Re: Got too greedy... dead machine?

From: Michael Grove <mgrove_at_iwvisp.com>
Date: Mon, 12 Apr 1999 08:21:49 -0700

Oliver Skelton wrote:
>
> Hi international rescuers...
>
> I got greedy after fitting the Nova & decided to clock the DSP to 50MHz
> (one of my long list of pending aims).

> 1) 50MHz oscillator wired to the 2 free solder pads near the DSP. I
> didn't cut the track between DMA & DSP clocks under the main board (or
> even look at bottom of main board at all) as the track was already cut
> on top - I just had to remove the 40MHz wire from Nemesis (folded back &
> insulated).

I don't remember seeing a trace that could be cut from the
top for the DSP
clock. Remember the SDMA gets a clock from the 32Mhz
oscillator too, so the
clock cannot be completely disabled. There is a trace on the
solder side of
the motherboard that when cut, separates the DSP clock in
from the 32Mhz
oscillator. Recheck the cut trace on top to be sure the SDMA
clock in is
not open (this is a different point than where the buffer
mod wire goes, but
I am not sure of the pin# on the SDMA for the input). If in
doubt, repair the
cut on top, and do the modification from the bottom (cut
trace, it's easily
seen going to the DSP underside). I did this a while back
and my machine
acted like yours when I failed to supply the SDMA with the
32Mhz clock.
Or, repair the top trace to the DSP and re-test using the
32Mhz clock.
 

> If I installed the new oscillator correctly - the corner with the dot is
> the corner with the pin which is not connected to anything.

Yup

> 3) I looked under Falcon 4mb memory card to see what's there. (TOS ROM,
> of course)

:). What I thought once was cool on the Falcon, is now a
pain trying to
bring up the 060:).


> 6) One of the 3 wires to the LED on the front of the CLAB case came off
> the chip it was soldered to. This chip is at the right front of the
> machine as you look at it in normal use. It came from the second pin
> from the left of the row of pins closest to the rear. I soldered it back
> and taped the lead down to try and prevent it happening again.

I never traced that out, but it is soldered to a couple of
pins on the SCSI
controller.

>
> 7) Removed & looked under ab040 card - looking at the Nemesis buffer to
> see what I might do with it in future.

Make a Aluminum foil sandwich ( I used two Hustler playing
cards and some
foil) to make a shield between the A/B and the PALS.(Careful
no foil is
exposed to cause a short).
Wrap some foil around the A/B ribbon connector and insulate
it with some
heat shrink. Tie the foil to ground with a wire.

>
> Then I got an ear bashing for playing with my Falcon instead of doing my
> expenses... so I hurriedly put the Falc together.

When my girlfriend yells, I just shut down the soldering
iron, throw a towel
over the machine, and shut off the lights. It's not worth
rushing:).

> Too hurriedly. I accidentally plugged the AB040 card in too far forward
> so that each row of pins was shifted to the wrong side of the
> connectors.
> (Half the pins hanging in the air)

I have done this before too, but have not performed an
analysis of what pins
could be shorted, only sighed a sigh of relief when the FUJI
appears.

Invest at least in a multimeter. I sure makes checking all
the cut traces on a
Falcon much easier. A completely dead machine (most of the
time) is caused by
a clock not getting somewhere. Make sure all clocks are
present. A crude check
is to look for about ~3V from the clock to where ever it is
going. Pay careful
attention to the Nemesis clocks. If you have a bad
connection.......it don't work;).

Mike
Received on ma. april 12 1999 - 18:00:48 CEST