🚀 go-pugleaf

In article <1344@ewu.UUCP> mpierce@ewu.UUCP (Mathew W. Pierce) writes:
>I just Dloaded the 14MHz speedup file by Leslie Ayling(s?) and was wondering
>whether or not it will work for the A2000?  It is supposedly meant for the 
>A500, but she says that it will work on other machines.

I've seen this speedup "design" and I'd be surprised if it works on _any_
machine. [For anyone who hasn't seen the design, it runs the 28MHz clock
through a a D flip-flop configured as a divide-by-2 and runs the resulting
14MHz clock into the bent-up clock pin of the 68000 instead of 7MHz.
That's it.]

A few weeks ago Dave Haynie posted a comprehensive description of why this
has little chance of working, and it boiled down to:

	1) Any accellerator must access normal Amiga RAM, ROM, custom chips,
	   expansion devices, etc., with the same timing as a 7MHz 68000,
	   synchronized with the master 7M clock.

	2) The 8520 Peripher Interface Adapter chips use the M6800 (yes,
	   68-hundred) interface, which requires a 0.7MHz E clock, and
	   VMA and VPA signals synchronized to it.

There are other problems with this particular design that diminish what
little faith I have in it:

	1) It refers to the MC68000P12A as a 16-MHz part.  My Motorola
	   data books say this should be a 12.5 MHz part. (It has a
	   fair chance of running at 14.3MHz, if you're lucky.)

	2) The schematic diagram includes a switch on the clock signal
	   to switch between 14MHz and 7MHz.  Now assuming that the
	   switch is mounted so it sticks out of the case, what we have
	   here is a (relatively) long, unshielded wire carrying a
	   high-frequency signal.  We EE's have a technical term for
	   this -- it's called an antenna. :-)  The amount of RFI that
	   it must generate and/or pick up must be incredible.

	3) It mentions something about floppy drives stepping too fast,
	   so the author replaced his with one with that can handle a
	   5ms track-to-track step rate.  I thought that the normal Amiga
	   step rate was about 3.6 ms, and any 3.5" drive should be able
	   to handle 3ms.

>Isa this a smart thing to do even if you do it right?

As you might guess, I think that to "do it right" you should start with a
better design.

--
Ed Hanway --- uunet!sisd!jeh
Use other side for additional listings.  Some assembly required.
Any resemblance to real persons, living or dead, is purely coincidental.