Dragon 64 reproduction motherboard "Issue 6"

Fairly straightforward assembly, most resistor and capacitor values are
annotated on the board.  I missed a couple, just check them in Kicad.

There are some decisions to make up front.

There are _four_ solder jumpers.  You must choose one connection on each.

LK1	Z7 select.  Was a jumper on the original board.  This isn't so relevant
	- it'll map two regions of 32K one way or the other.  Basically if
	there's a problem with one half of RAM, you use this ensure the lower
	32K is good.  In reality, it's a Dragon 64: if your RAM is part bad,
	you should probably replace it.

LK2	Video bias.  Was a jumper on the original board.  0V connects one side
	of the variable resistor to ground, 2V connects it to a pair of diodes
	(D21, D22), biasing by 2 voltage drops (2V-ish).  All my boards with
	this jumper seemed to have this set to 2V, so that's what I choose.
	However, I think I have one board at least that doesn't even have the
	diodes or jumper.

LK3	MS# select.  If IC13 is a 74LS32, you can connect this to FS#.  This is
	compatible with some lowercase boards that need the FS# signal, but use
	this pin instead.  If IC13 is a 74HCT32, you should connect this to +5V
	so that DA0 is never tristated (floating CMOS inputs are not a good
	idea)...

LK4	HSync select.  Connecting '64' replicates original Dragon 64 behaviour,
	where HSyncs in PAL padding lines don't make it to the PIA.  '32'
	routes the combined HSync there instead, giving Dragon 32 behaviour.
	Note that in a change from the original Dragon 64, the combined HSync
	is always routed to the SAM so that refresh occurs during the padding
	lines too.

Next, this reproduction supports using either a 14.31818MHz crystal (or
14.218MHz if you can find one: good luck) as the main clock source for the SAM
as per the original design, or a TTL oscillator.

If you choose a crystal, connect C65, C66 and R84 as annotated.  If you choose
a DIP8 TTL oscillator, populate R84 with a 100-200Ω resistor and bridge C65
(connecting OSCin on the SAM to ground).

C67 is annotated as "HARD WIRED" on the original schematic, which may actually
be "HAND WIRED".  Things seem to work without it, but it does tidy up the E
signal a little.

Be careful with the transistors.  Things got a bit confused between the pinouts
on substitute parts vs the originals, and things may not be as you expect.  In
each case pin 1 is towards the bottom of the board: TR1 is CBE, TR2 and TR3 are
both EBC, TR4 (PNP) is ECB.

The reset switch may still be a little low on the board.  It fits my case, but
it's a little off-centre through the hole.  You might be able to slightly tilt
it towards the top of the board to get it more central.