Win XP - Hard Drive Compatability

[GrahamW2]

PC - Dell Dimension 8200 - Pentium 4 - 2.8GHz - WinXP SP3 - "born" 2002

I have previously added an extra HDD from a Sky TV PVR. For several years I have had no problem with this 250GB drive.

Maxtor DiamondMax 10 - 6L250RO
250GB - PATA133
Jumpered to "Cable Select"

I am trying to swap this for a newer drive.

Western Digital Caviar Blue - WD5000AAKB
500GB - PATA /16MB Cache
Jumpered to "Cable Select"

I first put the new HDD in a USB caddy, deleted the existing partition, created a new one (NTFS) and let the PC use its default settings. I copied across all of my data onto the new drive whilst it was in the caddy.

No problems at all.

However, when I installed the HDD in the PC, it didn't work properly at all.

Boot time was doubled and the PC was "laggy" from the outset. I could "see" the new HDD (and I even changed its drive letter) but trying to copy to it was very slow indeed. Within a couple of mins the PC "hung" completely and went Blue Screen with an alarming message about "avoiding damage to the PC".

I thought that I must have damaged the PC in some way when I had it opened up. However, reapeated swapping back and forth has show that the PC performs just fine when this HDD is removed ... and fine when the old one is in. Whenever this new HDD is installed the PC exhibits this sluggish behaviour. On one occasion it crashed with a minute or two of booting up ... without me even having clicked on the new drive.

Am I doing something REALLY dumb ?

Can XP support a HDD of 500GB ? Is this HDD "too fast/new" ? Can I slow it down ? Should I have prepared it differently (i.e. formatted it when in the PC rather than in a USB caddy) ? Do I need to install a specific driver first, rather than just rely on Windows plug n play ? Is there some setting I should change somewhere in the o/s ?

I have gone back to the old drive for the time being, but it is nearly full (and quite old). I suppose I could use the new one in the caddy - but that's messy (and probably not very fast).

Any ideas ?

Regards

Graham

[gerbil]

500GB is certainly not too large. There are no drivers as such, the ATA interface handles all comms. Now... er... it's a while since i actually held a PATA drive in my hands, so i may be a little off-beam here...
When you inserted the new PATA, and started the sys for the first time did it recognize new hardware? [this is the bit I'm not sure about; I know a new SATA is recognized thus, I just don't recall about adding PATAs.. :( ]
Anyway, this is what I'm thinking: the sys knows your PATA from the USB, knows its signature and has associated it with a drive letter; perhaps now that it is seeing the disk from a different angle [the ATA interface] it is a little confused. So [and this will do no harm at all] try using regedit to open the HKLM\system\mounteddevices key. Find your PATA via its drive letter in the DosDevices section, note the value of its Data entry. Now go up the list and find the corresponding \??\Volume entry with matching Data value- delete it, then delete the DosDevices entry also. Restart your sys [a new entry will then be generated, and the USB origin will be forgotten].

[GrahamW2]

When you inserted the new PATA, and started the sys for the first time did it recognize new hardware? [this is the bit I'm not sure about; I know a new SATA is recognized thus, I just don't recall about adding PATAs.. :( ]
Anyway, this is what I'm thinking: the sys knows your PATA from the USB, knows its signature and has associated it with a drive letter; perhaps now that it is seeing the disk from a different angle [the ATA interface] it is a little confused. So [and this will do no harm at all] try using regedit to open the HKLM\system\mounteddevices key. Find your PATA via its drive letter in the DosDevices section, note the value of its Data entry. Now go up the list and find the corresponding \??\Volume entry with matching Data value- delete it, then delete the DosDevices entry also. Restart your sys [a new entry will then be generated, and the USB origin will be forgotten].

Hi Gerbil,

Thanks for the reply.

Let me just check here.

a) This will DO NO HARM AT ALL !

When I boot up my system (with either drive ... or indeed no drive) I can't spot anything like the "New System Device Found ... Looking for drivers" type stuff. The PC simply comes on and everything is there in "My Computer".

I manually assigned a drive letter (N) to the new drive (the same letter I previously used for the old drive). Could that cause a problem ?

Oh, and finally, can I just confirm that these deletions to the Registry ... WILL DO NO HARM AT ALL !

Regards

Graham

[gerbil]

:)
Nope. You might notice that there are a heck of a lot of \??\Volume entries - they are the devices you have once had connected, like any and all USB flashdrives, cameras etc. Windows remembers all. You can actually delete the whole MounteDevices key if you so wish; Windows at next restart will regenerate entries for the drives [volumes] that are then connected, and assign them all new letters. The problem with that is that you might have a particular order that you have used which is reflected in registry or other pgms, such as a backup referencing a certain drive. So then you have to change them. And it may get tricky if the system is not on C:
But simplest is to just go with the two related entries that I gave above. You're not walking through the valley of the shadow here....

[tigerbright]

Hi , i suggest install the second PATA drive on to second IDE onboard channel in place of the optical drive, and clone the first to the second using a floppy drive DOS bootable cloning tool - this will ensure both drives are configured the same by BIOS and avoid possible usb connected issues
Once cloned just replace the old primary drive with the new drive and reinstall the optical drive into the second ide channel
You will have one nominal 256GB partition as per the source hdd , so just initialize and format the remaining free space in Windows disk management after booting to Windows for 2 x 256GB partitions

[GrahamW2]

:)
Nope. You might notice that there are a heck of a lot of \??\Volume entries - they are the devices you have once had connected, like any and all USB flashdrives, cameras etc. Windows remembers all. You can actually delete the whole MounteDevices key if you so wish;

That's a neat trick. I removed all refs to the N: I was using.

However, the problem with the new drive always recurs ... even when I don't re-allocate it to N: and just leave it as F:

The problem seems to really kick off when I try and write to this drive ... although, in fairness, if I leave it alone for a bit the machine still goes belly up (and there's no getting away from the fact that the boot process takes longer when this drive is in). I performed numerous reboots, before and after deleting the registry entries and before and after re-allocating the new drive to N: Nothing makes the slightest difference.

TigerBright's idea is a bit beyond me. But it does suggest another idea. Installing the drive and then deleting the partition immediately. Then start over with a new partition and let the PC format the drive in situ. I could then put the old drive in a caddy and "copy backwards" from the caddy to the drive.

Might that work ?

I'm loathe to just "give it a go" because, if it doesn't work, I will have wasted a couple of really boring afternoons copying all the data across the first time.

If there some "low level" software problem (like that registry entry) then re-formatting the disk will be pointless - won't it ?

Any ideas welcome

Regards

Graham

[tigerbright]

Hi Graham, the cloning process is well tested and an easy process for those familiar with basic hardware upgrading
I suggest consult this site for additional information

http://www.xxclone.com/iwhatis.htm

The process will not erase the source primary drive - so you can remove that drive and test the target drive to confirm integrity

Once the target drive is confirmed good - it is suggested to either retain the source drive as the original oem spec hdd (presuming has a hidden oem recovery partition) or totally fully erase using tool from here (hhd vendor erase tools do not or should not erase the recovery partition which also contains oem related Windows licensing data)
http://hddguru.com/
(select hddguru erasure tools listed will erase all data of all partitions on the disk including hidden)
and use as a clean internal secondary drive or convert into a usb connected external drive

Gregory

[gerbil]

Okay... considering that you are loath to waste the time spent already loading the new PATA drive then perhaps you might run a quick check to see if Windows is fully recognising it and its data. Well, the actual check is quick, but you should spend time reading the help file.. :)
In that regard, might I suggest Grenier's Testdisk from http://www.cgsecurity.org/
Running it through its steps should confirm that the disk structure is as it should be.

[GrahamW2]

Okay... considering that you are loath to waste the time spent already loading the new PATA drive then perhaps you might run a quick check to see if Windows is fully recognising it and its data. Well, the actual check is quick, but you should spend time reading the help file.. :)
In that regard, might I suggest Grenier's Testdisk from http://www.cgsecurity.org/
Running it through its steps should confirm that the disk structure is as it should be.

Well, if I had any confidence that deleting the partition and re-formatting the drive in the PC (rather than in the USB caddy as before) would make it work properly, then I am happy to do it. However my feeling is that I simply need to change a setting somewhere.

For what it's worth, the drive seems to be completely "readable" when in the PC. It's just that writing to it (or using the PC for 5-10 mins) results in a "hang". And of course, when it's in the caddy, it is completely read/writeable without problem. It has been used in the caddy for many hours (over a couple of days) without problem (whilst it was loaded up with backups and photos totalling over 200GB !).

I'll put the drive back in the PC and run Testdisk on it when I get a chance and report back.

I don't want to "clone" a 250GB partition ... I want to end up with one 500GB drive.

Do you think that blanking it and re-formatting the drive in the PC (i.e. on the IDE channel rather than via USB) will have a high likelihood of making it work properly ? If it's a case of "That would have to work, otherwise the drive is faulty" then I might as well do it ... as the loading up time was wasted if the drive is duff anyway :o(

Regards

Graham

[tigerbright]

Here is a freeware partition manager whereby you can resize partitions
http://www.partition-tool.com/personal.htm

[tigerbright]

k, re your comment of issue ; However, when I installed the HDD in the PC, it didn't work properly at all.

Should not make a difference but check the optical drive configuration - should be CS, like the hdds
Also check the optical drive signal ribbon cable - if is uses 40wire , then I would suggest system is not designed for CS configuration (all drives on the onboard host bus adapter must be either CS or master/slave) and as such i would suggest resetting all drives to appropriate drive0/1 (master/slave) configurations or installing a 80wire IDE signal ribbon cable for the optical drive that supports cable select
(goes without saying for the hdd's as well)

[tigerbright]

I evaluated the specs of your dimension 8200 and is of a similar era to my gx270, which had 40wire signal cable to the optical dive, and if so for subject computer I would suggest that all onboard drives should be converted to master/slave configuration
40wire ide ribbon cables can not configure cable select
I would surmise was not such an issue when the secondary hdd was a usb connected drive not using the primary IDE host bus adapter channel but installed as the IDE primary channel secondary drive1 (slave) has created i/o issues

[GrahamW2]

I evaluated the specs of your dimension 8200 and is of a similar era to my gx270, which had 40wire signal cable to the optical dive, and if so for subject computer I would suggest that all onboard drives should be converted to master/slave configuration
40wire ide ribbon cables can not configure cable select
I would surmise was not such an issue when the secondary hdd was a usb connected drive not using the primary IDE host bus adapter channel but installed as the IDE primary channel secondary drive1 (slave) has created i/o issues

OP here.

I've almost given up.

I've installed the HDD in the PC directly (not in a USB caddy) and deleted the partition, then created a new simple one (with a Quick Format). The PC seems to work without hanging for a reasonable period of time (hours). Then I tried copying from my old 250GB HDD (which was in a USB caddy). As before, it worked well ... for a while. However, after a bit of copying (perhaps 50GB or so) the PC hung again.

I tried to format the disk, but Windows wouldn't allow it. I deleted various files on the disk relating to the recycling bin (in fact I have now turned recyling off on this drive) and managed to get Windows to do a full checkdisk on reboot. Same outcome.

Then I deleted the Volume (again !) and created a new one ... this time with a FULL FORMAT (which took hours !). Same outcome.

FWIW the "hang" usually happens during the transfer of large files (e.g. 10MB+).

So I have a drive which doesn't want to work - even when wiped and formated.

I like your idea of Master/Slave jumper settings. I have always set my secondary drives to "Cable Select". I have run a 250GB HDD for many years in this PC with the "Cable Select" setting. I'm not sure why this has worked for for a 250GB drive but won't for a 500GB drive.

The next (last ?) thing I'll try is to switch the jumper from CS to Slave. Will I need to recreate a Partition and do a full Format again ?

Regards

Graham

[gerbil]

Because you don't say what wire-count your cables are, i have to ask - are they 40- or 80-wire? CS does not work with 40-wire cables, unless you modify them yourself. Even though CS was available i used M and S jumpers, placing the Master drive at the end of the cable. But placing of the drives does not matter if you have two on the cable - putting the Master on the end was only important if the Master was all that was connected [removes reflections from an otherwise hanging end].
If you go with M and S jumping then no, no other disk work is necessary, it simply allows the controller to identify the Master drive. CS does it via wire 28 on the 80-wire cable - it is only connected to the mb and end connectors. So with an 80 wire cable and two drives you tend to put the Master on the end, for CS to work.
It has occurred to me that once i experienced a failure on a mb in which the South Bridge [which chip on an Intel chipset board handles disk communications] would overheat or likewise fail when tasked with a read/write job of more than a few MB ... the system would just crash. It was not actually a heatsink problem, at least nothing which could be solved by normal chip cooling [supplied heatsink with adequate paste, well seated and unobstructed], but one which a replacement mb solved. Sneaked in under the warranty. It would only allow loading of the OS if I had the case side off and a room fan blasting in air from an open window.... a frosty morning gave the most reliable result.
Perhaps your newer drive is placing a greater heat load on the S-bridge?
Btw, a full format is rarely necessary. A quick format deletes the file table info and runs a chkdsk. A full format does all that plus a surface check of every block. Yow.

[tigerbright]

hi , no reformat required , you are just adjusting the hdd"s configuration so that they can be positively identified on the same channel by the system.

I would surmise by the decription of the issue and model era of both computer and hdd that default CS config of single drive0 on primary channel was tolerated but not 2 CS configured drives on the same channel with I suspect the optical drive not being CS configured nor supported as evidenzed by i suspect a non-cs supported 40wire cable

I suspect the optical drive uses a "rough feel" 40 wire ribbon cable so as such does not support CS - all drives attached onboard need to be either CS and using a fine feel" 80wire CS supported cable (the configuration is also the cable form factor) or all drives configured master/slave using a 80wire cable for the high speed hdd's and some say best for the optical drive as well (although 40wire is within spec for an optical drive)

[tigerbright]

Hi , in the mean time you can test your western digital (and maxtor drive)

western digital

http://support.wdc.com/product/download.asp?groupid=505&sid=3&lang=en

maxtor

http://www.seagate.com/ww/v/index.jsp?locale=en-US&name=seatools-win&vgnextoid=552bd20cacdec010VgnVCM100000dd04090aRCRD

[Batsfile]

Can XP support a HDD of 500GB ?

yes sir

sir it doesn't matter if you use even 2 to 3 HHD i one PC..
its because i use in my pc 3 HHD 1 5oogb Sata type,1 200gb IDE Type Cable, 1 80gb Sata in may OS Window Xp Service pack 2..
and I have only Pentium 4 and 1.5GB memory card Processor 2.8..but the speed of my cpu its not fast.but ok to use...and tested many years..

in that case sir even you have a different type of Hard Drive it is ok...
all you have to do is....

1.reformat your Hard disk in NTFS Files.. and remove partition then install your New OS..and Install your Mother board Driver....

[GrahamW2]

Because you don't say what wire-count your cables are, i have to ask - are they 40- or 80-wire?

It has occurred to me that once i experienced a failure on a mb in which the South Bridge [which chip on an Intel chipset board handles disk communications] would overheat or likewise fail when tasked with a read/write job of more than a few MB ...

I don't know anything about 40 v 80 wires. AFAIK I've only ever encountered one sort of IDE ribbon cable (and one sort of connector - which has about 40 pins). The cables in my PC have one connector at one end (mother board) and two at the other (well, technically one at the end and one just short of the end). I can only assume that they are 40 pin.

I reconfigured from CS on both drives (which worked with a 100MB C: drive and a 250GB secondary drive for YEARS without problem) to a Master ("with Slave present") setting for C: and a Slave for the secondary. Result, exactly the same. When faced with the heavier load of a large file (say 10MB+) the PC invariably hangs.

I'm loath to blame any chips. Aftera all, writing to the C: drive never causes a problem ... just writing to the secondary.

As an elimination test I disconnected the USB caddy and just tried copying from C: to the secondary drive. Same problem.

I even tried downloading a large file from the Internet to the secondary drive - same problem.

So the issue seems to be restricted to the actual writing of large data files to the secondary drive when it's inside the PC - whether the file comes from a USB caddy, another HDD or the Internet. The problem continues whether or not there is a USB caddy connected to the PC.

It's a puzzle.

Thanks for the links to the HDD test programs. My HDD passes the quick test. I ran the extended test for an hour ... but it wanted another five hours to finish. I'm not sure that I am prepared to put my PC out of use for the best part of a day - and I'm pretty sure that it would pass the test anyway. ISTM that the HDD works perfectly well in terms of all of the sectors (?) being healthy - the problem seems to be a "dynamic" one caused by the activity of writing a large file (high data throughput ?).

I think that I am just going to have to accept that this HDD won't work in my PC. The cost of a new 500GB PATA HDD is about £50 (which isn't much less than a terabyte+ SATA HD !) so I'm not inclined to shell out. Especially when there is a significant chance that the problem will recur anyway.

Perhaps I'll hang onto the 500GB HDD and just keep it in the caddy. I'll do a bit of "deep archiving" of really old backup stuff (like my digital photos pre-2005).

Thanks again

Regards

Graham

[tigerbright]

Hi , five hours seems excessive for running those vendor supplied maintenance tools, which would imply major problems with the tested drive such as bad sectors
I presume you are writing about the usb connected drive attached as channel0, drive1(slave)
I would suggest after confirming the correct cable installed , that you save all valued personal files to removable storage, disconnect all drives , then attach the problem former usb connected drive as a channel0 , drive0 as single master drive (disconnect the optical drive if using the diskette version tool, if not confirm is set for master)
Then run quick zero (if supported), then full zero to erase all data
this will also repair the structure, patch bad sectors etc
Then reinstall all drives as required, copy back desired personal files

(You can run HDtune "error scan" while in Windows which will indicate any bad sectors
consulting "health" will indicate other looming problems - yellow or red highlights)

[GrahamW2]

Hi , five hours seems excessive for running those vendor supplied maintenance tools, which would imply major problems with the tested drive such as bad sectors
I presume you are writing about the usb connected drive attached as channel0, drive1(slave)
I would suggest after confirming the correct cable installed , that you save all valued personal files to removable storage, disconnect all drives , then attach the problem former usb connected drive as a channel0 , drive0 as single master drive (disconnect the optical drive if using the diskette version tool, if not confirm is set for master)
Then run quick zero (if supported), then full zero to erase all data
this will also repair the structure, patch bad sectors etc
Then reinstall all drives as required, copy back desired personal files

(You can run HDtune "error scan" while in Windows which will indicate any bad sectors
consulting "health" will indicate other looming problems - yellow or red highlights)

The quick scan was quick. The fuller scan came with a message saying that it would take several hours, depending on the size of the drive. Given that this is a ten year old PC I don't think five hours is so long. A full disk defrag takes a similar amount of time (and that's just on the older 250GB drive, not the 500GB "problem" one).

I'm giving up on this now. I may not even bother to try and squeeze some use out of the 500GB drive in a USB caddy. I can live without it. Life's far too short to waste any more hours on trying to get a bit of use out of the HDD left over from a PVR upgrade. Its paid for itself having done several years in the PVR !

There's no need for me to get involved in the downside risk of this drive failing in the future. Heck, the PC is long overdue for renewal now anyway.

I'm moving on.

Thanks for all your help though.

Regards

Graham

[tigerbright]

do you want to sell that drive ?

[gerbil]

Okay, Graham... I'm going to guess that the drive electronics cannot handle their potential full speed when inside the box, compared with the data rate achievable with USB when inside the caddy. How does it perform when it is Master on one controller [cable] and the the C: drive is Master on the other? Your optical drive can be slave on either.