21 Mar
2000
21 Mar
'00
4:16 a.m.
OOPS! I misread the previous post . . . it does say 1k x 4 bits.
The only 22-pin 4K-bit sram I remember from those days is one from EMM-SEMI,
but the number escapes me. Nevertheless, the only SRAM I remember in a
22-pin package other than the EMM-SEMI parts, was the 256x4 2101. There
were plenty of non-multiplexed DRAMs in that package, but few SRAMs.
If you have 22-pin parts and the memory diagnostic is failing on a 4K-byte
boundary, it's likely the parts are something wierd. The conclusion that
Philip has reached here is quite plausible. I'd explore his proposed
solution before going any further.
MOS Technology did build a few oddball parts, but I don't remember seeing
any of their parts as being THAT odd. My MOS data books don't include any
memory parts, though they did build some ROMs at that time.
Dick
-----Original Message-----
From: Philip.Belben(a)powertech.co.uk <Philip.Belben(a)powertech.co.uk>
To: classiccmp(a)classiccmp.org <classiccmp(a)classiccmp.org>
Date: Tuesday, March 21, 2000 10:51 AM
Subject: Are you sure the 6550 is dead? (was: Mos Technology RAM wanted)
2114s.
so
> I think the basic number you're looking for is "2101" from the same 1K
> series as the famous "2102" which is a 16-pin 1kx1 with separate in and
out.
> The 2101's I have are not fast enough to meet
the 200ns spec. However,
not
> much of anything that was readily available at the
time the PET model
2001
> came out was that fast. Either it was quite a bit
faster, e.g. 2147,
2115,
> etc, or it was slower, e.g. 2114, 21L02, TMS4044
etc, which were
typically
> 450 ns at that point in time. Those 450 ns parts
worked handsomely with
the
1 MHz 6502.
Perhaps you'd be able to use a 2101.
I would have thought a 2101 was far too small. These are 4kbit chips.
I agree 200ns probably isn't essential - my PET of that date uses 450ns On the 1MHz 6502 you have 500ns between the two clock
edges that govern RAM
timing - one guaranteeing a valid address, the second latching in the data.
(FWIW there were FOUR motherboard designs for the early PETs, based on all
permutations of 2114 or 6550 RAM and 2316 or 6540 ROM. Mine is late for an
old
style PET, and has 2114 and 6540)
>>> > One of the MCS6550 RAMs has gone west. Does anyone have a spare, or
an
>> >
equivalent, for sale? It's a 22-pin 1024 x 4 200ns static RAM.
>>
>> I'll have to check the RAMs in the spare PET that I
>> keep in the garage. Can't remember whether they're SRAMs
>> or DRAMs in the bigger PETs.
>
>Only the oldest 2001-x PETs use those SRAMs. I guess it's time to build
>that upgrade board (a PAL, a pair of EPROMS, and some 6264s, a 62256, or
>some cast-off PC cache).
That's right. When they revised the ROM code, they revised the motherboard that all ROMs were 2332s (and the sockets would take
2732s) and all RAM was
dynamic.
The thing that bothers me is that it says 3071 bytes free. This is EXACTLY
the
number of bytes free you get on a 4K PET.
My advice - Identify the suspect pair of chips (remember these are 4-bit
wide
parts) - I have the circuit diagrams if you want - and
swap them with the
corresponding pair for the top 1k of RAM. You should then get the message
"6143
BYTES FREE" when you power up. (Or try swapping
them with the video RAM -
you'll soon see if the chip is dead!)
I suspect, however, you will still get 3071 BYTES FREE.
Let me explain. On these early PETS the top four address lines come from
the
processor and go straight into a 74154 decoder, AND
NOWHERE ELSE. This
outputs
sixteen block select lines - one for each 4k of memory.
Your PET is
failing
memory test at exactly the point where it passes into
the second 4k block
from
the first.
In short, I think a failure at the block boundary is too much of a
coincidence.
I'd trace out what this line is doing from the
74154 (there is only one,
and
it's 24 pin, so you can't miss it!) to the RAM.
Failing this, I'd suspect
something else in the chip select logic. (It could, of course be a failure
that
affects a whole chip, but I'd try the other things
after you've swapped RAM
chips if the BYTES FREE doesn't change)
As I said, I can lend you my copy of the circuit diagrams if you haven't
already
got one.
Philip.
**********************************************************************
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
the system manager.
This footnote also confirms that this email message has been swept
for the presence of computer viruses.
Power Technology Centre, Ratcliffe-on-Soar,
Nottingham, NG11 0EE, UK
Tel: +44 (0)115 936 2000
http://www.powertech.co.uk
**********************************************************************