Re: Z80 type differences?

["Tristan Mumford" <nintendologist@nospam.gmail.nospam.com>]

On Mon, 06 Nov 2006 12:41:24 -0800, Michael J. Mahon wrote:

> Michael Black wrote:
>> "Tristan Mumford" (nintendologist@nospam.gmail.nospam.com) writes:
>> 
>>>G'day.
>>>
>>>I have a question that I haven't been able to answer easily. The question
>>>is based on a problem I have.
>>>
>>>My poor Apple//e Platimum's Z80 card is dead. Trying to use it locks it
>>>instantly. This is due to a smeared cap that shorted a couple of lines. I
>>>replaced the offending capacitor. There was no damage to the apple.
>>>
>>>However the card is still dead. Thankfully it's the standard Z80 Soft
>>>card. And everything is standard and socketed.
>>>First step I'm going to replace all the glue logic chips. If that doesn't
>>>do it there's the Z80B.
>>>It's the only Z80B I have. I have a stack of Z80A MPUs. And there are
>>>Z80-CTC MPUs on eBay. What's the difference and which can be used with the
>>>least modification, if at all?
>> 
>> 
>> The "A" and the "B" should denote the speed the CPU can run at.  That was
>> the case with other 8bit CPUs, and I'm pretty certain that applies to
>> the Z80.  The reference I have handy came out too early to list such
>> details, though.
>> 
>> My guess would be the "B" can run faster than the "A", but that might
>> not be relevant in the Z80 card where the maximum may not be hit.
> 
> Not true.  As Knut points out, the B version is a 6MHz part,
> and has shorter setup and hold times.  It is these times that
> are critical for reliable operation.
> 
> There were a few different "Softcards", and the original one
> used a clever syncopated clock scheme that worked with the
> ][ and ][+, but was marginal for the slightly altered bus timing
> on the //e.

Spot on. I'm using one lifted from a ][ that was terminally ill.
I even saw a likeness of it on some apple related site. I believe it is
the original clone card.


> 
> The //e versions of the Softcard use faster parts to deal with
> the marginal timing (just as the //e uses a 2MHz 65(C)02 to deal
> with the shorter setup and hold times, even though it still runs
> at 1MHz.).

Interesting. It'll be even more upset if I settle on the //e as the target
for an accelerator card I'm planning. Don't get too excited Though. It's
just going to be a cached 65816 based board with extra ram and a microusb
style USB interface. Already have the buffers, CPLD(s)(to keep things
small), and the CY[something]11HC ICs, misc logic etc. Anyway, back to the
topic.

I remember reading about the clocking problems with the Z80 cards.
My tired brain remembered another detail. Attempting to use the card
causes a second thing. Apart from the locking up, the LED that signals
activation lights and stays lit.

It's doubtful its a damaged board. More likely silicon. I reseated the
ICs, checked that the capacitors were capacitating (albeit roughly),
cleaned the edge connector and gave a careful inspection for any physical
damage.
All came up fine.
WHen I get a tract of usable time I'll go through, pull ICs as suggested,
substitute with new ones, and see what happens.

Thankyou for your help everyone.