19 Jul
2004
19 Jul
'04
9:56 p.m.
On Mon, 2004-07-19 at 22:43, Tony Duell wrote:
LM1889 chip it was - I now get coloured crap on the screen :)
The only spare LM1889 I had was in the Dragon 32, so I desoldered that
one... breaking one of the pins off right at the chip package in the
process - grrr! Out with the scalpel to cut away enough of the IC
plastic to expose metal and I've got it working... but bugger!
Absolutely. I'm confident that I've sorted that one though, although
I'll double-check (and make sure that the solder attempt by the previous
owner hasn't shorted anything it shouldn't)
ta - will do. I've grabbed the 6847 datasheet, and I should have the
6809 one already. Spare 6809's I have. The 6883 is socketed on both
Dragon 32 and 64 so if that's dead I could do a swap. Don't think I have
a spare 6847 though, and the one in the D32 is soldered. Of course, it's
likely to be something simpler I expect...
cheers,
Jules
Finally found the five minutes to see if my Dragon 64 works. It doesn't.
I get a white screen with a black border (which seems healthy enough),
Unusual, I would have expected something in green (but see below!) It'll give some kind of (Microsoft) BASIC sign-on
message. I am not sure
what the exact wording is. It should be black characters on a green
background.
Interesting. When the random characters stay together in the middle of
the screen I get a couple of A's and a D in the right place that made me
wonder if it was trying to say "Dragon Data", and then presumably
waiting for a keypress before dropping to BASIC (and giving the same
sort of text as the Dragon 32 does). Maybe that's just coincidence,
though.
is screwed -
but why the random characters should jump positions
sometimes I don't know (unless the design of the reset circuitry is not
very good)
One obvious fault I found was that pin 6 of the 40 pin 6847 IC was
DD3 -- Data bus bit 3. The machine will behave oddly without that! The only 18 pin chip I cna think of in the Dragon is
the colour encoder,
presumably an LM1889 or something. The fact that it's not there could
explain the lack of colour on the screen!
Yep :) Now running with a 17.5 pin chip... :-)
None of the
DRAM is getting warm (suggesting failure), and the fact that
there's something approaching normality on the display suggest that the
power rails are OK and the CPU's at least operating.
The design is similar to the CoCo, and is close to the classic Motorola
application circuit for the 6809/6883/6847. I would start with data
sheets on those chips. Check the power lines (I think all you need is +5V
in the 64), check the reset pin (is it stuck active?), check for CPU
clocks (E and Q, sourced from the 6883 SAM), check for memory address
activitiy, and so on. There's a lot you can do without a schematic.