I took apart a Mini iPod.

  • Thread starter Thread starter Matt Kim
  • Start date Start date
i decided to do what i would do if i were designing this drive into something for work: contact the applications engineer for the microdrive group at hitachi.

he says that the microdrive for the iPod is IDE-only, and was designed this way specifically so that it would NOT work in 99% of cameras out there.

this might be disinformation, but only in the sense that it may not be hard-wired into the device to be in IDE mode. i can understand hitachi's desire to do this so it does not cannibalize camera CF sales.

having said that, reading the specs on line i could not find any way to switch the mode of the hard disk and make it "stick" across power cycles. according to the specs as you power on the drive you change the state of certain pins in certain ways to cause it to go into one mode or the other. if the camera does not do this, or the drive does not respond to the command, you are hosed.

so i think we're probably stuck here. he says creative will soon be using these disks as well.
 
I definately think you are right about that... but I would not give up just yet, when I called Hitachi, the tech said there were the same exact drives. ? There is another post around here concerning a small program you can write to change the mode.
so i think we're probably stuck here. he says creative will soon be
using these disks as well.
--
Regards,
Matt

http://www.photo.net/photodb/user?user_id=386029
http://www.pbase.com/mattjkphoto/root
http://www.wherearethetoonsnow.com/
 
well again according to the spec there does not seem to be any non-volatile memory/register that holds the mode. but the spec could be incomplete. the mode apparently is set on each power on by the host computer/camera/whatever.

the guy i spoke with is the senior apps engineer for hitachi microdrives, so i'm sure he knows what he is talking about. again though its not in hitachi's interest to disclose all the information they have so what you are talking about may still be possible. he said that the drive does not carry the CF logo because it does not conform to the CF spec -- the memory and i/o modes are missing.

as an aside it may still be possible to put a flash card in the iPod mini, if you want a jog-proof unit. on the theory that the iPod expects an IDE mode device and does not set the right pins, apple may have left the important pin unconnected (pin 9 on the connector should be grounded). this means you would have to do some rework to solder one pin on the connector to ground and that would be difficiult to say the least :) that stuff is tiny.
 
Thanks for taking the time to find all this info out... it a bummer, but at least now we know :-/

I wonder what makes the Muvo's drive work... does IT have the CF logo? The P/N of the Muvo Drive and the Mini Drive are different (but the model numbers are the same)... maybe that tells us something?

I wonder if all the preorders for the next batch of muvos (the ones that dell says will ship in May) will have the same drive as the mini.
  • leo
he said that the drive does not carry the CF logo because
it does not conform to the CF spec -- the memory and i/o modes are
missing.
 
here is my guess:

the muvo drive works because it is a stock microdrive, which fully supports the CF standard (ide, memory and i/o modes). hitachi didnt care about using the stock drive there because they probably figured that creative's volumes wouldnt be that high, or they were not ready with the hacked drive in time for creative's plans, or maybe creative's asics don't speak IDE mode (yet)

along comes apple with 10x the volume of creative. hitachi is now more motivated to do a 'crippled' drive because of what we are trying to do here. and here we are :(

i'd wager that the next big batch of muvo2's will use this same hard drive.

its amusing how many people on other sites insist that the iPod mini drive will work or should work. i'm sure they havent tried it. we have 3 confirmed failures now that i know of. most people think its just a matter of formatting the card properly in the PC but everything i know now tells me this has nothing to do with the problem :(
I wonder what makes the Muvo's drive work... does IT have the CF
logo? The P/N of the Muvo Drive and the Mini Drive are different
(but the model numbers are the same)... maybe that tells us
something?

I wonder if all the preorders for the next batch of muvos (the ones
that dell says will ship in May) will have the same drive as the
mini.
  • leo
he said that the drive does not carry the CF logo because
it does not conform to the CF spec -- the memory and i/o modes are
missing.
 
Get a window version of Mini ipod and do it again
I took one apart tonight. It was quite easy and didn't do any
cosmetic damage. The best thing to do is get a popsicle stick and a
couple drops of hot glue, and stick it to the white plastic ends.
Wait for it to dry and then slowly lift until the covers come off.
The rest is easy.

To remove the hot glue, just heat it up with a hair dryer and then
peel the rest of the glue away.

Bad news is, I could not get the 4gb drive to work, or a 1gb
microdrive to work in the iPod. :(

Matt
 
here is my guess:
And did any of you guys that have the ipod buy a MuVo2? Did any body try to firmware the ipod drive in a MuVo? Do you need a link to Creative?

http://us.creative.com/shop/shopcategory.asp?category=3

And the firmware is posted in some threads on Canon EOS-10D/D60/D30 forum, or you can get it at the creative Japan site if you hurry before they take it down like they did on the USA site. Don't sit on your hands too long, Creative takes the Muvo off from time to time.

Dennis
the muvo drive works because it is a stock microdrive, which fully
supports the CF standard (ide, memory and i/o modes). hitachi didnt
care about using the stock drive there because they probably
figured that creative's volumes wouldnt be that high, or they were
not ready with the hacked drive in time for creative's plans, or
maybe creative's asics don't speak IDE mode (yet)

along comes apple with 10x the volume of creative. hitachi is now
more motivated to do a 'crippled' drive because of what we are
trying to do here. and here we are :(

i'd wager that the next big batch of muvo2's will use this same
hard drive.

its amusing how many people on other sites insist that the iPod
mini drive will work or should work. i'm sure they havent tried it.
we have 3 confirmed failures now that i know of. most people think
its just a matter of formatting the card properly in the PC but
everything i know now tells me this has nothing to do with the
problem :(
I wonder what makes the Muvo's drive work... does IT have the CF
logo? The P/N of the Muvo Drive and the Mini Drive are different
(but the model numbers are the same)... maybe that tells us
something?

I wonder if all the preorders for the next batch of muvos (the ones
that dell says will ship in May) will have the same drive as the
mini.
  • leo
he said that the drive does not carry the CF logo because
it does not conform to the CF spec -- the memory and i/o modes are
missing.
 
if you are referring to the muvo firmware that is not going to do much for us, its just a pattern of bytes written to the disk that the camera or ipod does not know what to do with.

on the other hand if you are referring to the microdrive firmware, then there might be hope. the IDE only cripple mode might be fixable with new drive firmware.

can you post some links to this firmware?
here is my guess:
And did any of you guys that have the ipod buy a MuVo2? Did any
body try to firmware the ipod drive in a MuVo? Do you need a link
to Creative?

http://us.creative.com/shop/shopcategory.asp?category=3

And the firmware is posted in some threads on Canon EOS-10D/D60/D30
forum, or you can get it at the creative Japan site if you hurry
before they take it down like they did on the USA site. Don't sit
on your hands too long, Creative takes the Muvo off from time to
time.

Dennis
the muvo drive works because it is a stock microdrive, which fully
supports the CF standard (ide, memory and i/o modes). hitachi didnt
care about using the stock drive there because they probably
figured that creative's volumes wouldnt be that high, or they were
not ready with the hacked drive in time for creative's plans, or
maybe creative's asics don't speak IDE mode (yet)

along comes apple with 10x the volume of creative. hitachi is now
more motivated to do a 'crippled' drive because of what we are
trying to do here. and here we are :(

i'd wager that the next big batch of muvo2's will use this same
hard drive.

its amusing how many people on other sites insist that the iPod
mini drive will work or should work. i'm sure they havent tried it.
we have 3 confirmed failures now that i know of. most people think
its just a matter of formatting the card properly in the PC but
everything i know now tells me this has nothing to do with the
problem :(
I wonder what makes the Muvo's drive work... does IT have the CF
logo? The P/N of the Muvo Drive and the Mini Drive are different
(but the model numbers are the same)... maybe that tells us
something?

I wonder if all the preorders for the next batch of muvos (the ones
that dell says will ship in May) will have the same drive as the
mini.
  • leo
he said that the drive does not carry the CF logo because
it does not conform to the CF spec -- the memory and i/o modes are
missing.
 
Get a window version of Mini ipod and do it again
So much misinformation out there.. there is no Mac vs. Windows version of the iPod Mini or even the regular iPod for that matter.

The iPod Mini includes both USB2 and Firewire cables.

The regular iPod comes formatted with a Mac filesystem, but the Update/Restore tool will reformat it for Win32 on a Windows machine. The Mac can still read it that way.

I don't know what filesystem is the default on iPod Mini, but it's probably FAT32 if they were selling it as Windows-compatible before the update tool was posted.

Besides, your camera doesn't care what was on there before if you "Format Card". The fact that NOBODY has claimed to get the iPod Mini drive working in a camera gives pretty strong evidence that it's not the same as the Muvo2 drive. Whether it can be changed.. that's the question.

So far the idea that makes the most sense is that the drive is lacking some CF functionality, and operates only in the one mode (IDE?) that the iPod requires.
 
thanks for the reply to this knucklehead. i couldnt bring myself to do it :)

you know, when apple started making windows iPods i figured that since OSX handles FAT partitions that they would just start making all iPods with FAT-formatted drives. but maybe OS9 does not support FAT and so the different models...

anyway my mini was purchased at Fry's on tuesday. out of the box it has a mac partition table and HFS+ partition, possibly even journaled. i didnt look in the box but i assume that there was a windows CD that probably had the software that they just posted on the web to transform it to FAT for windows...
I don't know what filesystem is the default on iPod Mini, but it's
probably FAT32 if they were selling it as Windows-compatible before
the update tool was posted.
 
Check this out.

http://forums.dpreview.com/forums/read.asp?forum=1019&message=7800231

Dennis
on the other hand if you are referring to the microdrive firmware,
then there might be hope. the IDE only cripple mode might be
fixable with new drive firmware.

can you post some links to this firmware?
here is my guess:
And did any of you guys that have the ipod buy a MuVo2? Did any
body try to firmware the ipod drive in a MuVo? Do you need a link
to Creative?

http://us.creative.com/shop/shopcategory.asp?category=3

And the firmware is posted in some threads on Canon EOS-10D/D60/D30
forum, or you can get it at the creative Japan site if you hurry
before they take it down like they did on the USA site. Don't sit
on your hands too long, Creative takes the Muvo off from time to
time.

Dennis
the muvo drive works because it is a stock microdrive, which fully
supports the CF standard (ide, memory and i/o modes). hitachi didnt
care about using the stock drive there because they probably
figured that creative's volumes wouldnt be that high, or they were
not ready with the hacked drive in time for creative's plans, or
maybe creative's asics don't speak IDE mode (yet)

along comes apple with 10x the volume of creative. hitachi is now
more motivated to do a 'crippled' drive because of what we are
trying to do here. and here we are :(

i'd wager that the next big batch of muvo2's will use this same
hard drive.

its amusing how many people on other sites insist that the iPod
mini drive will work or should work. i'm sure they havent tried it.
we have 3 confirmed failures now that i know of. most people think
its just a matter of formatting the card properly in the PC but
everything i know now tells me this has nothing to do with the
problem :(
I wonder what makes the Muvo's drive work... does IT have the CF
logo? The P/N of the Muvo Drive and the Mini Drive are different
(but the model numbers are the same)... maybe that tells us
something?

I wonder if all the preorders for the next batch of muvos (the ones
that dell says will ship in May) will have the same drive as the
mini.
  • leo
he said that the drive does not carry the CF logo because
it does not conform to the CF spec -- the memory and i/o modes are
missing.
 
thanks, but again i'm not sure how muvo firmware will help us, unless it also upgrades the disk firmware.

anyway along these lines i read the spec carefully for the drive and some portion of the buffer memory on board the drive contains the drive's firmware. there are commands to write to the buffer but they appear to be for writing data to the drive. and if that firmware is writable, and it is a flash memory, then there should be commands to erase the firmware and i don't see that either.

i think this is our best hope at this point, but i'm sure hitachi is not going to be helpful in this matter :)
 
I don't know whether this has your interest and I hope it isn't redundant information, but anyway it may of some use if you continue your hacking quest on the miniPod...

A few years ago in my old job I wrote device drivers for CF cards in some instrumentation devices. I have actually programmed source code to identify and access a CF disk so I have been hacking on ceveral of these cards from various vendors.

The CF standard is based on two parts: PCMCIA (aka. PC-CARD) and IDE.

Lets look at the PCMCIA part first: Since this standard is designed to support many types of units (such as modems, I/O ports, network adaptors, disks ...) it has a way of storing information describing a given PCMCIA card and how the host system should configure the interface. This information is stored in 'attribute memory' (typically a small seperate ROM in the card) as socalled 'CIS tuples' that has a specific data structure layout to hold the information. There is defined a multitude of these tuples to support all the different PCMCIA card types and each card holds a suitable subset of these to tell the host system how to operate the card. The information may device specifications (vendor ID, serial number), type specifications (modem parameters or whatever), interface specifications (address space, interrupt lines, etc), or power specifications (expected voltage and max current).

When a PCMCIA card is detected in a host system, the host reads the CIS tuples and establishes the relevant support. Since a CF card does exactly the same you'll always be able to put your CF card into an adaptor and plug it into a PCMCIA slot.

But CF is also defined to comply with the IDE hard drive standard. That means that a host system that KNOWS that it will only be dealing with CF cards could ignore all the CIS information because all the interface and driver stuff is stated by IDE (according to CF). Therefore you'll see that all CF cards have almost identical CIS information (i.e., corresponding to IDE).

So what can happen to make a host reject a CF card? Well, obviously you can put some 'invalid' stuff into the CIS information since most hosts will pay at least some attention to this to prevent an 'accident' if someone should put, say, a CF modem in there. This is a very easy fix for the vendor - it is just a ROM that should have a slightly different content.

Another possibility is to change the IDE interface that also provide the host with some configuration data (although on a mush smaller scale).

If you want to look into this with a specific card you'll need some CIS reading and decoding tools and some IDE verification tools.

I hope some of you find this information usefull or at least slightly interesting.

Good luck with the hacking!

H-Jessen
 
so do you know of any linux-based tools for this stuff?
I don't know whether this has your interest and I hope it isn't
redundant information, but anyway it may of some use if you
continue your hacking quest on the miniPod...

[snip]

If you want to look into this with a specific card you'll need some
CIS reading and decoding tools and some IDE verification tools.

I hope some of you find this information usefull or at least
slightly interesting.

Good luck with the hacking!

H-Jessen
 
try:
pccardc dumpcis

See if that shows all the tuples and such. I think you're on the right track.

If you have any PCCard driver package installed, you should have enough there to probe the card on insertion. What does dmesg show for kernel output when you insert the iPod drive?

Make sure you're using a passive PCMCIA card adapter, and not a 'smart' one like the new Delkin Cardbus controller.

The advanced adapters have drive controllers in them, while the cheaper $9 adapters are simply passive connector adapters. Those let the laptop talk directly to the card.
 
to be honest i never put the microdrive in my PC laptop running linux. i did put it in my powerbook running 10.3, using an IBM PC card microdrive adapter. the mac reported unknown vendor/unknown card. from what i have gathered if you plug in a CF card into a PCMCIA slot then more than likely the host will want to communicate with it using "i/o mode". if the iPod microdrive lacks this mode then that is consistent with what the powerbook said.

i have since given up and put the iPod back together... i dont really want to take it apart again...
try:
pccardc dumpcis

See if that shows all the tuples and such. I think you're on the
right track.

If you have any PCCard driver package installed, you should have
enough there to probe the card on insertion. What does dmesg show
for kernel output when you insert the iPod drive?

Make sure you're using a passive PCMCIA card adapter, and not a
'smart' one like the new Delkin Cardbus controller.

The advanced adapters have drive controllers in them, while the
cheaper $9 adapters are simply passive connector adapters. Those
let the laptop talk directly to the card.
 
thanks, but again i'm not sure how muvo firmware will help us,
unless it also upgrades the disk firmware.
Well that's what I am getting at. So visualize this. I get the Muvo2 and turn it on and make sure everything works as it should. I didn't have any songs to load onto it but it brought up the play list and all of the other menu items or whatever they are called. I take out the 4gig, put the 256 Sandisk in and it gives me a firmware error. So I'm thinking, hmmmm something on the 4gig went with it and I need it to get this thing working again because everybody else that has a MuVo2 and put a 256 card in got theirs to work. But that little piece of magic was the firmware. Once i found the firmware, the puter recognized the MuVo and flashed it up or whatever it does. And it works fine. I also didn't see anything on the 4gig when I put it in the card reader to check the contents. Dunno, but heh, you guys will never try till one of you tries it.
anyway along these lines i read the spec carefully for the drive
and some portion of the buffer memory on board the drive contains
the drive's firmware. there are commands to write to the buffer but
they appear to be for writing data to the drive. and if that
firmware is writable, and it is a flash memory, then there should
be commands to erase the firmware and i don't see that either.
So it is the same as when we upgrade the firmware on our 10D or whatever there is a firmware update for. It erases the old firmware and installs the new. Huh, what do you think??
i think this is our best hope at this point, but i'm sure hitachi
is not going to be helpful in this matter :)
Of course not, they are trying to get past the embarrasment of how much some people had to pay for the retail 4gig microdrive. As they already said before, the drive is the same, they only gave the ipod part numer a different number so they would know where it came from. And you want to know something else? I was looking at the back side of my 1gig IBM microdrive that I paid 250 bucks for and the only thing missing from the group of trademarks. The MIC trademark or icon is not on the 4gig. Also the drive is rotated 180 degrees from the 1gig in relation to the socket. The slides on the sides are made of plastic on the 4gig where they are a milled metal on the 1gig. I'll bet it's just as good a drive as the 1gig is, you guys just haven't found the magic that makes it work for the digicams. Maybe that bit of magic is the MuVo2 firmware that has nothing to do with the format of the drive, unless it has to be there in order for it to format. It's like a bootable partition. You can't install a Windows OS unless there is a bootable partition. I'm thinking this is why the ipod drives will not format.
 
yes, so let me translate what you have said:

"I put a blank 120GB hard drive in my PC and it didnt boot windows. so then i installed windows XP on that 120GB disk and the machine booted windows! so the windows XP installer must have changed the firmware on the disk so that whatever windows was already on that 120GB disk was enabled to boot!"

of course you can install windows without a bootable partition on your hard disk! you boot from the CDROM which was pressed with a bootable partition!

the muvo needs firmware on the flash or microdrive or whatever, so that it can boot into its own operating system. this is what happens when you run creative's program, or apple's iPod mini firmware installer for that matter -- the firmware for the muvo/ipod is written to the disk in a place where the muvo/ipod expects to find it.

there is some outside chance that the creative's program re-flashes the hard disk firmware itself, but that does not seem very likely to me.

again, i don't see why the hitachi applications engineer would lie about this. its logical that the iPod mini expects an ATA interface on the hard disk, since the previous generation iPod hard disk is almost certainly ATA.

i dont see how hitachi's retail pricing for this disk is an "embarrasment" to them. they are free to negotiate whatever OEM deals they see fit with apple, creative or anyone else. they are free to charge whatever they want at retail for this product. and because of this it would be in their best interest to change the firmware on the OEM disks such that what we are trying to do here is made impossible or at least very difficult. there is such a thing as economies of scale here and they may be able to make money selling 1,000,000 drives at $50; if they thought that they could only sell 100,000, they might charge $500...

there is one bit of information that will end this debate forever. there is a status register in the disk which shows the version of the firmware on board the disk. if we can get a muvo2 disk and an iPod disk and dump that register for both disks, we can see if the firmware is actually different.

the 'bit of magic' we are missing is the drive firmware. or it could be. its possible that they have modified the drives in some other way to make it physically impossible for the drive to go into Memory or I/O interface mode. or it is possible that the firmware in the microdrive is not 'flashable' and is instead in a ROM, in which case we are stuck.
thanks, but again i'm not sure how muvo firmware will help us,
unless it also upgrades the disk firmware.
Well that's what I am getting at. So visualize this. I get the
Muvo2 and turn it on and make sure everything works as it should.
I didn't have any songs to load onto it but it brought up the play
list and all of the other menu items or whatever they are called.
I take out the 4gig, put the 256 Sandisk in and it gives me a
firmware error. So I'm thinking, hmmmm something on the 4gig went
with it and I need it to get this thing working again because
everybody else that has a MuVo2 and put a 256 card in got theirs to
work. But that little piece of magic was the firmware. Once i
found the firmware, the puter recognized the MuVo and flashed it up
or whatever it does. And it works fine. I also didn't see
anything on the 4gig when I put it in the card reader to check the
contents. Dunno, but heh, you guys will never try till one of you
tries it.
anyway along these lines i read the spec carefully for the drive
and some portion of the buffer memory on board the drive contains
the drive's firmware. there are commands to write to the buffer but
they appear to be for writing data to the drive. and if that
firmware is writable, and it is a flash memory, then there should
be commands to erase the firmware and i don't see that either.
So it is the same as when we upgrade the firmware on our 10D or
whatever there is a firmware update for. It erases the old
firmware and installs the new. Huh, what do you think??
i think this is our best hope at this point, but i'm sure hitachi
is not going to be helpful in this matter :)
Of course not, they are trying to get past the embarrasment of how
much some people had to pay for the retail 4gig microdrive. As
they already said before, the drive is the same, they only gave the
ipod part numer a different number so they would know where it came
from. And you want to know something else? I was looking at the
back side of my 1gig IBM microdrive that I paid 250 bucks for and
the only thing missing from the group of trademarks. The MIC
trademark or icon is not on the 4gig. Also the drive is rotated
180 degrees from the 1gig in relation to the socket. The slides on
the sides are made of plastic on the 4gig where they are a milled
metal on the 1gig. I'll bet it's just as good a drive as the 1gig
is, you guys just haven't found the magic that makes it work for
the digicams. Maybe that bit of magic is the MuVo2 firmware that
has nothing to do with the format of the drive, unless it has to be
there in order for it to format. It's like a bootable partition.
You can't install a Windows OS unless there is a bootable
partition. I'm thinking this is why the ipod drives will not
format.
 
Any update?
i have since given up and put the iPod back together... i dont
really want to take it apart again...
try:
pccardc dumpcis

See if that shows all the tuples and such. I think you're on the
right track.

If you have any PCCard driver package installed, you should have
enough there to probe the card on insertion. What does dmesg show
for kernel output when you insert the iPod drive?

Make sure you're using a passive PCMCIA card adapter, and not a
'smart' one like the new Delkin Cardbus controller.

The advanced adapters have drive controllers in them, while the
cheaper $9 adapters are simply passive connector adapters. Those
let the laptop talk directly to the card.
 

Keyboard shortcuts

Back
Top