Article View: net.micro.atari
Article #1941Re: 1meg upgrade killer
From: info-atari
Date: Wed, 08 Jan 1986 19:08
Date: Wed, 08 Jan 1986 19:08
537 lines
26587 bytes
26587 bytes
------------------------ The "original" article ------------------------
>From drutx!ihnp4!mhuxn!mhuxr!ulysses!burl!clyde!bonnie!akgua!whuxlm!harpo!decvax!genrad!panda!talcott!harvard!seismo!lll-crg!gymble!umcp-cs!eneevax!ravi Mon Sep 9 18:14:26 1985
Relay-Version: version B 2.10.1 (Denver Mods 7/26/84) 6/24/83; site druhi.UUCP
Posting-Version: version B 2.10.1 6/24/83; site eneevax.UUCP
Path: druhi!drutx!ihnp4!mhuxn!mhuxr!ulysses!burl!clyde!bonnie!akgua!whuxlm!harpo!decvax!genrad!panda!talcott!harvard!seismo!lll-crg!gymble!umcp-cs!eneevax!ravi
From: ravi@eneevax.UUCP (Ravi Kulkarni)
Newsgroups: net.micro.atari
Subject: 1 Meg ST upgrade (long)
Message-ID: <360@eneevax.UUCP>
Date: Mon, 9-Sep-85 18:14:26 MDT
Article-I.D.: eneevax.360
Posted: Mon Sep 9 18:14:26 1985
Date-Received: Wed, 11-Sep-85 23:00:50 EDT
Distribution: net
Organization: U of Maryland, EE Dept., College Park, MD
Lines: 213
I am forwarding these instructions to the net as the original
author does not have access to usenet. So far he and I as far as
I know are the only ones who have upgraded to 1 meg. If you are
brave enough to try this upgrade you might want to let us know
if you succeeded or perish the thought failed.
-ravi
================================================================
Date: Wed, 28 Aug 85 20:53:05 pdt
From: Gert Slavenburg <gert@Pescadero>
Subject: manual to 520ST 1 Meg upgrade
To: info-st@eneevax
Status: RO
I had enough requests to justify sending this out to all of you at info-st.
Good luck with the modification. Cheers will be appreciated, as well as
examples of use of the full 1 Meg - it's lonely at the top.
Directions for expanding the Atari 520ST memory to 1 MByte
==========================================================
WARNING : This is a hardware modification that will void the warranty of
your 520ST. If you do not have the appropriate tools or experience you
have a substantial chance of ruining your 520ST. Proceed at your own risk !
This modification has been in my 520ST without any problems for 6 days now.
However, I have (of course) not checked with knowledgable sources at Atari
to verify if this modification endangers the long term machine reliability
and/or software compatibility (I suspect it may endanger their software
compatibility if enough of us do it !).
tools/components needed :
=========================
16 256k * 1 RAM chips, 150 nSec. access time type, e.g. NEC 41256C-15.
(available at e.g. Fry's Electronics, Sunnyvale, CA for $ 2.77 each)
A good quality, preferrably temperature controlled soldering iron, with
a miniature tip (tip should be narrow enough to avoid touching 2 I.C. pins
at the same time). E.g. Weller type soldering station.
Good quality resin core solder (thin).
Approximately 4 foot of wire-wrap wire and a good stripper for it.
(you will have to route 3 wire's over a sequence of I.C. pins. The
easiest way to do this is to have a stripper allowing you to shift the
isolation forward over the wire, solder the next point, measure new length,
shift over insulation etc. until the endpoint). The "No Nik" .014" (dark
green handle) wire wrap stripper is the best tool for this. Available
e.g. at Jensen tools, Phoenix, AZ (602) 968-6231 catalog no H4B305.
De-soldering wick and solder suction tool.
Philips type screwdriver (for opening your ST), tweezers, pliers etc.
A steady hand and self-confidence.
explanation of the modification :
=================================
(Please read the rest of this document before starting. It may save you
time and a 520ST)
The current memory inside the 520ST consists of 16 256k*1 RAM chips.
Address (A0..A8) lines are common to all those chips.
The WriteEnable line is also common to all chips.
Data (in and out) lines are (of course) individual. The RAS (Row-Address
Strobe) line is common to all chips. The 8 chips forming the high order byte
group have one common CAS line, and the 8 forming the low order byte group
have one common CAS line (CAS is used as enable for write operations,
suchthat WriteEnable can be common to both groups).
The high order group from MSB to LSB consists of U45,44,43,42,38,34,33,32.
The low order group of U30,29,28,25,24,18,17,16. Note that all chips are
adjacent, though the numbering has gaps. RAS0, CAS0H and CAS0L are supplied
from U15 pin 8,6 and 7 respectively. (The 0 indicates bank 0)
Bank 1 that you are going to build in will be "piggy-backed" on
top of the current chips, where all pins of the new chips EXCEPT RAS (pin 4)
and CAS (pin 15) are soldered to the old chips equivalent pin. Thus they will
end up sharing addresses, data, WriteEnable and power and ground with the
existing chips.
All RAS pins of the new chips are wired together and will be supplied with
the "RAS1" signal generated on pin 18 of U15 (the Memory Controller, marked
3H-2119CC or so).
The CAS pins of the 8 new high order byte chips (on top of U45..U32)
are wired together and supplied from the "CAS1H" signal generated on
pin 22 of U15.
Analogously, the CAS pins of the new U30 to U16 are wired together and
supplied with "CAS1L" from pin 21 of U15.
how to go about it :
====================
step 1 : Open up your 520ST, pull of the keyboard connector, and remove
the main circuit card from its top and bottom shielding.
Make sure to remember which screws go where and note the
keyboard connector orientation.
step 2 : Desolder all the capacitors adjacent to the existing RAM chips.
(DO NOT SKIP THIS STEP. You'll loose time if you do, and worse,
the modification will not be reliable since you can't solder
the pins obstructed by the capacitors reliably (if at all)).
To desolder them, I found it easiest to heat the island on the
non component side, and bend the wires straight. After doing that
for each capacitor, turn over to the component side and heat the
islands while pulling the capacitor out with tweezers.
step 3 : Open up the holes of all the desoldered capacitors, using a
combination of de-soldering wick and suction tool. Do this from
the non-component side. If certain holes are difficult to open
up, you may want to use a wood splinter. (push it through while
heating). Be carefull to remove all solder debris !
THE REASON for opening the holes NOW is that they will be less
accessible once you've done the other steps ! Patience is a virtue.
(NOTE : Step 2 and 3 are the only ones that may damage your ST PCboard.
Be sure not to use excessive force while pulling out the capacitors. If
you damage your PCboard anyway, cure the problem now and not later).
step 4 : In this step we will piggyback the new RAM's on top of the old ones.
Be sure to connect all pins except pin 4 (RAS) and 15 (CAS).
The best way to go about this is to do chip by chip. First bend
the pins of the new RAM's suchthat they are perpendicular to the
package (instead of having slightly spread "cowboy legs"). Use
pliers to bend pin 4 and 15 suchthat it comes out of the I.C. package
horizontal, and cut of the excess length of pins 4 and 15 (I mean
part of the pin, you still need to be able to solder to it!). Make
sure that the new RAM fits snugly on top of the old one (in the
same orientation !!!!), without intervening space and with the new
pins touching the old ones. Now solder
each pin (except the non-touching 4 and 15) to the other RAM's.
The best way to do this with least chance of damage is to touch
both the new RAM's pin and the old RAM's pin. Heat them both for
a second and add A LITTLE solder then. Wait till the solder flows.
After each IC, check all pins carefully to assure a good connection.
(use a magnifying glass)
NOTE : This step is crucial for the long term reliability of the
memory extension. A badly soldered joint may show up later as
sporadic memory errors. TAKE YOUR TIME.
(NOTE : until step 6 is finished, do not in any way apply power to your ST.
This intermediate state of affairs will damage your memory chips !)
step 5 : Re- mount all the desoldered capacitors. Bend the pins like they
were before re-soldering, suchthat they will not touch the lower
shielding. Solder from the non-component side.
step 6 : In this step you will route the 3 wires mentioned earlier. The
first wire connects pin 4 (RAS) of all new RAM's to pin 18 of U15.
The second wire connects pin 15(CAS) of the new U45 to U32 to pin
22 (CAS1H) of U15. The third wire connects pin 15 (CAS) of the
new U30 to U16 to pin 21 of U15.
The best way to do this is to use the stripper to remove 5 inches
of isolation. Solder the first IC pin to the end of the blank wire,
measure the distance to the next pin in sequence and shift over
that amount of isolation. Continue in this fashion untill all pins
in the sequence are done.
Work from U45 to the left, soldering directly to the leftover pins
on the new chips. Make sure that no wire or solder sticks out
above the top plane of the new chips, since they will almost
touch the top shielding !
Route the wires through the PC board hole below and to the left
of U15 to connect to U15 on the non-component side.
step 7 : Sit back. Use brain. Do you feel confident about the quality of
your work ? No mistakes ? Check everything once again if you are
but a little uncertain. Applying power with errors might make your
ST into a decorative, non-functional piece of art.
OK. Either rebuild your ST into its shielding and cabinet, or put
it onto a surface clear of wires and solder-remains and connect it
to monitor, disk and supply. Boot it.
If it boots, you're probably there. Test if the new memory works
by looking at the phystop variable ($42E) with SID if you have the
developper stuff. It should read $100000 (1M hex). Also note that
memcntlr ($424) now holds 5 instead of 4, and that v_bas_ad ($44E)
now holds $F80000 (screen bitmap origin).
If you don't have the developper stuff, try a single drive copy and
check that you get the whole disk in one buffer instead of two.
If the new memory does not seem to exist, use SID to deposit and
retrieve words on locations $80000 and up (1/2 M hex). If bit errors
occur, the ST bootROM did not detect the extension (it checks all
bits of 512 locations by testing a pseudo random sequence, before
accepting a memory bank). Try to pinpoint the faulty chip(s) and
remove the error.
If it doesn't boot, you're in trouble. I'm sorry. It is difficult
to give hints on what to do here. So many possibilities.
Desoldering the new chips probably won't work (if the old ones
were functional, the ST would still boot). Check for hidden
short-circuit on RAM pins. May also be that you have a flaky
new pin connection.
Good luck. I feel bad. Well, I might be able to help you a little,
call me when you're really stuck ! Evenings at (415) 965-7696.
Not after september - I'll be out of USA by then.
PLEASE :
========
Let me know if you did the modification and failed or succeeded. I'd like
to know how many people benefit from this and if it's safe enough for large
scale application. Mail to arpanet address gert@su-pescadero.
====================================================================
--
ARPA: eneevax!ravi@maryland
UUCP: [seismo,allegra]!umcp-cs!eneevax!ravi
--------------------------- The "second" article --------------------
> ~~~~~~~~~~~~~~~~~~~pop goes the ST~~~~~~~~~~~~~~~~~~~
>
> i just fried my ST and after talking to atari and my local dealer i
> have tracked down the problem. It is related to the 1 Meg upgrade
> and the new proms, it seems that the early postings of "how to
> upgrade ...." left out 4 critical resisters (60 ohm 10% tol); 2 must
> be placed on the CAS lines and 2 on the RAS lines, all 4 go on the
> MMU (i know it isnt really an MMU) side. the upgrade will work fine
> w/o the resistors until you put in the new proms, the difference in
> the current drain will cost you all your memory chips and possibly
> the MMU.
> --
> god bless Lily St. Cyr
> -Rocky Horror Picture Show
>
> Name: James Turner
> Mail: Imagen Corp. 2650 San Tomas Expressway, P.O. Box 58101
> Santa Clara, CA 95052-9400
> AT&T: (408) 986-9400
> UUCP: ...{decvax,ucbvax}!decwrl!imagen!turner
After reading the above article, I downloaded the revised procedure
from Compu Serve. I have had the upgrade for about 3 months, and has
worked without a flaw. I would have probably added the proms without
the resistors. Thanks to Mr. Turner for the warning.
Here is a copy of the revised procedure:
NOTE: This is an REVISED,TESTED version of the original text
downloaded from CompuServe. December 6, 1985
(This was REVISED AND TESTED by an annonymous engineer on
Atari's developement staff. The addition of the resistors
should provide a long life to your machine, but the warning
below is STILL IN EFFECT. This is not an official sanction
of the modification. USE WITH CARE!!!)
Here's the 1 Meg upgrade directions:
I have brought this over un-editted from the arpanet info-st
mailing list. I TAKE NO RESPONSIBILITY FOR ITS CONTENT OR
ACCURACY. I HAVE NOT TRIED THIS MODIFICATION ON MY OWN ST AS
YET. I AM PASSING THIS ALONG TO THOSE WHO DO WISH TO TRY IT.
FOLLOW THE DIRECTIONS AT YOUR OWN RISK.
--Dwight McKay (75776,1521)
From: gert@pescadero
WARNING: This is a hardware modification that will void the
warranty of your 520ST. If you do not have the appropriate
tools or experience you have a substantial chance of ruining
your 520ST. Proceed at your own risk! This modification has
been in my 520ST without any problems for 6 days now.
However, I have (of course) not checked with knowledgable
sources at Atari to verify if this modification endangers
the long term machine reliability and/or software
compatibility (I suspect it may endanger their software
compatibility if enough of us do it!)
Tools & components needed :
16 256k * 1 RAM chips, 150 ns access time type, e. g. NEC
41256C-15 (avilable at e. g. Fry's Electronics, Sunnyvale,
CA for $2.77 each)
A good quality, preferrably temperature controlled soldering
iron, with a minature tip (tip should be narrow enough to
avoid touching 2 I. C. pins at the same time). E. g. Weller
type soldering station.
Good quality resin core solder (thin).
Approximately 4 foot of #24 AWG insulated wire and a good
stripper for it and 2 feet of #22 AWG solid tinned copper
bus wire. You will have to route 3 wires over a sequence if
I.C. pins.
Desoldering wick and solder suction tool.
Philips type screwdriver (for opening your ST), tweezers,
pliers, etc.
A steady hand and self-confidence.
Explaination of the modification :
(Please read the rest of this document before starting. It
may save you time and an 520ST)
The current memory inside the 530ST consists of 16 256K*1
RAM chips. Address (A0..A8) lines are common to all those
chips. The WriteEnable line is also common to all chips.
Data (in and out) lines are of course individual. The RAS
(row-address strobe) line is common to all chips. The 8 chis
foring the high order byte group have one common CAS line,
and the 8 forming the low order byte group have one common
CAS line (CAS is used as enable for write operations, such
that WriteEnable can be common to both groups). The high
order group from MSB to LSB consists of U45, 44, 43, 42, 38,
34, 33, 32. The low order group of U30, 29, 28, 25, 24, 28,
27, 26. Note that all chips are adjacent, though the
numbering has gaps. RAS0, CAS0H, and CAS0L are supplied from
U1 pin 8,6 and 7 respectively (The 0 indicates bank 0)
Bank 1 that you are going to build in will be "piggy-backed"
on top of the current chips, where all pins of the new chips
EXCEPT RAS (pin 4) and CAS (pin 15) are soldered to the old
chips equivalent pins. Thus they will end up sharing
addresses, data, WriteEnable and power and ground with the
existing chips.
All RAS pis of the new chips are wired together and will be
supplied with the "RAS1" signal generated on pin 18 of U15
(the memory controller, marked 3H-2119C or so). The CAS pins
of the 8 new high order byte chips (on top of U45..U32) are
wired together and supplied from the "CAS1H" signal
generated on pin 22 of U15. Analogously, the CAS pins of the
new U30 to U16 are wired together and supplied with "CAS1L"
from pin 21 of U15.
How to go about it:
Step 1: Open up your 520ST, pull off the keyboard connector
and remove the main circuit card from its top and bottom
shielding. Make sure to remember which screws go where and
note the keyboard connector orientation.
Step 2: Desolder all of the capacitors adjacent to the
existing RAM chips. (DO NOT SKIP THIS STEP. You'll lose time
if you do, and worse, the modification will no be reliable
since you can't solder pins obstructed by the capacitors
reliably (if at all)). To desolder them, I found it easiest
to heat the island on the non component side, and bend the
wires straight. After doing that or each capacitor, turn
over to the component side and heat the islands wile pulling
the capacitor out with the tweezers.
Step 3: Open up the holes of all the desoldered capacitors,
using a combiation of de-soldering wick and suction tool. Do
this from the non component side. If certain holes are
difficult to open up, you may want to use a wood splinter.
(push it through while heating). Be carefull to remove all
solder debris!! THE REASON for opening the holes NOW is that
they will be less accessible once you've done the other
steps! Patience is a virtue.
(NOTE: Step 2 & 3 are the only ones that may damage your ST
PC board. Be sure not to use excessive force while pullling
out the capacitors. If you damage your PC board anyway, cure
the problem now and not later).
Step 4: In this step we will piggyback the new RAM's on top
of the old oes. Be sure to connect all pins except pin 4
(RAS) and 15 (CAS). The best way to go about this is to do
chip by chip. First, bend the pins of the new RAM's suchthat
hey are perpendicular to the package (instead of having
slightly spread "cowboy legs"). Use pliers to bend pin 4 and
15 such that the legs are 180 degrees from their normal
position, so they stick up in the air above the plane of the
top surface of the chips. Don't make an absolute sharp 180
degree bend since some manufacturers' pins may snap off.
Leave a little curve in the leg, but insure that is above
the plane of the top surface of the chip.
Using #22 AWG to #16 AWG tinned solid copper wire you will
form three buses along the top surface of the new d-rams.
Cut a #22 AWG solid copper wire the length of the 16 d-rams
on the PCB. The RAS bus is formed by soldering all the pin
4's of the new d-rams to the solid copper wire. The bus wire
must be seated against the top surface of the new d-rams
without a gap. This insures clearance between the top shield
and the pins of the d-rams.
After soldering all 16 d-rams to the bus clip off any
portion of the pins that extend above the top of the bus
wire. Now cut a #22 AWG solid copper wire the length of the
16 d-rams. Place the bus wire along the top surface of the
new d-rams in contact with all the pin 15's of the new
d-rams. Solder every pin 15 to this bus and as above insure
that the wire is seated solidly against the top surface of
the new d-rams. Cut off all excess pin length sticking up
above the top of the bus wire. Using diagonal cutters remove
the section of the bus connecting the new U30 pin 15 to the
new U32 pin 15. This divides the bus in half with the new
U16, 17, 18, 24, 28, 29 having a common pin 15. The new U32,
33, 34, 38, 42, 43, 44, 45 now have a common pin 15,
seperated from the other common bus.
(NOTE: until step 6 is finished, do no in any way apply
power to your ST. This intermediate state of affairs will
damage your memory chips!!)
Step 5: Remount all the desoldered capacitors. Bend the pins
like they were before resoldering, suchthat they will not
touch the lower shielding. Solder from the non component
side.
Step 6: Orient the 520ST PCB so that you are looking at the
solder side of the PCB (non-component side), with the row of
d-rams nearest you. Find the double square pattern of pads
at the 68-pin socket of the memory controller, U15 (3H2119).
The following is a guide to locating the six memory
controller pins necessary to complete the wiring. The socket
is numbered conterclockwise, starting with pin 1, the square
pad (look closely) in the middle of the bottom outside row.
The sequence, moving counterclockwise from pin 1, first on
the outside square ONLY: (NOTE: the sequence ")(" means to
make a 90-degree turn counterclockwise, i.e. around the
corner)
1,3,5,7,9)(10,12,14,16,18,20,22,24,26)(27,29,31,33,35,37,39,
41,43)(44,46,48,50,52,54,56,58,60)(61,63,65,67
The sequence, moving counterclockwise along the inside
square only, and starting with the left side of the bottom
row:
(62,64,66,68,2,4,6,8)(11,13,15,17,19,21,23,25)(28,30,32,34,3
6,38,40,42)(45,47,49,51,53,55,57,59)
Six 68-ohm 1/4W plus/minus 10% carbon film resistors must be
added when adding memory. These series terminating resistors
minimize undershoot which may damage BOTH BANKS of d-rams if
omitted. Solder a 68-ohm resistor to pin 18 of U15, RAS1.
Solder a #24 AWG stranded wire from the remaining end of the
68-ohm resistor to the pin 4 bus (RAS) of all the new
d-rams. that is the new U16, 17, 18, 24, 25, 28, 29, 30, 32,
33, 34, 38, 42, 43, 44, and 45.
Solder a 68-ohm resistor to pin 22 of U15, CASH1. Solder a
#24 AWG stranded wire from the remaining end of the 68-ohm
resistor to pin 15 bus (CAS) of the new
U45,44,43,42,38,34,33,32.
Solder a 68-ohm resistor to pin 21 of U15, CASIL. Solder a
#24 AWG stranded wire from the remaining end of the 68-ohm
resistor to pin 15 bus (CAS) of the new U30,
29,28,25,24,18,17,16.
For best results in all three cases above solder the wires
coming from the resistors to the middle of the three bus
wires in a "T" fashion rather than at one end of the buses.
Use a continuity tester to find the following three traces
-- do not depend on visual inspection. Now install three
68-ohm series terminating resistors in the original 512K
bank of ram. Be very careful while soldering to these narrow
traces, since excessive heat can easily lift a trace from
the board. Use an Exacto knife to gently remove solder mask
from traces.
Cut the trace leading from pin 8, RAS0, of U15 near U15.
Solder a 68-ohm resistor in series with the trace.
Cut the trace leading from pin 6, CAS0H, of U15 near U15.
Solder a 68-ohm resistor in series with the trace.
Cut the trace leading from pin 7, CAS0L, of U15 near U15.
Solder a 68-ohm resistor in series with the trace.
Step 7: Sit back. Use Brain. Do you feel confident about the
quality of your work? No mistakes? Check evrything once
again if you are but a little uncertain. Applying power with
errors might make your ST into a decorative, nonfunctional
piece of art. OK. Either rebuild your ST into its shielding
and cabinet, or put it onto a surface clear of wires and
solder remians and connect it to monitor, disk and supply.
Boot it.
It it boots, you're probably there. Test if the new memory
works by looking at the phystop variable ($42E) with SID if
you have the developer stuff. It should read $100000 (1M
hex). Also note that memcntlr ($424) now holds 5 instead of
4, and that v_bas_ad ($44E) now holds $F80000 (screen bitmap
origin). If you don't have the developer stuff, try a single
drive copy and check that you get the whole disk in one
buffer instead of two.
If the new memory does not seem to exist, use SID to deposit
and retrieve words on locations $80000 and up (1/2 Meg hex).
If bit errors occur, the ST bootROM did not detect the
extension (it checks all bits of 512 locations by testing a
psedo random sequence, before accepting a memory bank). Try
to pin point the faulty chip(s) and remove the error.
If it doesn't boot, you're in trouble. I'm sorry. It is
difficult to give hints on what to do here. So many
possibilities. Desoldering the new chips probably won't work
(if the old ones were functional, the ST would still boot).
Check for hidden short:circuit on the RAM pins. May also be
that you have a flaky new pin connection.
That's all there is...
--
Jwahar R. Bammi
Usenet: .....!decvax!cwruecmp!bammi
CSnet: bammi@case
Arpa: bammi%case@csnet-relay
CompuServe: 71515,155
------------------------- That's all folks! ------------------------
Technical Details
Message-ID:
<8601081908.AA15608@ucbvax.berkeley.edu>Path:
arc1-ca.pugleaf.net!utzoo!watmath!clyde!cbosgd!ihnp4!ucbvax!info-atariReferences:
<8601071848.AA02085@sdcc6.ARPA>Show Full Headers
Complete Article Headers
Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site ucbvax.BERKELEY.EDU Path: arc1-ca.pugleaf.net!utzoo!watmath!clyde!cbosgd!ihnp4!ucbvax!info-atari From: info-atari@ucbvax.UUCP Newsgroups: net.micro.atari Subject: Re: 1meg upgrade killer Message-ID: <8601081908.AA15608@ucbvax.berkeley.edu> Date: Wed, 8-Jan-86 14:08:21 EST Article-I.D.: ucbvax.8601081908.AA15608 Posted: Wed Jan 8 14:08:21 1986 Date-Received: Thu, 9-Jan-86 03:46:24 EST References: <8601071848.AA02085@sdcc6.ARPA> Sender: daemon@ucbvax.BERKELEY.EDU Organization: The ARPA Internet Lines: 537