nntp.giganews.com!news.zanker.org!npeer.de.kpn-eurorings.net!newsfeed.tpinternet.pl!newsfeed.atman.pl!news.intercom.pl!f124.n480!f127.n480!f112.n480!f200.n2432!f605.n774!f500.n123!f514.n250!not-for-mail
Newsgroups: fido.science
Distribution: fido
From: DAVID WILLIAMS 
Date: Tue, 24 Aug 04 08:15:40 +0200
Subject: Pythagorean triples
Message-ID: 
Organization: The Bayman BBS,Toronto, (416)698-6573 - 1:250/514
 111
 300
Lines: 37
Xref: number1.nntp.dca.giganews.com fido.science:321

->  DW> I know that on the VIC 20, it was possible to POKE some addresses and 
->  DW> make fine adjustments to the *speed* of the clock. I think the clock 
->  DW> speed (interrupt frequency) was generated by dividing the frequency (1 
->  DW> MHz) of the crystal that clocked the CPU. 
-> interesting... 
 
-> No colour on the vic-20? you need a pretty good crystal to produce colour 
-> composite video (and therfore TV). 
 
-> Jasen 
  
Yes. There was colour. But the VIC had a separate modulator box that 
had to be inserted between the computer and a TV set, if the TV was to 
be used as a monitor. Alternatively, the computer could be connected 
directly to a monitor that had separate RGB inputs. So the computer 
didn't need to produce composite video. 
  
Incidentally, I made a slip in my last message. I said that the TI 
clock's speed could be adjusted to make it accurate to within a few 
seconds per year. I meant "per *day*". The low byte of the divisor was 
held in RAM, which could be POKEd. The divisor took the frequency down 
from (about) 1 MHz to 60 Hz, i.e. by about 17,000 to 1. So altering the 
divisor by 1 changed the TI speed by about 1 part in 17,000, which is 
about 5 seconds per day. By selecting the best value for the divisor, 
it was possible to make the clock accurate to within about 2.5 seconds 
per day. 
  
My heliostat program had a combined clock-adjustment feature. While the 
thing was running, I could hit the "+" or the "-" key, which would make 
the clock jump forward or backward by 1 minute. But the program kept a 
record of the number of days that had passed since the last adjustment 
was done, which allowed it to adjust the speed of the clock to make it 
as accurate as possible. It worked very well. When I first set the 
thing up, it would need a clock adjustment after a few days, But after 
that, it would keep time within a minute for a month or two. 
  
                          dow