Prev: Housing two 2.5-inch hard disks in one 3.5-inch drive bay?
Next: Problem 80 wire cable for hard disk
From: VanguardLH on 1 Mar 2010 13:36
ClueLess wrote:

> Posted to comp.sys.ibm.pc.hardware.storage, 24hoursupport.helpdesk
>
> Hi Friends
>
> Here is a strange problem, when I connect my hard disk using the 80
> wire cable my computer does not recognize it.
>
> It happened this way. I had to check another hard disk and so I
> swapped the disks (IDE0) and found it was not recognized. Then I
> connected the original hard disk and it was also not recognized. I
> tried both the IDE0 and IDE1 channels with the same result. I tried
> also other hard disks as well other 80 wire cables, still no go.
>
> Then I found an old 40 wire cable and when connected with this, bingo,
> all the hard disks are recognized.
>
> No pins broken, no hardware damage. All the cablea are in good
> condition. In fact the machine for years worked with only the 80 wire
> cable.
>
> How does this happen? This is just a banana motherboard and the
> problem is in the bios stage itself. (bios says "Not installed")
>
> If any of you can give me an explanation or a solution please do

Signalling wise, there is no difference between a 40- and 80-wire IDE cable.
Only 40 signal wires are used in both. The 80-wire ribbon cable has a
ground wire between each signal wire to reduce cross-talk and noise. It is
still a 40-pin connector to a 40-pin connector setup.

There are some cables that are physically altered (by snipping out a very
short length of one signal wire, or not connecting that signal to the
connector pin, or with a missing connector pin) to make them "cable select"
cables. Rather than you configuring your drives as master and slave, you
configure both to "cable select" and the lack of the signal from controller
to one of the drives makes it the slave. Unless you using a meter to test
each pin of the 40-pin connector at the mobo end to each matching pin in the
other 40-pin connectors to see if all have continuity or if one is open,
just test using a different 80-wire, 40-pin IDE cable.

It's also possible that the 80-wire, 40-pin cable is damaged so one of its
signal lines is disconnected. Just because it worked for years doesn't mean
it was a perfect cable. The connector's pins aren't soldered to the wires.
The ends of the connector's pins are simply split to scissor through the
insulation to make contact with the wired by squeesing the two tangs against
the wire. It's a mechanical connection. This is how, for example, a
Scotchlock splicing connector works. It isn't the most secure or durable
connection. Also, you pulling on the ribbon cable to remove the cable can
cause damage, especially if you wiggle from side to side to waddle the
connector off the header. That's why the better cables come with a nylon or
plastic pull tab that is part of the connector assembly and you pull on that
rather than the ribbon cable.

You never mentioned if the connectors were polarized. If the connector has
no tang (to slide into a matching notch in the shroud around the header
pins) or doesn't use a blank pin 20 (a solid spot where pin 20 would be but
with no hole for a pin to slide into), or the shroud around the headers is
depolarized by having notches on both sides, or the shroud is missing, it's
possible you attached the connector upside down. It's also possible you
didn't slide the connector all the way onto the header pins. It's also
possible one of the female pins inside the connector is loose and didn't
make good contact. Did you ever try a different 80-wire, 40-pin cable?
From: VanguardLH on 1 Mar 2010 16:08
Rod Speed wrote:

>> Signalling wise, there is no difference between a 40- and 80-wire IDE
>> cable.
<which is qualified in a following paragraph regarding cable select>
>
> Wrong. 80 wire cables are usually cable select cables, 40 wire cables
> usually are not.

Get an ohmmeter to test.

http://www.unitechelectronics.com/ide44pinout.gif
(this is the drive header, same pinout as for mobo header; remember to
mirror the image for ribbon connector)

Cable select cables require a pin (#28) not be connected on one of the
device-side connectors. I already mentioned in my prior post how this can
be done. The host adapter grounds this signal (i.e., when connected, the
device sees a ground line). If pin 28 is connected at the device (i.e., the
device sees this signal is grounded), it is the master device. If this pin
is open (floating), that device is the slave.

Just get a continuity tester to determine in pin 28 is open on one of the
device connectors on the ribbon cable. I have 80-wire, 40-pin cables that
have pin 28 connected and some where it is open. I have not seen a
predominance of one or the other to claim a "usual" configuration, and
because of this is why I test. It all depends on who manufactured the
cable. For example, the manufacturer might want that same 80-wire cable to
be usable in 40-wire cable setups so they feed pin 28 to all connectors.
The same holds true as to whether the connector or shroud are polarized with
a tang or notch, if depolarized by the absense of a tang or having notches
on both sides of the shroud, or if pin 20 is used as a polarizing key (by
pin 20 missing in header and blocked or solid in the connector). You have
to look. For cable-select or not enforced in the manufacture of a 80-wire
ribbon cable, you have to check with a continuity check.

If the connectors on the ribbon cable are color coded, I would suspect
cable-select was enforced through the physical connections as follows:

- Blue: Motherboard. Pin 28 connected.
- Black: Master device. Pin 28 connected.
- Gray: Slave device. Pin 28 not connected.

But I'd still check pin 28 with a continuity tester to make sure. You may
not have cable-enforced cable-select with some 80-wire ribbon cables (i.e.,
pine 28 goes to all connectors) when you want to use cable-select on the
device jumpers. You may not want cable-enforced cable-select if you are
using jumpers on the devices to configure them as master and slave. Whether
you have a cable-select cable is something you should really test.
From: Desk Rabbit on 1 Mar 2010 16:38
On 01/03/2010 18:50, Rod Speed wrote:
> Desk Rabbit wrote:
>> On 01/03/2010 14:21, ClueLess wrote:
>>> Posted to comp.sys.ibm.pc.hardware.storage, 24hoursupport.helpdesk
>>>
>>> Hi Friends
>>>
>>> Here is a strange problem, when I connect my hard disk using the 80
>>> wire cable my computer does not recognize it.
>>>
>>> It happened this way. I had to check another hard disk and so I
>>> swapped the disks (IDE0) and found it was not recognized. Then I
>>> connected the original hard disk and it was also not recognized. I
>>> tried both the IDE0 and IDE1 channels with the same result. I tried
>>> also other hard disks as well other 80 wire cables, still no go.
>>>
>>> Then I found an old 40 wire cable and when connected with this,
>>> bingo, all the hard disks are recognized.
>>>
>>> No pins broken, no hardware damage. All the cablea are in good
>>> condition. In fact the machine for years worked with only the 80 wire
>>> cable.
>>>
>>> How does this happen? This is just a banana motherboard and the
>>> problem is in the bios stage itself. (bios says "Not installed")
>>>
>>> If any of you can give me an explanation or a solution please do
>>>
>> And the hard drive make/model is?
>> And the motherboard make/model is?
>> And the BIOS version is?
>
> All irrelevant given that the original hard drive isnt recognised anymore.
>
>
Any information is useful at this point. It may be a known problem but
without knowing what he's got the chances of finding out are less than zero.
From: thund3rstruck on 1 Mar 2010 16:42
richard wrote:

> I'm amazed that you could even get the 40 wire cable to connect to the same
> connector as the 80 wire. They're generally designed so that you CAN'T do
> that.

Connectors are the same for the 80 as the 40

> Was the first drive you checked properly initialized and formatted per
> maniufacturer's instructions with the provided disk? That's generally why
> it isn't recognized. Did the other drives work elswhere?

The bios should see it if it's configured right via
jumpers/connector/etc. Formatting or not does not matter

> In the device manager, do you get any errors?

If the drive on ide0 isn't seen, how can he boot to Windows, much less
check DevMgr?

n0i
From: Rod Speed on 1 Mar 2010 18:58
VanguardLH wrote
> Rod Speed wrote

>>> Signalling wise, there is no difference between a 40- and 80-wire IDE cable.

> <which is qualified in a following paragraph regarding cable select>

So that claim is just plain wrong. There is in fact that very real difference.

>> Wrong. 80 wire cables are usually cable select cables, 40 wire cables usually are not.

> Get an ohmmeter to test.

Dont need one. I use the ATA standard.

<reams of you proving what I said flushed where it belongs>



First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5
Prev: Housing two 2.5-inch hard disks in one 3.5-inch drive bay?
Next: Problem 80 wire cable for hard disk