Amiga HDD to WinUAE

[NoX1911]

Hallo,
i have a Quantum LP52S (52MB) amiga scsi harddrive connected to my PC (XP). The harddisk shows up in device manager but not in winuae (add harddrive) nor in 'disk management'. Any attempt to access the device with winhex (physical access) or Acronis Trueimage CD (linux) doesn't work either (empty or inaccessible). The harddisk shows up at the SCSI controller bootscreen (PC) with some kind of delay. Maybe the controller even has some problems detecting it properly. The hdd responds to 'SCSI reset' command (clicking, upspin) during windows boot and bios bootscreen. This should be normal i think. I've seen similar behavior on my scsi cd-rom's before.



Regarding the way its meant to be done... should the drive show up in winuae or is it generally impossible to access/mount real amiga hdds (and 'add harddrive' isn't meant for real amiga devices)?

If i remember correctly the hdd didn't even start up the last time i tested it on my real amiga but the amiga controller could be defective as well.

The hdd has been created on an A500 with Kickstart 1.3 (custom/guardian) and Macro Systems Evolution controller. It's around 18 years old i think. I'm not sure if there is some essential data written on track 0 or something (cyl, heads,...) that doesn't allow proper usage on PCs. On the other hand the hdd is maybe just broken.

I will take my scsi card and quantum hdd to a second pc system to make sure there is no problem with my hardware (PCI INT IRQ <> SCSI Card).

I just want to make sure i don't miss a thing. Is everything done right so far? Any other ideas how i could access the drive? How much would professional data recovery cost (just in case...)?

Win XPSP3
WinUAE 1.5.3 beta 0.1 (2008.09.26)
Scsi controller Dawicontrol DC-2980
Amiga HDD Quantum LP52S 950509405 (52MB)
(A500 Kick1.3 / Macro Systems Evolution controller)

[Toni Wilen]

Attach winuaebootlog.txt after you have clicked "Add harddrive" button. Perhaps the drive is really broken or it may be simply buggy/incompatible with "too new" scsi adapters..

[NoX1911]

Full Log attached.

opening device '\\?\scsi#disk&ven_quantum&prod_lp52s_950509405&rev_2.8#5&12e0bfd4&0&000#{53f56307-b6bf-11d0-94f2-00a0c91efb8b}'
device id string: 'QUANTUM LP52S 950509405 2.8 '
IOCTL_DISK_GET_LENGTH_INFO failed with error code 1117.
IOCTL_DISK_GET_LENGTH_INFO not supported, detected disk size may not be correct.
BPS=512 Cyls=0 TPC=0 SPT=0 MediaType=12
device size 0 (0x0) bytes
IOCTL_DISK_GET_DRIVE_LAYOUT failed with error code 21.

That's the dos maintenance utility for the SCSI controller (Dawicontrol/LSI Symbios). Looks like CHS (Cylinder-head-sector) are unknown.


Seems i can enter CHS values manually. I will try to find the correct ones but i will not change anything yet. Maybe these changes are fatal.

So looks like the device is just not detected properly (or defective). Maybe i can get another PC SCSI controller or another on-HDD controller (who knows..).

Would be greatly appreciated if anyone can help me with the real CHS values. I've found nothing reliable so far.

Edit:
Looks like these ones could be correct. Gives a total of 51.178.496 bytes. Seems to be a one platter (2 heads).

Jan 26 00:12:23 inca /netbsd: ahsc0 targ 6 lun 0: <QUANTUM, LP52S 950509405, 2. 8> SCSI2 0/direct fixed
Jan 26 00:12:23 inca /netbsd: sd2 at scsibus0: 49MB, 1219 cyl, 2 head, 41 sec, 512 bytes/sec

Edit2:
Just cosmetics. Altering the values doesn't change anything. Maybe tool specific only. There's a message prior this menu 'medium not accessible'. That's the main problem.
If anyone having a Quantum LP52S please contact me.

[prowler]

@Nox1911:

I'm sorry it has taken me three days to reply to your request, and I'm not even sure whether this will be of any help to you in solving your problem. But, on the basis that you've had no other visible reply in the meantime and I'm sure I would welcome all possible avenues of help if the problem were mine, I've decided to go ahead now anyway.

First, I can confirm that the values you have assumed for your Quantum LP52S hard disk drive parameters are correct, i.e. 1219 cylinders, 2 heads and 41 sectors per track.

Failure to access the drive when connected to your PC via the DC-2980 SCSI controller indicates that either the drive mechanism or (more likely, unless the drive has suffered a knock) the attached drive electronics card has developed a fault.

Is your main concern retrieving data from the disk? Professional data recovery is very expensive, but will be necessary only if the drive mechanism is physically damaged, and should be considered only as a last resort. (Then I think you would view the $160 refurbished drive on ebay much more favourably! )

Does the drive spin up when power to the PC is switched on? Is the process accompanied by a repetetive clattering sound? These symptoms usually indicate that the problem lies with the drive electronics.

Quantum SCSI hard drives like this one were very common in their day, as they were very popular with computing equipment manufacturers, including Apple, so it should not be too difficult to find one.

In addition, the IDE variants of these drives (in this case, the LP52AT) were physically identical, with the drive electronics providing an IDE interface instead of SCSI, so that if you could get hold of a working Quantum LP52AT hard disk drive or just the drive electronics from one of these, then this would give you the option of connecting your drive to your computer's IDE controller to access the data.

If this sounds like an acceptable alternative, then it would better (possibly double) your chances of finding a suitable drive electronics card.

However, these disk drives are very slow when compared to their modern EIDE and Ultrawide SCSI counterparts, and due to their probable incompatibility with the speed of modern computers and the rapid response times required by them of attached devices, it would be a clear advantage to be able to use use an old computer (say 80486-33MHz or older) to test such drives and copy the data from them.

Now, although I don't have a Quantum LP52S hard drive (or even an LP52AT), I do have some experience with disk drives of this type. I have the complete 50-pin, 3.5-inch drive electronics cards from two Quantum PRO-120S SCSI hard drives (120MB: 1123 cyclinders, 5 heads, 48 sectors), and a PRO-210S (210MB: 1156 cylinders, 4 heads, 52 sectors), the half-height drive mechanisms of which wore out.

It might be possible to attach one of these cards to your working drive mechanism in order to copy the data it contains to another medium. This would make your disk drive appear as one of a higher capacity with an large number of bad sectors. To be successful, your disk drive must have ribbon cables compatible with these drive electronics cards and it must contain no important data beyond the 1156th cylinder (track).

You are welcome to try this if you have no success with finding anything better. You'll have to let me know the nature of your drive's ribbon cables in this case, so that I can confirm their compatibility with these cards, which each accept one 14-way and one 20-way ribbon socket.


Good luck!

[NoX1911]

That are some really good news, Prowler. Your post is very informative. That kind of help is really appreciated. You're right, i just want to recover my data from the drive.

I have revived my Amiga and recorded a real environment video (with sound) to show you what happens on a real amiga. It doesn't boot up there neither like the last time i tried but maybe you can hear something. I've also attached a picture of the pcb. You can see in the corners that the pcb was literally "screwed" (damaged) by overlong screws. That was accidentally done by a friend of mine long time ago. I fixed it as good as i could (currently i'm wondering why i didn't choose simple wire connection...) and the hdd worked flawlessly again till the end of the glory days. On the picture it looks like one track is broken but it's not (its just somewhat bent down partially and looks missing on the picture). I double checked with a voltmeter and the connections seem still properly connected from start- to endpoint (0 Ohm between ICs) and not shorted. But i can't say for sure its working like it should. As i remember correctly that was the same behavior like when the tracks were cut. Not sure about 'SCSI reset' though.

A kind person from Canada contacted me so far regarding a spare drive. He might have one. In the meantime i will check for some alternative versions. Now, that i know there are some more compatible versions i will take another look in the web.

Thank you so far for replying and check out the attached video below. I hope you don't have any hassle with MP4 formats (h264/AAC).

[thomas]

I might be wrong but I fear changing the electronics does not help here. The drive obviously respnds to SCSI commands, it even spins down on the SCSI reset command. So the eletronics seems to be ok. The drive just hangs when it attempts to read some data. So it probably has a real magnetical read error on one of its platters, just like an old floppy disk. One that it is not able to recover from by itself.

[NoX1911]

Possible... nevertheless, i will solder/bypass the two damaged tracks again. After that i will swap the whole pcb (if i can get one). That's the least i should try i think.

Isn't there some kind of service mode where the hdd blindly reads from first to last sector as good as possible so that one could capture a raw image? What would professional data recovery companies do in case of weak/bad magnetics?

[hit]

try a linux live distribution. dsl ( http://www.damnsmalllinux.org/ )for example. any other not so old one, will do (ubuntu, etc).
newer will mount ntfs partition without hassle. so you can write the image of the drive to your prefered windows partition (if ithis one is mounted correctly).
once started, try: fdisk -l (lists all available drives and its partitions)
the command dd will do the trick:
dd if=/dev/sda of=/tmp/drive-backup.img

Edit: sda just as an example, probably it will be sdc (again only an example)

[NoX1911]

I tried dd and ddrescue but the device itself is not accessible. Any copy operation results in 0 bytes. Syslog tells device is unknown but a device has been assigned (sdc).

I think there's something wrong with the controller. Unless drives characteristics are stored on track 0 (there are drives that do that i think) the controller should know about all drive essentials (like CHS) and it looks like it does not. Maybe the eeprom/flash is corrupt or something else. The device knows it name though showing on scsi bios.

I hope i can find a new controller...

[prowler]

@NoX1911:

Thankyou for posting the video and high resolution pcb scan attachments.

I found I could not play the h264/AAC codec MP4 format video with my standard WinXPSP2 setup, but I've installed the Media Player Codec Pack v2.2.0 (17.5MB) I found here:
http://www.soft32.com/download_218080.html
and now it plays OK.

Of course, it is never possible to predict exactly what symptoms a faulty pcb might be causing and the sound is inconclusive. Is there another hard drive which can be heard as well as the Quantum? Is the drive spinning up continously when the power is connected or only briefly in response to the SCSI reset command?

Either way, there is no definite indication that the drive mechanism is faulty, and in my experience it is the electronics card which is far more likely to be the culprit when the drive fails. (The repetitive clattering noise I have mentioned sounds rather like a self-destruct process has been triggered, but it still turned out to be the drive electronics at fault! )

The pcb scan shows the serious damage you have described near the top right-hand corner in the region of IC U210, where two, possibly three, tracks appear to have been broken. How about the underside of the card? Is this OK? (The drive electronics cards I have mentioned are populated in this area with surface-mounted components on the underside.)

Unfortunately, the picture also shows that my drive electronics cards are not compatible with your hard drive mechanism. Your pcb has three sockets (P201,P202,P203) accepting ribbon edge connections to the drive mechanism, whereas mine each have plugs which connect with sockets soldered to pads at the end of only two ribbons. The reason for this discrepancy is probably that my Quantum Pro drives are older (1988/9).

Let's hope that the person from Canada who has contacted you actually does have one of these drives. After all, this would be the best solution.

There is one other point I should mention. If you aren't successful in finding an LP52S but you do manage to get hold of an LP52AT, you may have to try it with a motherboard which has no auto-sensing hard drive detection in the BIOS, or at least allows you to override the auto-sensing with user-defined hard drive parameters if the correct drive geometry is not detected by default. (I can elaborate on this later if necessary.)

[prowler]

@NoX1911:

Hey, it's all gone rather quiet here just lately!

I really think it would be worth taking a look at the underside of the pcb in the area(s) where it is damaged.

I accept that you were able to fix the damage to the tracks after they were broken and that it worked flawlessly after this, but if this area is populated with surface-mounted components on the underside, then one or more of these may have lifted slightly when the over-long screw put a strain on the board and, although no problems were apparent at that time, the weakened joint(s) and connection(s) that may have resulted could have corroded and/or separated in the meantime while the drive was in storage, and this may be all that's causing the present problem.

If this is what has happened, and you were to flow some solder over the tracks adjacent to the components on the underside of the board in this area, it shouldn't be too difficult and would likely fix the problem.

Has the chap from Canada got back to you in connection with his hard drive? I hope so.

Good luck!

[NoX1911]

I checked the underside. It looks untouched and properly. But i've found that there was some copper shimmering around the fixed lines. It is ground. Its possible that the outer lane "hit the ground" and something was seriously damaged. I also removed the fixed lines completely but the hdd behaves exactly the same in both states. So no difference between fixed lines and no lines at all. SCSI reset still works though. I used direct cables this time to fix it but didn't change anything.

I got some picture in the meantime from the guy in canada. It is the right 52S and looks exactly the same like mine. I also found a small shop in the US that sells a refurbished 52AT for 10$. Doesn't look too bad to get a new pcb...

I also opened the hdd to check interior condition. It looks fine to me. At power-on the heads move to a specific position and after a second back to the outer position. No signs of read-errors or multiple read attempts. To be more exactly, it looks like no read attempts at all.

The hdd is a two platter (with 1 dummy on top. The 105 will most likely use both 2 platters). The upper platter is unprocessed and not polished at all. The lower is shiny and glossy. At least that part that i could see. The hdd cap/cover looks completely clean.

[prowler]

@Nox1911:

Hey, you really shouldn't have opened up that drive mechanism if you're still hoping to read the data off it!

Those units are hermetically sealed to keep them free of contaminants.

The gap between the heads and platters is typically so small that a single particle of dust or even cigarette smoke can cause damage if it gets between them...

So get it sealed up again quick and hope that no harm has been done while you've had it apart.

You're best hope is that, being an old drive, it's not made to the same exacting tolerances as modern drives...

Those illustrations you see of disk internals in magazine advertisements are just to make them more interesting to the eye... They're not meant as an invitation to actually open them up!

What next, I wonder...

[NoX1911]

I'm aware of this but i had to make sure its state is proper inside (condense, water or something). There is also a massive rotation speed (air flow) that will clean the surface during spin up and its a low density platter. I would never open up my 333GB perpendicular platter drive. Chances that the hdd is completely crashed due to this attempt is very low. The drive is not meant to be used again anyway. Once i got the physical image its just a memorial part of my amiga collection again.

[prowler]

OK then, but still fingers crossed, eh?

[prowler]

@NoX1911:

Yesterday, I was so alarmed to see that you'd opened up the drive mechanism that I forgot to read the rest of your post in my haste to post a quick-fire reply with a warning of the possible consequences, for which I apologize.

Now, since you seem to have a choice between a replacement SCSI or IDE interface for your hard drive, I don't blame you for wanting to check the condition of the drive mechanism assembly before parting with hard-earned cash.

I was shocked by the appearance of the upper platter in your pictures until I realised that it was a dummy and therefore not of good finish! You are right in your assumption that the LP105S uses both platters; the drive parameters in this case are 1219 cylinders, 4 heads and 41 sectors per track.

As the drive mechanism appears to be in good shape, I think you should go ahead.

To keep things simple, it would be better to replace like with like and go for a SCSI interface, but either will do.

Good luck, and let me know how you get on.

[prowler]

@NoX1911:

Did you get the Quantum hard drive resurrected?

prowler

[NoX1911]

In the meantime i got two IDE drives. The first one was broken as well. It didn't even start up properly on my system (no access, kind of similar symptoms). Then i got another one some weeks ago but i didn't test it yet. The owner told me he tested it successfully before shipping. I will do it in the next days. Stay tuned.

[Thorham]

Originally Posted by NoX1911

In the meantime i got two IDE drives. The first one was broken as well. It didn't even start up properly on my system (no access, kind of similar symptoms). Then i got another one some weeks ago but i didn't test it yet. The owner told me he tested it successfully before shipping. I will do it in the next days. Stay tuned.

This may not be useful advice, but if those drives are old second hand drives, than don't bother. Drives wear out, and if your IDE interface works properly, you're much better off just buying a new IDE drive for 50 euros at the local peecee shoppe (same applies to things like old CD Rom drives, don't bother, and buy a new DVD Rom drive)

[NoX1911]

Quantum is dead for years. There are no new drives, some refurbished ones though. My drive did wear off already. I'm just trying to give it a last kick by attaching another hdd controller. I'm not interested in the 50MB storage device, just the data on it.