4 Apr
2005
4 Apr
'05
2:21 p.m.
OK, I posted about this a few weeks ago and now I have some more
information.
The drives in question are a ST-225 20MB (615/4/17) that came out of an
XT (so would have been using a 8-bit controller) and a ST-238R 30MB
(615/4/26) of unknown origin.
I don't have or even know what the original controllers were, the
machines were gutted/given away years ago by the owners.
I have a 386 testbed which I'm using, and plan to laplink the files over
to another machine once they are accessible.
The only card in the 386 is a VGA card. Turbo is off.
I have the following controllers:
- WD1002A-27X 8-bit RLL
- Adaptec ACB-2370A S2 16-bit RLL/floppy
- WD1002SV-SR2 16-bit RLL/floppy
- Everex EV-346 16-bit MFM/floppy
I also have two Miniscribe 8438 30MB RLL drives. However, they were
used with the 27X, which means the geometry is ambiguous depending on
how the jumpers were set when it was formatted.
Now, I'm either doing something wrong or all these drives are dead.
I am using a floppy cable with a twist in it and a 20-pin data cable.
All of the drives spin up (the Miniscribes needed a little coaxing) and
sound "healthy" as they dance their little self-test jigs. Here's where
the trouble starts.
I tried the ST-238R with all three of the RLL controllers. With the 27X
and with no hard drive entered in the NVRAM setup, since the BIOS can't
be disabled, it runs the BIOS, which can't figure out what's going on
and returns a 1701 POST code.
With both of the 16-bit RLL controllers, 615/4/26 entered in the NVRAM
and the controller BIOS disabled, the drive makes a repetitive seeking
sound like an ECC retry about twice a second, until eventually the BIOS
gives up and returns Drive C: failure. What this seems to indicate to
me is that the drive was formatted with different geometry, or the
tracks have drifted so bad it can't get its bearings.
I also tried the Miniscribe drives with the 27X for kicks. Everything
sounds normal, including the long growl that I remember that
drive/controller combo doing during POST. But, a 1701 is returned,
which isn't normal - this drive *used* to work with this exact
controller.
The only MFM controller I have is the Everex, so I tried the ST-225 with
that one, entering 615/4/17 in the NVRAM and disabling its BIOS.
Unfortunately, I get mostly the same behavior as with the ST-238R;
except that with this drive, the retry clicks are about a second or two
apart. Eventually the BIOS gives up anyway.
I'm fresh out of ideas at this point. Maybe I have a bad cable? Do I
need to find the exact controller the drives were paired with - what
were the most common ones for each drive? Is the 386 the problem, too
fast for these cards?
--
Ryan Underwood, <nemesis at icequake.net>
4 Apr
4 Apr
2:38 p.m.
On Mon, 2005-04-04 at 16:21 -0500, Ryan Underwood wrote:
Now, I'm either doing something wrong or all these
drives are dead.
I am using a floppy cable with a twist in it and a 20-pin data cable.
I don't remember ever using a twisted cable with a hard disk, only with
floppy drive (and then only in the PC world). I think both control /
data cables should just be straight through.
cheers
Jules
2:57 p.m.
On Mon, 2005-04-04 at 16:21 -0500, Ryan Underwood
wrote:
> Now, I'm either doing something wrong or all these drives are dead.
> I am using a floppy cable with a twist in it and a 20-pin data cable.
On Mon,
4 Apr 2005, Jules Richardson wrote:
I don't remember ever using a twisted cable with a
hard disk, only with
floppy drive (and then only in the PC world). I think both control /
data cables should just be straight through.
IBM used a twisted cable with their hard disks - in order to avoid having
to explain "Drive Select" jumpers to their [Computerland level] dealers.
The hard drive cable twist is NOT the same twist as for floppies!
--
Grumpy Ol' Fred cisin at xenosoft.com
3:13 p.m.
From: "Fred Cisin" <cisin at xenosoft.com>
Sent: Monday, April 04, 2005 4:57 PM
If all you have is a twisted floppy cable use the "B-drive" connector
(before the twist). Make sure the drive has the first drive selected.
As stated IBM used a different twisted cable for hard drives and as with
floppies both possible hard drives used the second drive select line.
Unlike the floppy you can use a straight cable and use the first drive
select, the floppy cable changes more than just drive select.
Randy
www.s100-manuals.com
On Mon,
2005-04-04 at 16:21 -0500, Ryan Underwood wrote:
> Now, I'm either doing something wrong or all these drives are dead.
> I am using a floppy cable with a twist in it and a 20-pin data cable.
On Mon,
4 Apr 2005, Jules Richardson wrote:
I don't remember ever using a twisted cable
with a hard disk, only with
floppy drive (and then only in the PC world). I think both control /
data cables should just be straight through.
IBM used a twisted cable with their hard disks - in order to avoid having
to explain "Drive Select" jumpers to their [Computerland level] dealers.
The hard drive cable twist is NOT the same twist as for floppies!
--
Grumpy Ol' Fred cisin at xenosoft.com 
4:36 p.m.
At 02:57 PM 4/4/2005 -0700, you wrote:
Amen brother! Get a straight-thru cable and jumper the drive for DS0
[Computing] "Unix has been feverishly evolving for over 20 years,
sort of like bacteria in a cesspool, only less attractive." --Unix for
Dummies
--... ...-- -.. . -. ----. --.- --.- -...
tpeters at nospam.mixcom.com (remove "nospam") N9QQB (amateur radio)
"HEY YOU" (loud shouting) WEB ADDRESS http//www.mixweb.com/tpeters
43? 7' 17.2" N by 88? 6' 28.9" W, Elevation 815', Grid Square
EN53wc
WAN/LAN/Telcom Analyst, Tech Writer, MCP, CCNA, Registered Linux User 385531
On Mon,
2005-04-04 at 16:21 -0500, Ryan Underwood wrote:
> Now, I'm either doing something wrong or all these drives are dead.
> I am using a floppy cable with a twist in it and a 20-pin data cable.
On Mon,
4 Apr 2005, Jules Richardson wrote:
I don't remember ever using a twisted cable
with a hard disk, only with
floppy drive (and then only in the PC world). I think both control /
data cables should just be straight through.
IBM used a twisted cable with their hard disks - in order to avoid having
to explain "Drive Select" jumpers to their [Computerland level] dealers.
The hard drive cable twist is NOT the same twist as for floppies! 
5 Apr
5 Apr
11:50 a.m.
Jules Richardson wrote:
On Mon, 2005-04-04 at 16:21 -0500, Ryan Underwood
wrote:
My MFM controller/drive setup has two cables and neither are twisted...
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Want to help an ambitious games project? http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
Now, I'm either doing something wrong or all
these drives are dead.
I am using a floppy cable with a twist in it and a 20-pin data cable.
I don't remember ever using a twisted cable with a hard disk, only with
floppy drive (and then only in the PC world). I think both control /
data cables should just be straight through. 
4 Apr
4 Apr
3:04 p.m.
On Monday 04 April 2005 14:21, Ryan Underwood wrote:
I am using a floppy cable with a twist in it and a
20-pin data cable.
Do not use a floppy cable with a twist - you must use straight through cables
on both MFM and RLL drives for both control and data.
Lyle
--
Lyle Bickley
Bickley Consulting West Inc.
http://bickleywest.com
"Black holes are where God is dividing by zero"
3:55 p.m.
Floppy cable won't work. If you want to use the cable cut the second
connector off and make it a single connector straight through cable.
Another problem I had with early MFM hard drives was they needed to be
read with the same model of controller that formatted them.
For the ST225 look for a DTC 5150 or a western digital 1002-WX1 HD
controller. These were two of the most common HD controllers for XTs.
I doubt that you can read them with any of the 16 bit cards.
When I put away any XT drives that I wanted the data off I kept the
cables and controller card with the drive.
--
Paxton Hoag
Astoria, OR
USA
3:18 p.m.
From: "Ryan Underwood" <nemesis-lists at icequake.net>
Sent: Monday, April 04, 2005 4:21 PM
OK, I posted about this a few weeks ago and now I have some more
information.
The drives in question are a ST-225 20MB (615/4/17) that came out of an
XT (so would have been using a 8-bit controller) and a ST-238R 30MB
(615/4/26) of unknown origin.
I don't have or even know what the original controllers were, the
machines were gutted/given away years ago by the owners.
I have a 386 testbed which I'm using, and plan to laplink the files over
to another machine once they are accessible.
The only card in the 386 is a VGA card. Turbo is off.
I have the following controllers:
- WD1002A-27X 8-bit RLL
- Adaptec ACB-2370A S2 16-bit RLL/floppy
- WD1002SV-SR2 16-bit RLL/floppy
- Everex EV-346 16-bit MFM/floppy
I also have two Miniscribe 8438 30MB RLL drives. However, they were
used with the 27X, which means the geometry is ambiguous depending on
how the jumpers were set when it was formatted.
<snip>
If you are trying to recover files you are probably wasting your time, the
XT controllers used unique formatting and you will never read the data off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset.
Randy
www.s100-manuals.com
4:14 p.m.
On Mon, 4 Apr 2005, Randy McLaughlin wrote:
If you are trying to recover files you are probably
wasting your time, the
XT controllers used unique formatting and you will never read the data off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset.
This doesn't make much sense. Can you please explain further?
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
5 p.m.
From: "Vintage Computer Festival" <vcf at siconic.com>
Sent: Monday, April 04, 2005 6:14 PM
On Mon, 4 Apr 2005, Randy McLaughlin wrote:
With Vintage as a domain name you should be familiar with the fact that
standards develop slowly.
Hard drive interfaces for the PC "standardized" with the AT, before that
people did what ever they wanted. Sometimes one controller would read a
disk from a different type of controller but it was just luck.
Most hard disk controllers formatted to 17 sectors of 512 bytes per track
but that doesn't say much about how it is formatted at the hardware level
(bit for bit). Another point is that before the AT the drive geometry was
normally kept on the drive and different controllers stored the information
differently.
With the PC many people hooked up drives before the XT came out but once IBM
came out with the AT they had chosen a "standard" hard drive controller
everyone worked to emulate until RLL controllers came out. Even controllers
like the Everex controllers that did track buffering to allow 1:1 interleave
were still "compatible" with the AT formatting.
If you have an MFM drive used on an AT or better you have a good chance of
reading it on any AT or better system with most any controller.
RLL or 8 bit based systems have no such "standard".
Randy
www.s100-manuals.com
If you are trying to recover files you are
probably wasting your time,
the
XT controllers used unique formatting and you will never read the data
off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset.
This doesn't make much sense. Can you please explain further?
--
Sellam Ismail Vintage Computer
Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger
http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage
mputers ]
[ and academia at www.VintageTech.com || at
http://marketplace.vintage.org ]
5:23 p.m.
On Mon, 2005-04-04 at 19:00 -0500, Randy McLaughlin wrote:
From: "Vintage Computer Festival" <vcf at
siconic.com>
Sent: Monday, April 04, 2005 6:14 PM
I was confused by the statement too. Are you saying that the same
*model* of controller won't necessarily work (due to different
tolerances in components say) - or just that different controllers that
happen to use a few common chips won't work?
The former would be a little alarming - the latter makes much more sense
and would seem like a given. (Maybe that's where Sellam was getting
confused?)
cheers
Jules
On Mon, 4 Apr 2005, Randy McLaughlin wrote:
With Vintage as a domain name you should be familiar with the fact that
standards develop slowly.
Hard drive interfaces for the PC "standardized" with the AT, before that
people did what ever they wanted. Sometimes one controller would read a
disk from a different type of controller but it was just luck. If you are trying to recover files you are
probably wasting your time,
the
XT controllers used unique formatting and you will never read the data
off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset.
This doesn't make much sense. Can you please explain further?
5:47 p.m.
From: "Jules Richardson" <julesrichardsonuk at yahoo.co.uk>
Sent: Monday, April 04, 2005 7:23 PM
On Mon, 2005-04-04 at 19:00 -0500, Randy McLaughlin
wrote:
The incompatibilities are not a tolerance problem, when the same chipset is
used the BIOS can be different and not be compatible. But I've always found
that the same model controllers were interchangeable with no problems.
A few times I've seen different PC controllers from the same manufacturer
(i.e. WD or whatever) work from one model to another but not be able to boot
without a low-level formatting (booting from a floppy allows access).
Other times no compatibility at all and must be low-level formatted to do
anything.
As was pointed out if you know the drives were formatted on a 8 bit
controller just set the BIOS to no hard drive and don't even try the 16 bit
controllers, they would just be a waste of time. Your best bet would be to
try different 8 bit cards and boot to a floppy then do a "dir c:", repeat
until you've tried all cards.
Randy
www.s100-manuals.com
From: "Vintage Computer Festival"
<vcf at siconic.com>
Sent: Monday, April 04, 2005 6:14 PM
I was confused by the statement too. Are you saying that the same
*model* of controller won't necessarily work (due to different
tolerances in components say) - or just that different controllers that
happen to use a few common chips won't work?
The former would be a little alarming - the latter makes much more sense
and would seem like a given. (Maybe that's where Sellam was getting
confused?)
cheers
Jules On Mon, 4 Apr 2005, Randy McLaughlin wrote:
With Vintage as a domain name you should be familiar with the fact that
standards develop slowly.
Hard drive interfaces for the PC "standardized" with the AT, before that
people did what ever they wanted. Sometimes one controller would read a
disk from a different type of controller but it was just luck. If you are trying to recover files you are
probably wasting your time,
the
XT controllers used unique formatting and you will never read the data
off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even
chipset.
This doesn't make much sense. Can you please explain further?
5:58 p.m.
On Tue, 5 Apr 2005, Jules Richardson wrote:
I was confused by the statement too. Are you saying
that the same
*model* of controller won't necessarily work (due to different
tolerances in components say) - or just that different controllers that
happen to use a few common chips won't work?
From my recollections of working on XT systems in the last 80s, when
swapping MFM controllers, you had to make sure they used the same chipset.
For instance, a drive formatted by a WD controller would probably NOT be
readable by a DTC controller.
Mike Loewen mloewen at cpumagic.scol.pa.us
The Dixie Lion Jazz Band http://ripsaw.cac.psu.edu/dixie.html
The B9 Robot Builders Club B9-0014 http://ripsaw.cac.psu.edu/~mloewen/B9/
Old Technology http://ripsaw.cac.psu.edu/~mloewen/Oldtech/
5:01 p.m.
At 04:14 PM 4/4/2005 -0700, you wrote:
On Mon, 4 Apr 2005, Randy McLaughlin wrote:
I agree.
I've had problems with controllers that can be jumpered to lie about drive
geometry being mismatched with the geometry used to format the drive in the
first place, but WD MFM controllers of the same era could usually be
counted on to read any given MFM drive within their specs, and the RLL
controllers likewise (with RLL drives). Else how did I replace drives, or
controllers, sometimes separately, on many a service call?
"Not at all tricky. We do this sort of stuff every day before
breakfast. Then I fly to work on my winged pig, Swilma." --David Chase
--... ...-- -.. . -. ----. --.- --.- -...
tpeters at nospam.mixcom.com (remove "nospam") N9QQB (amateur radio)
"HEY YOU" (loud shouting) WEB ADDRESS http//www.mixweb.com/tpeters
43? 7' 17.2" N by 88? 6' 28.9" W, Elevation 815', Grid Square
EN53wc
WAN/LAN/Telcom Analyst, Tech Writer, MCP, CCNA, Registered Linux User 385531
If you are trying to recover files you are
probably wasting your time, the
XT controllers used unique formatting and you will never read the data off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset.
This doesn't make much sense. Can you please explain further?
4:38 p.m.
At 05:18 PM 4/4/2005 -0500, you wrote:
From: "Ryan Underwood" <nemesis-lists at
icequake.net>
Sent: Monday, April 04, 2005 4:21 PM
I have been able to read many old MFM drives and the occasional RLL drive.
When the original controller was jumpered to lie about head or cylinder
count or other facts of drive geometry, I have be largely unable to read
those. This was several years ago.
OK, I posted about this a few weeks ago and now I have some more
information.
The drives in question are a ST-225 20MB (615/4/17) that came out of an
XT (so would have been using a 8-bit controller) and a ST-238R 30MB
(615/4/26) of unknown origin.
I don't have or even know what the original controllers were, the
machines were gutted/given away years ago by the owners.
I have a 386 testbed which I'm using, and plan to laplink the files over
to another machine once they are accessible.
The only card in the 386 is a VGA card. Turbo is off.
I have the following controllers:
- WD1002A-27X 8-bit RLL
- Adaptec ACB-2370A S2 16-bit RLL/floppy
- WD1002SV-SR2 16-bit RLL/floppy
- Everex EV-346 16-bit MFM/floppy
I also have two Miniscribe 8438 30MB RLL drives. However, they were
used with the 27X, which means the geometry is ambiguous depending on
how the jumpers were set when it was formatted.
<snip>
If you are trying to recover files you are probably wasting your time, the
XT controllers used unique formatting and you will never read the data off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset. Randy
www.s100-manuals.com
[Government]I don't make jokes. I just watch the government and
report the facts. --Will Rogers
--... ...-- -.. . -. ----. --.- --.- -...
tpeters at nospam.mixcom.com (remove "nospam") N9QQB (amateur radio)
"HEY YOU" (loud shouting) WEB ADDRESS http//www.mixweb.com/tpeters
43? 7' 17.2" N by 88? 6' 28.9" W, Elevation 815', Grid Square
EN53wc
WAN/LAN/Telcom Analyst, Tech Writer, MCP, CCNA, Registered Linux User 385531
5:04 p.m.
From: "Tom Peters" <tpeters at mixcom.com>
Sent: Monday, April 04, 2005 6:38 PM
<snip>
I would be willing to bet you are talking about 16 bit controllers, most 16
bit controllers are interchangeable. Very few 8 bit controllers are
interchangeable.
Randy
www.s100-manuals.com
If you are
trying to recover files you are probably wasting your time, the
XT controllers used unique formatting and you will never read the data off
of the drives without using controllers indentical with ones used
originally. It is not good enough to use the same brand or even chipset.
I have been able to read many old MFM drives and the occasional RLL drive.
When the original controller was jumpered to lie about head or cylinder
count or other facts of drive geometry, I have be largely unable to read
those. This was several years ago.
>Randy
>www.s100-manuals.com 
9 p.m.
On Mon, 4 Apr 2005 19:04:24 -0500
"Randy McLaughlin" <cctalk at randy482.com> wrote:
From: "Tom Peters" <tpeters at
mixcom.com>
Sent: Monday, April 04, 2005 6:38 PM
<snip>
I would be willing to bet you are talking about 16 bit controllers,
most 16 bit controllers are interchangeable. Very few 8 bit
controllers are interchangeable.
I thought it would be worth adding, to clarify this for some people who
aren't familiar with the tech involved, some of the background regarding
AT versus XT HD controllers. The IBM PC-AT and AT clones have direct
BIOS support for the hard disk and controller built into the
motherboard. XT and XT-clone motherboards have no direct built-in
support for hard drives. The Hard Drive BIOS support is included as a
'BIOS extension' on a ROM on the 8-bit controller card. Said BIOS
extension code can and is widely varied from controller to controller.
(For further background: the 'BIOS extension' is a feature in the PC
BIOS whereas the built-in BIOS on the motherboard scans a certain
address space looking for a specific signature that indicates a plug-in
card on the I/O channel has inserted BIOS extension ROM at that address,
then loads it as part of the BIOS before proceeding to boot.)
I may not be completely accurate in the above. Anybody, please add to
or correct me as appropriate.
-Scott
If you are
trying to recover files you are probably wasting your
time, the >XT controllers
used unique formatting and you will never
read the data off >of the drives without using controllers indentical
with ones used >originally. It is not good enough to use the same
brand or even chipset.
I have been able to read many old MFM drives and the occasional RLL
drive. When the original controller was jumpered to lie about head
or cylinder count or other facts of drive geometry, I have be
largely unable to read those. This was several years ago.
>Randy
>www.s100-manuals.com
9:39 p.m.
From: "Scott Stevens" <chenmel at earthlink.net>
Sent: Monday, April 04, 2005 11:00 PM
<snip>
I thought it would be worth adding, to clarify this
for some people who
aren't familiar with the tech involved, some of the background regarding
AT versus XT HD controllers. The IBM PC-AT and AT clones have direct
BIOS support for the hard disk and controller built into the
motherboard. XT and XT-clone motherboards have no direct built-in
support for hard drives. The Hard Drive BIOS support is included as a
'BIOS extension' on a ROM on the 8-bit controller card. Said BIOS
extension code can and is widely varied from controller to controller.
(For further background: the 'BIOS extension' is a feature in the PC
BIOS whereas the built-in BIOS on the motherboard scans a certain
address space looking for a specific signature that indicates a plug-in
card on the I/O channel has inserted BIOS extension ROM at that address,
then loads it as part of the BIOS before proceeding to boot.)
I may not be completely accurate in the above. Anybody, please add to
or correct me as appropriate.
-Scott
Exactly, any variation in hardware or firmware causes compatibility
problems, with an AT 99% of that disappears.
The problem re-emerged on the AT once on board BIOS's and/or different
hardware was used such as RLL.
On 8 bit PC's trying to recover data when a controller dies is much more of
an issue than with 16 bit systems.
I am not trying to say it can't be done but it is a problem which is proven
by the fact that I have seen lots of ideas flying around but have yet to
hear that any of the data has been recovered.
If it was me and I knew there was info I absolutely had to have I might try
a 16 bit controller and see if I could physically read any sectors, if so an
assembly language routine could be written to transfer all readable sectors
to a file on a second hard drive. I would bet money that it wouldn't work
since by definition the hardware is different. Trying it with the wrong 8
bit controller would be even harder.
The best chance of recovery is as said in a previous post to keep trying
different 8 bit controllers and booting with a floppy.
The things working against you:
If it was from a specialty computer like a Tandy they used a different
incompatible set of controllers. Some 8 bit computers had their own hard
disk BIOS's on the system board and required disabling the controllers BIOS
in that case you must find the same type computer and the same type
controller.
Finding the right controller.
Does the controller require a real 8 bit machine, yes not all 8 bit
controllers even worked in 16 bit computers.
Is the drive bad.
Has the drive formatting been messed up.
Is the drive terminated properly.
Is the drive jumpered properly.
To work everything has to work, to fail only one thing has to fail.
Can anyone guess why they charge so much for data recovery?
On IDE all of the truly low level stuff is 100% hidden so the only
compatibility problems with IDE tended to be drive geometry (early problems
when the same drive could be addressed by different geometries) .
Good luck,
Randy
www.s100-manuals.com
10:01 p.m.
New subject: OT: Removing Electrolytic cap residue
A friend of mine has a problem with Electrolytic caps on his motherboard
having gone "pop" & deposited their contents on some contacts...
Although this is a PeeCee (spit) that's affected, I thought I'd ask the
assembled knowledgeable types here, as it's also a problem which COULD
affect our Olde Worlde computers quite easily...
Quote
"Does anyone know how best to remove dried capacitor electrolyte from a
surface? I have tried a number of solvents, including Ultrsolve and Super
Servisol, but the stuff seems is set hard.
A couple of capacitors leaked on a motherboard and have dripped into one of
the PCI slots. They are insulating some of the pins, so the slot cannot be
used, but I need the slot. The capacitors were easily replaced (just
pennies and half an hour in front of the TV) but I'm really not sure how to
clean those PCI slot contacts.
The electrolyte dries to a brown varnish-like film. There seem to be a
bunch of these exploded capacitors coming in from PCs manufactured around
2002 (I reckon the worldwide supply of dodgy caps from the 1999/2000
production were put in storage for a few years, then bought up by Packard
Bell and friends)."
Endquote
Thanks in advance, all.
Cheers,
Ade.
5 Apr
5 Apr
12:58 a.m.
New subject: OT: Removing Electrolytic cap residue
Adrian wrote:
A friend of mine has a problem with Electrolytic caps
on his motherboard
having gone "pop" & deposited their contents on some contacts...
[...]
Does anyone know how best to remove dried capacitor
electrolyte from a
surface?
I don't know, but my solution to this has been to replace the affected
motherboards. Those capacitors really are necessary for the board to
be reliable, as they are part of the power supply for the internal CPU
operating voltage. If some of them went bad, you can be pretty sure that
the rest will too, so they all need to be replaced. The local surplus
places have old motherboards for about $20; the cost of a set of new
capacitors is more than that. When I factor in my time to replace them,
it would even be cost-effective to buy a brand new motherboard.
This normally happens to motherboards that were made using cheap
Taiwanese electrolytic capacitors. Someone stole the electrolyte
formula from one of the Japanese capacitor manufacturers, but they
apparently missed some ingredients that were essential to stabilize
it, or perhaps they left those out deliberately to make the stuff
cheaper. The result is that multiple Taiwanese capacitor vendors
were making defective capacitors for quite a long time.
I'll note that I have never seen this problem with any of my Asus or
Tyan motherboards, only with cheap crap like ECS which other people buy
then expect me to support.
Eric
4:32 a.m.
New subject: OT: Removing Electrolytic cap residue
On Tue, 2005-04-05 at 00:58 -0700, Eric Smith wrote:
Adrian wrote:
If the motherboard is "valuable" in some way (containing the right slot
layout for the user, or having a decent BIOS, or onboard SCSI or
something) then one option might just be to replace the damaged PCI
slot.
I've never tried this, but it can't be much harder than snipping a dead
IC from a board, cleaning everything up, and replacing. Either buy a new
PCI socket from a component place, or try removing one from a dead
motherboard. I've never tried the latter (only with ISA sockets) - but a
blowtorch on the underside of the donor board might do the trick without
damaging the PCI slot....
A friend of mine has a problem with Electrolytic
caps on his motherboard
having gone "pop" & deposited their contents on some contacts...
[...]
Does anyone know how best to remove dried
capacitor electrolyte from a
surface?
I don't know, but my solution to this has been to replace the affected
motherboards. I'll note that I have never seen this problem with
any of my Asus or
Tyan motherboards
FWIW I've seen it on a couple of dual-CPU Asus boards before now (but
have seen other Asus boards from the same era that have been fine).
cheers
Jules
5:02 a.m.
New subject: OT: Removing Electrolytic cap residue
Ok, working on thousands of mobo's and PC's (ick, i know) over the past
decade, I must agree and disagree. Most Asus and Tyan boards have been ok,
but, about 3-4 years ago, they also had the bad cap issues. The Asus A7V
(early Socket A AMD Athlon/Duron line) and some of the 266 series had a
leaky cap/cap failure issue for a while. Ended up replacing about 20 or so
boards because of this (shop was too cheap to just replace caps when
another board was on hand and took less time to swap boards than to fix the
damned problem). As with the Tyans, some of the P-III socket 370 boards
from 2000-2002 had such an issue as well, though, less
so than Asus and
obviously, much less than other manufacturers, such as Abit (the
whole
KT7/KT7A series was known to be an 8/10 failure rate for that time frame.
later revisions, I believe called the 1.3 to 1.6 revisions, did not have
these issues as different caps were used.).
As far as I could remember, if the caps leaked that badly, there was a good
chance that there was damage to other circuitry on the board and any
attempts to repair should just go to replacing the board at that point to
limit headaches from voltage and circuit instability caused by voltage flux
from the bad caps. Once had an Iwill board in the shop
I had worked in do
that from caps going bad and could not figure out why it was
giving errors
in diags/Winblows until we re-examined the board with a magnifying glass
and swapped out all the other components (ram, cpu, vid card, etc) and
found there was a small trace going from an area near one of the caps to
the northbridge that was a bit messed up. Replaced the board and things
were still a bit messed up, had to replace the ram too and then everything
cleared up.
Voltage flux is nasty as sometimes, it is an unseen killer of parts. Best
bet, if unsure if stuff is still safe, replace it and let someone else
worry about headaches.
-John Boffemmyer IV
At 07:32 AM 4/5/2005, you wrote:
On Tue, 2005-04-05 at 00:58 -0700, Eric Smith wrote:
Adrian wrote:
If the motherboard is "valuable" in some way (containing the right slot
layout for the user, or having a decent BIOS, or onboard SCSI or
something) then one option might just be to replace the damaged PCI
slot.
I've never tried this, but it can't be much harder than snipping a dead
IC from a board, cleaning everything up, and replacing. Either buy a new
PCI socket from a component place, or try removing one from a dead
motherboard. I've never tried the latter (only with ISA sockets) - but a
blowtorch on the underside of the donor board might do the trick without
damaging the PCI slot....
A friend of mine has a problem with Electrolytic
caps on his motherboard
having gone "pop" & deposited their contents on some contacts...
[...]
Does anyone know how best to remove dried
capacitor electrolyte from a
surface?
I don't know, but my solution to this has been to replace the affected
motherboards. I'll note that I have never seen this problem
with any of my Asus or
Tyan motherboards
FWIW I've seen it on a couple of dual-CPU Asus boards before now (but
have seen other Asus boards from the same era that have been fine).
cheers
Jules 
10:41 a.m.
New subject: OT: Removing Electrolytic cap residue
Jules Richardson wrote:
If the motherboard is "valuable" in some way
(containing the
right slot layout for the user, or having a decent BIOS, or
onboard SCSI or something) then one option might just be to
replace the damaged PCI slot.
I've never tried this, but it can't be much harder than
snipping a dead IC from a board, cleaning everything up, and
replacing. Either buy a new PCI socket from a component place,
or try removing one from a dead motherboard. I've never tried
the latter (only with ISA sockets) - but a blowtorch on the
underside of the donor board might do the trick without
damaging the PCI slot....
I've removed a few ISA and PCI slots using a hot-air gun.
My success rate for ISA was quite good but PCI was less good:
I did salvage a few but most of the time part of the socket
also melted and one or more connectors moved.
I was doing this "just because I could" (I was scrounging
other bits at the time) so I don't know whether the
salvaged ISA or PCI slots would have worked again.
A blow-torch seems even more destructive than a hot air gun!
Antonio
--
---------------
Antonio Carlini arcarlini at iee.org
3:58 p.m.
New subject: OT: Removing Electrolytic cap residue
I don't know, but my solution to this has been to
replace the affected
motherboards. Those capacitors really are necessary for the board to
be reliable, as they are part of the power supply for the internal CPU
operating voltage. If some of them went bad, you can be pretty sure that
the rest will too, so they all need to be replaced. The local surplus
places have old motherboards for about $20; the cost of a set of new
capacitors is more than that. When I factor in my time to replace them,
it would even be cost-effective to buy a brand new motherboard.
This assumes, though, that the capacitors on that replacement motherboard
are of reasonable qulaity. Personally, I'd probably replace the
capacitors anyway on any second-hand motherboard before they caused any
problems. A handful of electrolytics are a lot cheaper than the time
wasted and hassle casued if the machien fails at the wrong moment.
-tony
5:59 a.m.
New subject: OT: Removing Electrolytic cap residue
I would try Lime-Away. I haven't tried it for caps but it works good on
cleaning up the residue from NiCad batteries and I think the cap residue is
a similar material.
Joe
At 06:01 AM 4/5/05 +0100, you wrote:
A friend of mine has a problem with Electrolytic caps
on his motherboard
having gone "pop" & deposited their contents on some contacts...
Although this is a PeeCee (spit) that's affected, I thought I'd ask the
assembled knowledgeable types here, as it's also a problem which COULD
affect our Olde Worlde computers quite easily...
Quote
"Does anyone know how best to remove dried capacitor electrolyte from a
surface? I have tried a number of solvents, including Ultrsolve and Super
Servisol, but the stuff seems is set hard.
A couple of capacitors leaked on a motherboard and have dripped into one of
the PCI slots. They are insulating some of the pins, so the slot cannot be
used, but I need the slot. The capacitors were easily replaced (just
pennies and half an hour in front of the TV) but I'm really not sure how to
clean those PCI slot contacts.
The electrolyte dries to a brown varnish-like film. There seem to be a
bunch of these exploded capacitors coming in from PCs manufactured around
2002 (I reckon the worldwide supply of dodgy caps from the 1999/2000
production were put in storage for a few years, then bought up by Packard
Bell and friends)."
Endquote
Thanks in advance, all.
Cheers,
Ade.
12:07 p.m.
New subject: Removing Electrolytic cap residue
Flux cleaner may work, or just wash in distilled water.
Jim.
----- Original Message -----
From: "Adrian Vickers" <javickers at solutionengineers.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, April 05, 2005 6:01 AM
Subject: OT: Removing Electrolytic cap residue
A friend of mine has a problem with Electrolytic caps
on his motherboard
having gone "pop" & deposited their contents on some contacts...
Although this is a PeeCee (spit) that's affected, I thought I'd ask the
assembled knowledgeable types here, as it's also a problem which COULD
affect our Olde Worlde computers quite easily...
Quote
"Does anyone know how best to remove dried capacitor electrolyte from a
surface? I have tried a number of solvents, including Ultrsolve and Super
Servisol, but the stuff seems is set hard.
A couple of capacitors leaked on a motherboard and have dripped into one
of
the PCI slots. They are insulating some of the pins,
so the slot cannot be
used, but I need the slot. The capacitors were easily replaced (just
pennies and half an hour in front of the TV) but I'm really not sure how
to
clean those PCI slot contacts.
The electrolyte dries to a brown varnish-like film. There seem to be a
bunch of these exploded capacitors coming in from PCs manufactured around
2002 (I reckon the worldwide supply of dodgy caps from the 1999/2000
production were put in storage for a few years, then bought up by Packard
Bell and friends)."
Endquote
Thanks in advance, all.
Cheers,
Ade.
4 Apr
4 Apr
4:35 p.m.
At 04:21 PM 4/4/2005 -0500, you wrote:
The drives in question are a ST-225 20MB (615/4/17)
that came out of an
XT (so would have been using a 8-bit controller) and a ST-238R 30MB
(615/4/26) of unknown origin.
Need RLL controller for the ST-238 and MFM for the ST-225. Sorry if that
was already abundantly obvious.
I don't have or even know what the original
controllers were, the
machines were gutted/given away years ago by the owners.
I have a 386 testbed which I'm using, and plan to laplink the files over
to another machine once they are accessible.
The only card in the 386 is a VGA card. Turbo is off.
I have the following controllers:
- WD1002A-27X 8-bit RLL
- Adaptec ACB-2370A S2 16-bit RLL/floppy
- WD1002SV-SR2 16-bit RLL/floppy
- Everex EV-346 16-bit MFM/floppy
I have docs for the WD cards. The 27X sounds (without looking) like an RLL
controller.
I also have two Miniscribe 8438 30MB RLL drives.
However, they were
used with the 27X, which means the geometry is ambiguous depending on
how the jumpers were set when it was formatted.
Now, I'm either doing something wrong or all these drives are dead.
I am using a floppy cable with a twist in it and a 20-pin data cable.
That certainly will NOT work. The floppy cable is for floppy drives. Get a
MFM data cable.
All of the drives spin up (the Miniscribes needed a
little coaxing) and
sound "healthy" as they dance their little self-test jigs. Here's where
the trouble starts.
I tried the ST-238R with all three of the RLL
controllers. With the 27X
and with no hard drive entered in the NVRAM setup, since the BIOS can't
be disabled, it runs the BIOS, which can't figure out what's going on
and returns a 1701 POST code.
With both of the 16-bit RLL controllers, 615/4/26 entered in the NVRAM
and the controller BIOS disabled, the drive makes a repetitive seeking
sound like an ECC retry about twice a second, until eventually the BIOS
gives up and returns Drive C: failure. What this seems to indicate to
me is that the drive was formatted with different geometry, or the
tracks have drifted so bad it can't get its bearings.
I also tried the Miniscribe drives with the 27X for kicks. Everything
sounds normal, including the long growl that I remember that
drive/controller combo doing during POST. But, a 1701 is returned,
which isn't normal - this drive *used* to work with this exact
controller.
Wrong 34-pin cable. Also check the markings on the controllers and the
drives very carefully for the pin 1 identification. Perhaps all the solder
pads are round, except for one which is deliberately square, (that's pin 1)
or maybe it's silk screened on the board/drive.
The only MFM controller I have is the Everex, so I
tried the ST-225 with
that one, entering 615/4/17 in the NVRAM and disabling its BIOS.
Unfortunately, I get mostly the same behavior as with the ST-238R;
except that with this drive, the retry clicks are about a second or two
apart. Eventually the BIOS gives up anyway.
I'm fresh out of ideas at this point. Maybe I have a bad cable? Do I
need to find the exact controller the drives were paired with - what
were the most common ones for each drive? Is the 386 the problem, too
fast for these cards?
No, it's not too fast. Try the right cable and the Western Digital
controllers.
[Peace] It is an unfortunate fact that we can secure peace only by
preparing for war. --John F. Kennedy
--... ...-- -.. . -. ----. --.- --.- -...
tpeters at nospam.mixcom.com (remove "nospam") N9QQB (amateur radio)
"HEY YOU" (loud shouting) WEB ADDRESS http//www.mixweb.com/tpeters
43? 7' 17.2" N by 88? 6' 28.9" W, Elevation 815', Grid Square
EN53wc
WAN/LAN/Telcom Analyst, Tech Writer, MCP, CCNA, Registered Linux User 385531
7204
days inactive
7205
days old
27 comments
18 participants
participants (18)
-
a.carlini＠ntlworld.com -
ard＠p850ug1.demon.co.uk -
cctalk＠randy482.com -
chenmel＠earthlink.net -
cisin＠xenosoft.com -
eric＠brouhaha.com -
innfoclassics＠gmail.com -
javickers＠solutionengineers.com -
jim＠g1jbg.co.uk -
john_boffemmyer_iv＠boff-net.dhs.org -
julesrichardsonuk＠yahoo.co.uk -
lbickley＠bickleywest.com -
mloewen＠cpumagic.scol.pa.us -
nemesis-lists＠icequake.net -
rigdonj＠cfl.rr.com -
tpeters＠mixcom.com -
trixter＠oldskool.org -
vcf＠siconic.com