8 Jun
2002
8 Jun
'02
4:20 p.m.
I'm stripping interesting parts off of these stat mux boards
that I recently got around the corner - lots of socketed 6502s
and 65c02s, 2532s and the like. Many of the boards have a 24-pin
DIP MC3242 which I can find little about - Google only gives me
one meaningful hit that it's a "Motorola Mux and Refresh". I'm
trying to determine if it's worth desoldering any. I'm currently
using a hot-air desolderer, so it's not a _lot_ of effort, but,
for example, I'm not bothering to pull the popcorn TTL parts,
just memory and "interesting" parts.
So... anyone have any information on the MC3242 and why I might want
to save a couple dozen of them?
-ethan
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
8 Jun
8 Jun
5:40 p.m.
According to the Motorola Microcomputer Data Library (1978):
"The MC3242A is an address multiplexer and refresh counter for 16-pin
16K dynamic RAMs that require a 128-cycle refresh... such as the
MCM4116."
Louis
On Sat, 8 Jun 2002 18:20:39 -0700 (PDT), Ethan Dicks wrote:
#So... anyone have any information on the MC3242 and why I might want
#to save a couple dozen of them?
5:45 p.m.
--- Louis Schulman <louiss(a)gate.net> wrote:
According to the Motorola Microcomputer Data Library
(1978):
"The MC3242A is an address multiplexer and refresh counter for 16-pin
16K dynamic RAMs that require a 128-cycle refresh... such as the
MCM4116."
Well that certainly makes sense. A bunch of these boards have 4116s.
Seeing as all the 4116-based-6502 designs I've seen are refreshed
through discrete TTL, I guess I don't need to save these. I still
might grab a few, just in case.
I have pulled a few 4116s - probably have to test them in a dynamic
PET or something similar.
-ethan
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
6:31 p.m.
It's a 7-bit mux and refresh counter, and, if you ask me, it's a waste of
silicon. I used lots of 'em over the years and it isn't really that valuable.
If you have an application (one or two Intel Multibus-1 DRAM boards used 'em).
The 3242 is just a 7-bit mux and a 7-bit counter, with controls. It's useful
only if you specifically need it. I prefer to use normal muxes when I'm
reworking old troubled DRAM circuits.
If you really want the details, perhaps I can scan 'em for you. The 3242,
BTW, is not a MOT original, being an Intel/MMI design.
Dick
----- Original Message -----
From: "Ethan Dicks" <erd_6502(a)yahoo.com>
To: <cctalk(a)classiccmp.org>
Sent: Saturday, June 08, 2002 7:20 PM
Subject: Anyone have a data sheet on the MC3242?
I'm stripping interesting parts off of these stat mux boards
that I recently got around the corner - lots of socketed 6502s
and 65c02s, 2532s and the like. Many of the boards have a 24-pin
DIP MC3242 which I can find little about - Google only gives me
one meaningful hit that it's a "Motorola Mux and Refresh". I'm
trying to determine if it's worth desoldering any. I'm currently
using a hot-air desolderer, so it's not a _lot_ of effort, but,
for example, I'm not bothering to pull the popcorn TTL parts,
just memory and "interesting" parts.
So... anyone have any information on the MC3242 and why I might want
to save a couple dozen of them?
-ethan
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
9 Jun
9 Jun
10:55 a.m.
Richard Erlacher wrote:
It's a 7-bit mux and refresh counter, and, if you ask me, it's a waste of
silicon. I used lots of 'em over the years and it isn't really that valuable.
If you have an application (one or two Intel Multibus-1 DRAM boards used 'em).
The 3242 is just a 7-bit mux and a 7-bit counter, with controls. It's useful
only if you specifically need it.
I.e. : save some. I would, since I have at least one
rare machine ( Wirth's Lilith ) that relies on them.
I would not like to have to come up with a replacement
circuit if one fails.....
Jos
1:47 p.m.
Jos Dreesen wrote:
I.e. : save some. I would, since I have at least one
rare machine ( Wirth's Lilith ) that relies on them.
^^^^^^^^^^^^^^^
I've heard of these machines, but never seen a picture of
one. Do you have a web site with photos, by any chance?
--
John Honniball
coredump(a)gifford.co.uk
9:58 a.m.
A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 1589 bytes
Desc: not available
Url :
http://www.classiccmp.org/pipermail/cctalk/attachments/20020609/3bbf6b02/at…
12:38 p.m.
I have a bunch of small boards that use this baby, but the fact that its
relatively slow and relies on relatively conservative timing makes it easy to
design an application badly and not at all easy to design one well with it.
It saves you having to provide two LS257's and an LS590. I dont' find that
much of a saving. Those are all three 16-pin parts and cost well less than
the 3242 did, regardless of whether you used Intel's or Motorola's. Sadly,
the thing didn't provide much to help you with the timing itself, but that was
pretty easy. There was a companion part, the MC3480, which was helpful with
timing, though, as I said, it helped you do it badly and just got in the way
of doing it well.
I've got the data somewhere, but in a databook and not in a loose sheet.
Dick
----- Original Message -----
From: "Tony Duell" <ard(a)p850ug1.demon.co.uk>
To: <cctalk(a)classiccmp.org>
Sent: Sunday, June 09, 2002 12:58 PM
Subject: Re: Anyone have a data sheet on the MC3242?
I'm stripping interesting parts off of these stat mux boards
that I recently got around the corner - lots of socketed 6502s
and 65c02s, 2532s and the like. Many of the boards have a 24-pin
DIP MC3242 which I can find little about - Google only gives me
Intel also made it (or at least they made something called a 3242 with
the same functions from what I can see).
It's basically the address path logic for 16K DRAMs (7 address lines to
the RAM). It's a 7-wide 3-input multiplexer and 7 bit counter (for the
refresh address). The 3 inputs to the mux are the high 7 address inputs,
the low 7 address inputs, and the output of the counter.
While it didn't make DRAM control easy, it reduced the number of packages
you needed.
The pinout is :
1: COunt/ (clock input to refresh counter)
2: Refresh Enable (High = address out from refresh counter, low = from
row/col mux)
3: Row Enable (High = row/column mux passes low 7 address lines, low = high
7 address lines)
4: N/C
5: A1 input
6: A8
7: A2
8: A9
9: A0
10: A7
11: O0/ (Address 0 Output, inverted)
12: O2/
13: O1/
14: Gnd
15: Zero Detect/ (low if (refresh counter == 0))
16: O5/
17: O4/
18: O3/
19: O6/
20: A10
21: A3
22: A11
23: A4
24: A12
25: A5
26: A13
27: A6
28: Vcc (+5V)
So... anyone have any information on the MC3242
and why I might want
to save a couple dozen of them?
Well, I wouldn't design using one now, but it was commonly used on old
DRAM boards (Acorn used it, I think, in the 'System' machines, I've seen
it on S100 boards, etc), and thus it might be worth saving a few.
-tony
12:38 p.m.
Tony Duell wrote:
Intel also made it (or at least they made something
called a 3242 with
the same functions from what I can see).
It's basically the address path logic for 16K DRAMs (7 address lines to
the RAM). It's a 7-wide 3-input multiplexer and 7 bit counter (for the
refresh address). The 3 inputs to the mux are the high 7 address inputs,
the low 7 address inputs, and the output of the counter.
Did anybody make something for 64K DRAMS as well.
Most 6502/6800 machines used video access for refresh and the Z80
had built in refresh so only the 8080 really needed such a chip.
Back then a 4k/8k/16k static memory board was a option too rather
than DRAM board. Sigh ... Remember when 4x1K 450 ns static ram
was the best thing since sliced bread?
--
Ben Franchuk - Dawn * 12/24 bit cpu *
www.jetnet.ab.ca/users/bfranchuk/index.html
1:55 p.m.
New subject: 74S409 (was Re: Anyone have a data sheet on the MC3242?)
--- Ben Franchuk <bfranchuk(a)jetnet.ab.ca> wrote:
Did anybody make something for 64K DRAMS as well.
We used the 74S409 for our 64K and 256K DRAMs on COMBOARDs. I have
enough to last several lifetimes, if anyone wants to play with one.
We learned the hard way that 33 Ohm resistors were required inline
with the CAS and RAS lines to minimize undershoot, but other than
that, they are rock-solid. A 74S409 went on every COMBOARD after
the COMBOARD-I (which had 32KW of 2114 SRAM). I think we typically
put 16-18 chips per board (the COMBOARD-II has parity) - it worked
out like this...
COMBOARD-II - 128KB (4164 x 18)
COMBOARD-Q - 512KB (41256 x 16)
COMBOARD-BI - 2048KB (44256 x 16)
The reason why I mentioned the number of chips is that the inline
resistors are "tuned" to the load of the number of DRAMs they have
to drive. We had a cancelled project with a 68000 + 68451 in a
desk-top box with up to 32 serial ports and 512K - 2048K of 4164s.
The number of populated banks affected the values of the resistors.
-ethan
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
8113
days inactive
8114
days old
9 comments
7 participants
participants (7)
-
ard＠p850ug1.demon.co.uk -
bfranchuk＠jetnet.ab.ca -
coredump＠gifford.co.uk -
edick＠idcomm.com -
erd_6502＠yahoo.com -
jos.mar＠bluewin.ch -
louiss＠gate.net