Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

A5000 TRM's & Schematics.

42 views
Skip to first unread message

Charlie

unread,
Feb 9, 2010, 7:17:35 AM2/9/10
to
I wonder if any kind person here can point me to a (working) link for
these?

As usual I have an admittedly pointless project this time involving
the A5000. It would be eased greatly by not having to manually trace
tracks round the motherboard...

...Read on at your peril:

In a effort to see how far I can o/c my A5000 I've noticed a few
things:
-Replacing the 36mhz (Memory & VIDC) crystal with a 60mhz works -
surprisingly.
-Even the ROMs work reliably @ 75ns (using a slower crystal) - big
surprise.
-Sadly the VIDC can't cope with more than 42mhz, even with cooling -
ho, hum.
(As it's plugged into a TFT that's an issue)

So:
I could just open the VIDC link on the gen-lock header and fix the
VIDC @ 36mhz regardless but that opens up several other cans of
worms...
Or:
Find some way to split the memory clock from the fastest VIDC clock &
'there you have it'.

To me the two most obvious ways of doing this are:
-Cut the 'selection line' for 36mhz & feed that to a VIDC-Enhancer
with a 36mhz crystal via the gen-lock header - doable, transparent,
yet another VIDC enhancer to build.
or:
-The more sneaky method would be to keep the 36mhz crystal (or maybe a
40mhz one) for the sake of the VIDC and cut the memory clock track...
...A5000 memory as standard is clocked @ 36mhz/3=12mhz. So if I take
that divider out of the equation and replace it with a D-type flip-
flop the memory will run @ 36mhz/2=18mhz - all will be sweetness and
light.

This brings me back to my request - it all gets a bit tricky if I
don't know what tracks to cut.

My thanks in advance for any help / pearls of wisdom.

Charles

z0m8ied0g

unread,
Feb 9, 2010, 7:30:23 AM2/9/10
to
There are a few for the A5000 on Chris Why's site
http://acorn.chriswhy.co.uk/docs/Acorn/Manuals/Manuals.html#TRM

If not I may have them somewhere if I can dig them out.

Hope this helps.

Cheers

Stephen

Theo Markettos

unread,
Feb 9, 2010, 7:38:39 AM2/9/10
to
Charlie <charle...@googlemail.com> wrote:
> I wonder if any kind person here can point me to a (working) link for
> these?

Schematics:
http://acorn.chriswhy.co.uk/docs/Acorn/Manuals/Manuals.html#TRM

> As usual I have an admittedly pointless project this time involving
> the A5000. It would be eased greatly by not having to manually trace
> tracks round the motherboard...

> Or:
> Find some way to split the memory clock from the fastest VIDC clock &
> 'there you have it'.

Have you read:
http://www.chiark.greenend.org.uk/~theom/riscos/docs/a5kspeed.pdf

Theo

Charlie

unread,
Feb 9, 2010, 8:13:05 AM2/9/10
to
That was quick! Thanks very much chaps - TRM downloaded. :-)
(I wonder why I couldn't find it on Chris's Acorns? Ho, hum)

Yes, thanks - most useful document.
The tricky bit is my A5000's VIDC can't cope with the speed of crystal
that the memory can, hence my looking to either split the 'fast' VIDC
clock from the memory clock or change the memory divider.

Of course if I had any sense I'd do something more productive with my
spare time...

Charlie

unread,
Feb 9, 2010, 2:34:27 PM2/9/10
to
Ok, ready to begin:
Having had a look it seems splitting the 'FastVIDC' and MEMC clocks
will be very easy indeed - no need to faff with piggy-back
VIDCEnhancers or memory clock dividers. It should be a very simple job
to split either of the lines at a resistor, retaining the on-board
crystal for the other line and using a further crystal for the split
line. If using a Mk2 motherboard it will hopefully be possible to use
that 'spare' area for the new crystal - we'll see.
SMT resistor can stay in-situ:
-De-solder & turn through 90deg so only the end nearest the crystal is
disconnected - this will then leave a 'patch point' for the new
crystal while splitting the line.

In case any others are interested (and are even more incapable of
reading schematics than me), here's my notes on the subject:

Seperate MEMC1a and Fast VIDC lines
---------------------------------------------------------

X2(36mhz) To Memory and Fast VIDC.
Two lines:
To MEMC - Via R115 (68R)
To VIDC - Via R93 (22R)
-(X3) 24mhz - Via R119 (22R)
-(X1) 25.*mhz Via R126 (22R)

Options:
Memory line
-Turn R115, so cutting the memory clock.
-Plumb new crystal to memory clock via a (68R) resistor.
VIDC line
-Turn R93, so cutting the fast VIDC clock.
-Plumb new crystal to fast VIDC clock via a (22R) resistor.

Where to put the new crystal?
X1 unpopulated in Mk2 A5000, alternate crystal present - potential for
locating new crystal here?
Or
Piggyback crystals with seperate clock lines.
Or
Fully seperate sockets.


Maybe I'll get round to doing this tomorrow & let you all know how it
went.

Regards,

Charlie

unread,
Feb 10, 2010, 4:30:49 PM2/10/10
to
Just in case anyone's interested I worked out how the split the MEMC &
'Fast VIDC' clocks on the A5000.

No need for fancy piggy-back VIDC-Enhancers or clock dividers...

-Add a socket for the extra crystal
-Turn R115 through 90deg
-Solder a wire between the free'd end of R115 to the lower pad of R126

Here's the nub of it as performed on a Mk2 motherboard:
http://grope.thruhere.net/Qube/projects/images/CrystalA5000Mock.jpg

The same should be possible with a Mk1 motherboard by soldering a
socket on top of a PLASTIC crystal placed in X2...
...with all but the output legs soldered together a wire could then be
taken from the free leg on the socket to the free end of R115.

My thanks for the helpful replies that got me to this point - my A5000
now has a 52mhz crystal supplying memory while a 40mhz one supplies
the VIDC as that works better with my monitor.

Regards,
Charles.

0 new messages