Is HP BD01865CC4 18.2 GB drive really a 36.7 GB Fujitsu MAP3367NC?

[Arie Bant]

Hello,

I recently aquired a HP BD01865CC4 18.2 GB 10K rpm Ultra 3 SCSI drive.
To my surprise the label reveals the drive is MAP3367NC, a Fujitsu 36.4
GB 10K rpm Ultra 320 drive.
I also have a HP BD03695CC8 36.4 GB 10K rpm Ultra 320 drive, which is a
rebranded Fujitsu MAP3367NP, only difference is the physical interface,
68-pin instead of SCA-2.

Does anyone know whether a HP BD01865CC4 drive can be made to behave
like the drive it originally was, i.e. a 36.4 GB 10K rpm Ultra 320 drive?

The firmware on both drives is the same, i.e. HPB6, which suggests it
cannot be simply a different firmware version.

I would like to try if it is possible to change the number of locical
blocks in the drives mode select data, but at the moment I only have a
MS Windows machine available with LVD SCSI HBA.
Does anyone know of a good free program to read the SCSI mode sense
data/write the mode select data under MS Windows? (XP)

TIA.

Arie Bant.

 

[Folkert Rienstra]

Hello,

I recently acquired a HP BD01865CC4 18.2 GB 10K rpm Ultra 3 SCSI drive.
To my surprise the label reveals the drive is MAP3367NC, a Fujitsu 36.4
GB 10K rpm Ultra 320 drive.
I also have a HP BD03695CC8 36.4 GB 10K rpm Ultra 320 drive, which is a
rebranded Fujitsu MAP3367NP, only difference is the physical interface,
68-pin instead of SCA-2.

Does anyone know whether a HP BD01865CC4 drive can be made to behave
like the drive it originally was, i.e. a 36.4 GB 10K rpm Ultra 320 drive?

Depends on how it was done, by Short Stroke or by Factory Low Level Format.
In the latter case it can not be changed.

The firmware on both drives is the same, i.e. HPB6, which suggests it
cannot be simply a different firmware version.

It depends on what capacity info was written to the platters that doesn't come
with the firmware. That cannot be changed.

I would like to try if it is possible to change the number of logical
blocks in the drives mode select data, but at the moment I only have a
MS Windows machine available with LVD SCSI HBA.

WD's SCSI bench has an option under utilities called 'Short stroke drive'.
http://support.wdc.com/download/index.asp?cxml=n&pid=24&swid=25
Click 'maximum capacity'.
Alternatively IBM/s Codeupdt can set the number of blocks in
Header and block descriptor under option 6 Mode Sense/Select
http://www.nu2.nu/scsitool/

Does anyone know of a good free program to read the SCSI mode sense
data/write the mode select data under MS Windows? (XP)

Above runs under W9x, don't know about XP.

 

[Arie Bant]

Depends on how it was done, by Short Stroke or by Factory Low Level Format.
In the latter case it can not be changed.




It depends on what capacity info was written to the platters that doesn't come
with the firmware. That cannot be changed.




WD's SCSI bench has an option under utilities called 'Short stroke drive'.
http://support.wdc.com/download/index.asp?cxml=n&pid=24&swid=25
Click 'maximum capacity'.
Alternatively IBM/s Codeupdt can set the number of blocks in
Header and block descriptor under option 6 Mode Sense/Select
http://www.nu2.nu/scsitool/




Above runs under W9x, don't know about XP.

The WD utility does not see the MAP3367Nx drives under MS Windows XP. I
assume this is caused by the fact that the aspi-driver does not see
either them or the Plextor CD-Writer. The aspi-driver does see the
other Compaq/Seagate drive and the Pioneer DVD-reader.
MS Windows XP SP2 does see all those units.

Using Bart's scsitool from a DOS-flop:

The numbers of cylinders (48122), heads (2), bytes/sector (512) and
sectors/track (936) are all the same on both drives according to the
mode sense parameters.
Yet the one MAP3367Nx (=18.2 GB HP BD01865CC4) reports 35.565.080 data
blocks, the other MAP3367Nx (=36.4 HP BD03695CC8) reports 71.132.000
data blocks.
I have tried to set the number of data blocks on the 18 GB drive the
same as on the other through the block descriptor parameters of the mode
select command, but the drive reports a "value out of range" error, as
it does for some smaller values I tried.

Any other ideas Folkert, Rod, Ron, or anybody else of course?

TIA.

Arie.

 

[Folkert Rienstra]

The WD utility does not see the MAP3367Nx drives under MS Windows XP. I
assume this is caused by the fact that the aspi-driver does not see
either them or the Plextor CD-Writer. The aspi-driver does see the
other Compaq/Seagate drive and the Pioneer DVD-reader.
MS Windows XP SP2 does see all those units.

Using Bart's scsitool from a DOS-flop:

The numbers of cylinders (48122), heads (2), bytes/sector (512) and
sectors/track (936) are all the same on both drives according to the
mode sense parameters.
Yet the one MAP3367Nx (=18.2 GB HP BD01865CC4) reports 35.565.080 data
blocks, the other MAP3367Nx (=36.4 HP BD03695CC8) reports 71.132.000
data blocks.
I have tried to set the number of data blocks on the 18 GB drive the
same as on the other through the block descriptor parameters of the mode
select command, but the drive reports a "value out of range" error, as
it does for some smaller values I tried.

Any other ideas Folkert,

Nope, that is the case where the recorded capacity on the platters been set as
18 GB. Only Fujitsu can change that. Well, if they do it the same way as IBM did.

Or, as seems te case with Seagate, the capacity is recorded in
the firmware itself and you need the 36 GB version of the firmware.
Wether the drive will accept the firmware or that the flash utility
will accept that the drive isn't reporting as Fujitsu, who is to say.

 

[Arie Bant]

Nope, that is the case where the recorded capacity on the platters beenset as
18 GB. Only Fujitsu can change that. Well, if they do it the same way as IBM did.

Or, as seems te case with Seagate, the capacity is recorded in
the firmware itself and you need the 36 GB version of the firmware.
Wether the drive will accept the firmware or that the flash utility
will accept that the drive isn't reporting as Fujitsu, who is to say.

I could try flashing the 18GB drive with the firmware of the 36GB one.
Both have a HP number and a Compaq identifier, both are recognised by a
Fujitsu disk test program.
I´ll think a bout that. No time until the weekend.

Cheers,

Arie.

 

[Folkert Rienstra]

A dozen of added empty lines removed. Surely that doesn't need to be.

I could try flashing the 18GB drive with the firmware of the 36GB one.

That assumes that you can extract it. That usually isn't possible.
If it *is* possible I would extract from both and compare them first
before possibly flashing the 18GB to death and then find out that they
were not even different. Or find out from Fujitsu, either by asking the
question or by ordering the 18GB and 36GB version of the firmware
and find out for yourself by comparing what you get.
This won't be an easy call since your drives also differ in interface.

 

[Rolf Blom]

Hello,

I recently aquired a HP BD01865CC4 18.2 GB 10K rpm Ultra 3 SCSI drive.
To my surprise the label reveals the drive is MAP3367NC, a Fujitsu 36.4
GB 10K rpm Ultra 320 drive.
I also have a HP BD03695CC8 36.4 GB 10K rpm Ultra 320 drive, which is a
rebranded Fujitsu MAP3367NP, only difference is the physical interface,
68-pin instead of SCA-2.

Does anyone know whether a HP BD01865CC4 drive can be made to behave
like the drive it originally was, i.e. a 36.4 GB 10K rpm Ultra 320 drive?

The firmware on both drives is the same, i.e. HPB6, which suggests it
cannot be simply a different firmware version.

I would like to try if it is possible to change the number of locical
blocks in the drives mode select data, but at the moment I only have a
MS Windows machine available with LVD SCSI HBA.
Does anyone know of a good free program to read the SCSI mode sense
data/write the mode select data under MS Windows? (XP)

TIA.

Arie Bant.

I think you need a tool from Fujitsu to change the number of logical
blocks on the drive. The drive could have been reduced in size to fit as
a sparepart in a hw raid system where the original drives were smaller.

/Rolf

 

[Arie Bant]

A dozen of added empty lines removed. Surely that doesn't need to be.




That assumes that you can extract it. That usually isn't possible.
If it *is* possible I would extract from both and compare them first
before possibly flashing the 18GB to death and then find out that they
were not even different. Or find out from Fujitsu, either by asking the
question or by ordering the 18GB and 36GB version of the firmware
and find out for yourself by comparing what you get.
This won't be an easy call since your drives also differ in interface.

Thanks, thinking about it I had already decided that comparing them is
the right way to go before flashing comes into it.
Fujitsu is not very helpful, it is an OEM-drive and all support has to
come from HP/Compaq.

 

[Arie Bant]

I think you need a tool from Fujitsu to change the number of logical
blocks on the drive. The drive could have been reduced in size to fit as
a sparepart in a hw raid system where the original drives were smaller.

/Rolf

That is indeed what it seems like. I did not know the drive was a
rebadged MAP3367NC before I actually got it. Fujitsu does not want to
play ball though. Any support I want has to come from HP/Compaq, who of
course claim that it is a HP BD01865CC4 18.2 GB 10K rpm Ultra 3 SCSI
drive and it cannot be changed.

Hence my posting, to see if someone had come accross these drives before
and, drawing the same conclusion, had found out if it can be done and how.

I have a tool to change the number of logical blocks in the mode select
pages but I cannot increase the number, just decrease it. All that is
normal for any SCSI drive. There must be a catch somewhere, possibly in
the firmware, but as I said before, they have the same version number.

 

[Folkert Rienstra]

That is indeed what it seems like. I did not know the drive was a
rebadged MAP3367NC before I actually got it. Fujitsu does not want to
play ball though. Any support I want has to come from HP/Compaq, who of
course claim that it is a HP BD01865CC4 18.2 GB 10K rpm Ultra 3 SCSI
drive and it cannot be changed.

Hence my posting, to see if someone had come across these drives before
and, drawing the same conclusion, had found out if it can be done and how.

I have a tool to change the number of logical blocks in the mode select
pages but I cannot increase the number, just decrease it.

Which means that the (reduced) capacity is recorded, either on the platters
or hard coded in the firmware.

All that is normal for any SCSI drive.

It also means that the drive's capacity wasn't reduced by means of
short stroke but is recorded permanently.

There must be a catch somewhere, possibly in the firmware,

For you, hopefully. Otherwise you can forget it.

but as I said before, they have the same version number.

And as I said before, Seagate has the same version different firm-
ware that is to be specifically ordered for the right drive capacity.
The model and capacity is hard coded in the drive firmware.

With IBM though, capacity and Model are recorded on the platters
and all capacities work with the same firmware image.

 

[Folkert Rienstra]

Please do something about it.

 

[Folkert Rienstra]

Just another thought:
Instead of flashing you can exchange the firmware by exchanging
the electronics boards instead. If the capacity is recorded in the
firmware rather than on the platters the 36GB board should be able
to see it at it's full capacity, either directly or after a mode select.
If no change, then the capacity must have been recorded on the platters.

 

[Rod Speed]

Go and **** yourself

Please do something about it.

 

[Arie Bant]

A dozen of added empty lines removed. Surely that doesn't need to be.


On 29/06/05 23:20, Folkert Rienstra wrote:


On 28/06/05 23:16, Folkert Rienstra wrote

"Arie Bant" <[email protected]> wrote in message news:[email protected]

[snip]

I could try flashing the 18GB drive with the firmware of the 36GB one.

That assumes that you can extract it. That usually isn't possible.
If it *is* possible I would extract from both and compare them first
before possibly flashing the 18GB to death and then find out that they
were not even different. Or find out from Fujitsu, either by asking the
question or by ordering the 18GB and 36GB version of the firmware
and find out for yourself by comparing what you get.
This won't be an easy call since your drives also differ in interface.

Just another thought:
Instead of flashing you can exchange the firmware by exchanging
the electronics boards instead. If the capacity is recorded in the
firmware rather than on the platters the 36GB board should be able
to see it at it's full capacity, either directly or after a mode select.
If no change, then the capacity must have been recorded on the platters.

Thanks Folkert,

The thought had crossed my mind but I rejected it because these are nice
new drives, nothing wrong with them. I have tried the board-switch
method before with two older, but almost identical Barracuda's. One,
68-pin interface, seemed to have a drive problem, the other with SCA
interface worked fine. I tried the working one with the 68-pin board.
Not only did it not work, but when I put the boards back, neither drive
ever worked again. I am resonably good with electronics, often repaired
boards and know about static etc. so I do not assume I damaged the good
board or drive in the process.

Hence I am somewhat reluctant to try that trick.

Arie.

 

[Folkert Rienstra]

Still lots of excess empty lines

On 30/06/05 14:33, Folkert Rienstra wrote:
A dozen of added empty lines removed. Surely that doesn't need to be.
On 29/06/05 23:20, Folkert Rienstra wrote:
On 28/06/05 23:16, Folkert Rienstra wrote
[snip]

I could try flashing the 18GB drive with the firmware of the 36GB one.

That assumes that you can extract it. That usually isn't possible.
If it *is* possible I would extract from both and compare them first
before possibly flashing the 18GB to death and then find out that they
were not even different. Or find out from Fujitsu, either by asking the
question or by ordering the 18GB and 36GB version of the firmware
and find out for yourself by comparing what you get.
This won't be an easy call since your drives also differ in interface.

Just another thought:
Instead of flashing you can exchange the firmware by exchanging
the electronics boards instead. If the capacity is recorded in the
firmware rather than on the platters the 36GB board should be able
to see it at it's full capacity, either directly or after a mode select.
If no change, then the capacity must have been recorded on the platters.

Thanks Folkert,

The thought had crossed my mind but I rejected it because these are nice
new drives, nothing wrong with them.
I have tried the board-switch method before with two older, but almost
identical Barracuda's. One, 68-pin interface, seemed to have a drive
problem, the other with SCA interface worked fine.
I tried the working one with the 68-pin board. Not only did it not work,
but when I put the boards back, neither drive ever worked again.

That's odd. I don't have the same experience.
I have had no problems to get one working Barracuda 18LP ST39175LW
out of 2 bad ones. I'll have to say though that both were 68-pin and they
had the same FW rev.
I didn't have luck with a 9GB Cheetah ST39103LC, not sure anymore whether
the good one was LC too, I think it was 68-pin. The good one was an 18GB.
To complicate matters the 9GB is IBM badged. It didn't work.
The 9GB called itself the 18GB but didn't work anyway which may not be
all that surprising and the 18GB now called itself the 9GB but appeared
to come up well. However first access got it into a frenzy of clicks.
Fortunately the 18GB still worked afterwards. Since I only had it on loan
for inspection/checkout it was returned.

I am reasonably good with electronics, often repaired boards and know
about static etc. so I do not assume I damaged the good board or drive
in the process.

Hence I am somewhat reluctant to try that trick.

Yep, it is hair raising stuff, especially if you have no warranty to fall back on.

 

[Rolf Blom]

Which means that the (reduced) capacity is recorded, either on the platters
or hard coded in the firmware.




It also means that the drive's capacity wasn't reduced by means of
short stroke but is recorded permanently.




For you, hopefully. Otherwise you can forget it.




And as I said before, Seagate has the same version different firm-
ware that is to be specifically ordered for the right drive capacity.
The model and capacity is hard coded in the drive firmware.

With IBM though, capacity and Model are recorded on the platters
and all capacities work with the same firmware image.

We did use 'scsitoolbox' here (not free, and too expensive for a single
use, but your Bart's toolbox would probably be as good for the purpose.)

In the scsitoolbox, the sequence was:

Mode select

offset xx yyyyyyy (where yyyyyyy is number of logical blocks)
(the offset number was 04 with our Fujitsu drive, could be different)

save parameters
low-level format

both the 'save parameters' and 'low-level format'
in the same run was neccessary, or the drive didn't change.

/Rolf

 

[Folkert Rienstra]

And yet another one with that POS newsclient.


[Since you don't ever respond to the quoted text I have snipped it for you]

We did use 'scsitoolbox' here (not free, and too expensive for a single
use, but your Bart's toolbox would probably be as good for the purpose.)

As I said before, it can't (do you ever read the messages that you respond to)?

In the scsitoolbox, the sequence was:

offset xx yyyyyyy (where yyyyyyy is number of logical blocks)
(the offset number was 04 with our Fujitsu drive, could be different)

both the 'save parameters' and 'low-level format'
in the same run was neccessary, or the drive didn't change.

Must have been a quirk of SCSIToolbox.
Setting the top LBA has nothing to do with Low Level Format.

 

[chrisv]

Go and **** yourself

Well, that's not very nice, Ron^Hd...

 

[Arie Bant]

We did use 'scsitoolbox' here (not free, and too expensive for a single
use, but your Bart's toolbox would probably be as good for the purpose.)

In the scsitoolbox, the sequence was:

offset xx yyyyyyy (where yyyyyyy is number of logical blocks)
(the offset number was 04 with our Fujitsu drive, could be different)

both the 'save parameters' and 'low-level format'
in the same run was neccessary, or the drive didn't change.

/Rolf

Thanks Rolf,

Sounds much the same as I did. Except that in this case the drive
returned an error that means "logical block number out of range".
I did format the drive anyway on the off-chance it might work, but it
made no difference at all.

It seems there is an absolute maximum value hidden somewhere either in
the firmware or in the non-user space on the platter, where it keeps all
the other mode select data.
When I find the right tool and the time I will copy the firmware of both
drives to files and compare those, see what that brings.

Of course, if someone new of a non-documented mode select page for the
Fujitsu MAP3367Nx drives that could be used to change the absolute
maximum or some other way to achieve the same result, we would be back
in business.

Otherwise I will have to force myself to consider how much time I can
reasonably waste on a probably fruitless chase to save 18 GB storage
space on a drive I bought as an 18 GB in the first place.
I just like technical puzzles and once I have the bit between the teeth...

Arie.

 

[Rolf Blom]

On 2005-07-01 15:43, Arie Bant wrote:

-snip-

Otherwise I will have to force myself to consider how much time I can
reasonably waste on a probably fruitless chase to save 18 GB storage
space on a drive I bought as an 18 GB in the first place.
I just like technical puzzles and once I have the bit between the teeth...

Arie.

One thing to try is a lower than max value, maybe max-1, it could be
different tools have counting start at 0 instead of at 1.

Sometimes it's more fun to fiddle with things for a bit, than to have
the results readily served
Still 18GB is better than a broken drive, so don't try too hard.

/Rolf