15 Aug
2010
15 Aug
'10
1:43 a.m.
Hi all --
I finally have procured enough hardware to get my IMSAI 8080 running
again, albeit in a very barebones state. I currently have only 16K of
working (and compatible) memory and am trying to get an additional 32K
static RAM board working again (It's an IMS board, see:
http://www.s100computers.com/Hardware%20Folder/IMS/32K%20Static%20RAM/32K%2…)
Mine is missing a number of memory chips; these are 5257 4Kx1 (I believe
from my research the TMS4044 is equivalent). Can
anyone suggest a
reputable place to pick up half a dozen or so of these for a
reasonable
price? There are a few on ebay for $6 each, which seems a little steep
(but perhaps my expectations are out of order...)
I'm also looking for a manual for a PSS RAM65. I believe this is a 16K
DRAM board. It's the only S-100 board in my collection that I haven't
managed to find a manual or schematic for...
Once I get the memory situation resolved I hope to get a floppy disk
system working. I currently have a Tarbell 1011D in unknown condition
-- is this a decent card to work with (and can it be made to work with
5.25" drives?), or would I be better off finding something else?
Thanks as always,
Josh
15 Aug
15 Aug
3:18 a.m.
On 14 Aug 2010 at 18:43, Josh Dersch wrote:
Mine is missing a number of memory chips; these are
5257 4Kx1 (I
believe from my research the TMS4044 is equivalent). Can anyone
suggest a reputable place to pick up half a dozen or so of these for a
reasonable price? There are a few on ebay for $6 each, which seems a
little steep (but perhaps my expectations are out of order...)
I believe that the Intel 2147 is a pin-compatible alternative.
Unicorn has these for $3 and change the each. Perhaps eBay has them
cheaper...
--Chuck
3:56 a.m.
From: cclist at sydex.com
On 14 Aug 2010 at 18:43, Josh Dersch wrote:
The 2147s draw a lot more power than the 4044s
Dwight
Mine is missing a number of memory chips; these
are 5257 4Kx1 (I
believe from my research the TMS4044 is equivalent). Can anyone
suggest a reputable place to pick up half a dozen or so of these for a
reasonable price? There are a few on ebay for $6 each, which seems a
little steep (but perhaps my expectations are out of order...)
I believe that the Intel 2147 is a pin-compatible alternative.
Unicorn has these for $3 and change the each. Perhaps eBay has them
cheaper...
--Chuck
4:32 a.m.
On 14 Aug 2010 at 20:56, dwight elvey wrote:
The 2147s draw a lot more power than the 4044s
Well, there are lower consumption parts available; the Intel 2147L or
C2147AL; the AMD Am21L47. On the other hand, if you've got the
supply headroom, the 2147 is only 2 bucks and change at Arcade
Components.
But good luck finding the low-power variety.
4:47 a.m.
Chuck Guzis wrote:
On 14 Aug 2010 at 20:56, dwight elvey wrote:
The board I'm populating needs 64 chips, but I'm only missing six of
them. Is valid to assume that 2147s will play nicely with the existing
5257s? In that case the extra power consumption from 6 2147s is
probably not going to break the bank.
Thanks,
Josh
The 2147s draw a lot more power than the 4044s
Well, there are lower consumption parts available; the Intel 2147L or
C2147AL; the AMD Am21L47. On the other hand, if you've got the
supply headroom, the 2147 is only 2 bucks and change at Arcade
Components.
But good luck finding the low-power variety.
5:22 a.m.
On 14 Aug 2010 at 21:47, Josh Dersch wrote:
The board I'm populating needs 64 chips, but
I'm only missing six of
them. Is valid to assume that 2147s will play nicely with the
existing 5257s? In that case the extra power consumption from 6 2147s
is probably not going to break the bank.
I'd probably replace a row of 8, not just the 6. The 2147 is a lot
faster than the MM5257 (by about a factor of 5) which probably
accounts for the difference in power consumption. If you've got the
power, you should be fine.
--Chuck
5:12 p.m.
On Sat, Aug 14, 2010 at 10:22 PM, Chuck Guzis <cclist at sydex.com> wrote:
On 14 Aug 2010 at 21:47, Josh Dersch wrote:
One other thing that might help the situation is that the 2147 enters
an automatic power-down mode when the CE# input is at a high level.
For example the datasheet for the CY2147-35 says the max operating
current is 125ma while the max standby current is 25ma.
Do MM5257 parts enter a standby mode when not selected? I don't have
a datasheet handy. Which would draw more current total, two banks of
MM5257 parts, or two banks of CY2147 parts where at least one of the
two banks should be in standby mode?
-Glen
The board I'm populating needs 64 chips, but
I'm only missing six of
them. ?Is valid to assume that 2147s will play nicely with the
existing 5257s? ?In that case the extra power consumption from 6 2147s
is probably not going to break the bank.
I'd probably replace a row of 8, not just the 6. ?The 2147 is a lot
faster than the MM5257 (by about a factor of 5) which probably
accounts for the difference in power consumption. ?If you've got the
power, you should be fine.
--Chuck
7:02 p.m.
On 15 Aug 2010 at 10:12, Glen Slick wrote:
One other thing that might help the situation is that
the 2147 enters
an automatic power-down mode when the CE# input is at a high level.
For example the datasheet for the CY2147-35 says the max operating
current is 125ma while the max standby current is 25ma.
Do MM5257 parts enter a standby mode when not selected? I don't have
a datasheet handy. Which would draw more current total, two banks of
MM5257 parts, or two banks of CY2147 parts where at least one of the
two banks should be in standby mode?
If the 5257 parts are like the TMS4044 parts, standy is entered by
reducing Vcc to 2.4V. Otherwise, maximum operating power dissipation
is given as 303 mW. I somehow doubt that the memory board in
question has the smarts to drop Vcc.
There also exist CMOS versions of the same chip--the Motorola
MCM146504, the Harris HM6504 or the Intersil IM6504 is one such, with
a standby power of 10mW. These are the same speed (450nsec) as the
TMS4044s. So more grist for your mill...
--Chuck
8:58 p.m.
On 08/15/2010 03:02 PM, Chuck Guzis wrote:
On 15 Aug 2010 at 10:12, Glen Slick wrote:
Watch out,
Some parts like the 6504 are not fully static and the
logic is somewhat dynamic with the transition of CS
performing the precharge of the gating logic around a
static cell. That means you must cycle CS for every
address change withing a read/write or RMW cycle.
Allison
One other thing that might help the situation is
that the 2147 enters
an automatic power-down mode when the CE# input is at a high level.
For example the datasheet for the CY2147-35 says the max operating
current is 125ma while the max standby current is 25ma.
Do MM5257 parts enter a standby mode when not selected? I don't have
a datasheet handy. Which would draw more current total, two banks of
MM5257 parts, or two banks of CY2147 parts where at least one of the
two banks should be in standby mode?
If the 5257 parts are like the TMS4044 parts, standy is entered by
reducing Vcc to 2.4V. Otherwise, maximum operating power dissipation
is given as 303 mW. I somehow doubt that the memory board in
question has the smarts to drop Vcc.
There also exist CMOS versions of the same chip--the Motorola
MCM146504, the Harris HM6504 or the Intersil IM6504 is one such, with
a standby power of 10mW. These are the same speed (450nsec) as the
TMS4044s. So more grist for your mill...
--Chuck
11:43 p.m.
On 15 Aug 2010 at 16:58, allison wrote:
Watch out,
Some parts like the 6504 are not fully static and the
logic is somewhat dynamic with the transition of CS
performing the precharge of the gating logic around a
static cell. That means you must cycle CS for every
address change withing a read/write or RMW cycle.
I think you're saying that the address is latched on the edge of the
enable/chip select line. That's probably okay in this situation, I'm
guessing.
--Chuck
7:14 p.m.
Another option is the InMOS IMS1203 CMOS SRAM. Same pinout (it's a
JEDEC standard, so other examples probably abound), fast (45 nsec)
and low power (80/15 mA active/standby current).
Shame that you need the 4K parts. I have a whole potload of the 16K
20-pin parts.
FWIW,
Chuck
5250
days inactive
5250
days old
10 comments
5 participants
participants (5)
-
ajp166＠verizon.net -
cclist＠sydex.com -
derschjo＠mail.msu.edu -
dkelvey＠hotmail.com -
glen.slick＠gmail.com