n_K Posted January 3, 2015 Share Posted January 3, 2015 So, I've had a few SCSI drives in use for a while in my desktop PC, each with different OS which I switch between (the drives are 33.6GB Fujitsu MAJ3364MC). Today whilst on my W7 drive in DOSBOX, I got a BSOD, it didn't wait there for more than half a second and the PC rebooted, drive still powered up and detected by the SCSI card, only I couldn't boot from it (No OS installed), so I powered it off and powered it on - the result is it no longer powers up, the LED on the drive PCB just flashes, and nothing is detected by the SCSI card. I have had no problems with this drive, no read errors, no unusual noises, it's been spot on perfect except for this. I've tried swapping the drive PCBs around with the other ones I have and doing so causes the drive to power up - GREAT! Except the fujitsu engineers were so incompetant that they must have stored the drive's serial number on the disc and in the firmware because using a different drive's PCB gives a '16 - HARDWARE ERROR' response when the SCSI card detects it, and it refuses to work. So now I'm left with a drive that does obviously work but a PCB for it that refuses to boot up. I'm not sure if I can flash a firmware upgrade to it (might have got corrupted?) or what to fix it and was posting this for suggestions or ideas. I've been looking up what all the parts are on the PCB and there is possibly an EEPROM next to the flash module for the firmware, so maybe they store it in that. I don't have access to a soldering iron for a while but next time I do I might desolder it, read the contents and see if they store something in there. Link to comment Share on other sites More sharing options...
+BudMan MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 Why would anyone go through so much trouble for a failed disk?? Buy a new one - ALL disks fail, they might give you warning - they might just DIE. For what possible reason could you have in spending such effort on something that is designed to be throw away. Lets just hope you spend 1/10 of this effort on your backups and your golden.. 33 gig?? So my flash drive I carry around in my pocket that I got for $15.. T3X4S and Aergan 2 Share Link to comment Share on other sites More sharing options...
T3X4S Posted January 3, 2015 Share Posted January 3, 2015 SCSI stuff is really sensitive to cabling - are you sure you are OK in that aspect ? Its always been my philosophy to think of the simplest things first. I know you said it has given 0 problems, but when was the last time you did a reboot ? But like Budman mentioned - its an old, obsolete drive - toss it Link to comment Share on other sites More sharing options...
ToneKnee Posted January 3, 2015 Share Posted January 3, 2015 That drive was released in 2000, I say it's had a good 15 year run. goretsky, MorganX and Aergan 3 Share Link to comment Share on other sites More sharing options...
n_K Posted January 3, 2015 Author Share Posted January 3, 2015 Why would anyone go through so much trouble for a failed disk?? Buy a new one - ALL disks fail, they might give you warning - they might just DIE. For what possible reason could you have in spending such effort on something that is designed to be throw away. Lets just hope you spend 1/10 of this effort on your backups and your golden.. 33 gig?? So my flash drive I carry around in my pocket that I got for $15.. The data on it isn't important, it's a clone of another drive but with steam, etc. I don't care about the size of it. And why bother? Because it's something to do, heck I'll learn something from it. SCSI stuff is really sensitive to cabling - are you sure you are OK in that aspect ? Its always been my philosophy to think of the simplest things first. I know you said it has given 0 problems, but when was the last time you did a reboot ? But like Budman mentioned - its an old, obsolete drive - toss it Cablings fine, it's just got one drive on it with a terminator, the other 2 drives I have here both work fine. I turn this PC on daily, it's off over night. That drive was released in 2000, I say it's had a good 15 year run. Yeah it's been a good drive, but I want to know why this has happened, how to fix it and how to prevent it from happening again. Link to comment Share on other sites More sharing options...
Anibal P Posted January 3, 2015 Share Posted January 3, 2015 The data on it isn't important, it's a clone of another drive but with steam, etc. I don't care about the size of it. And why bother? Because it's something to do, heck I'll learn something from it. Cablings fine, it's just got one drive on it with a terminator, the other 2 drives I have here both work fine. I turn this PC on daily, it's off over night. Yeah it's been a good drive, but I want to know why this has happened, how to fix it and how to prevent it from happening again. It's a 15 yo drive, it died, you can't prevent disks from dying, they just do Link to comment Share on other sites More sharing options...
n_K Posted January 3, 2015 Author Share Posted January 3, 2015 It's a 15 yo drive, it died, you can't prevent disks from dying, they just do As I've said, with the original PCB the drive light flickers and drive does not spin up, with a different PCB it spins up fine but fails a check (most likely serial number related). The drives not dead, the PCBs dead. Link to comment Share on other sites More sharing options...
+BudMan MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 "The drives not dead, the PCBs dead." Same thing.. While if a hobby of yours keeping old disks that have no use in modern computing alive great.. I keep meaning to fire up my old commodore 64 To be honest this is not the best place to ask for help with your hobby.. I doubt you will find anyone here of sim interest in keeping a 15 year old disk up and running. For 1 many of the readers here are prob just a couple of years older than the disk your working on if that hehehe Good Luck.. You will prob find better help on a antique computing board of some kind. "how to prevent it from happening again." Sorry dude but old things die.. that is just how it is - we all will face it ourselves.. You do understand these disks are not old cars where you can swap parts in an out and keep them running.. Do you fix a lightbulb? Aergan and T3X4S 2 Share Link to comment Share on other sites More sharing options...
n_K Posted January 3, 2015 Author Share Posted January 3, 2015 It's not that it's old that matters though really. An enterprise drive's PCB randomly failed, yes it's old and that might have played a big part in it, but I need to find out why, that's just how I am. I've got a server running, if one of the drives in it exhibited this exact problem [not a drive dying but the PCB firmware failing] (they're SAS drives so yes, much newer) I would be utterly gobsmacked. Even worse if it happens in a data centre. This is the first time I've seen where a PCB swap from an identical drive will not work. I bought a 2.5" SAS seagate drive years ago and it was DOA, I kept it though because I have a few of the same drives incase something goes wrong with them (I swapped over the PCB to check if the PCB was dead, and it worked). Anyway, I just assumed there would be people interested in hacking things and finding out why problems occur. If I do find out a solution to the problem one day I'll come back and post it in case anyone is interested. Link to comment Share on other sites More sharing options...
+Fahim S. MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 It's not that it's old that matters though really. An enterprise drive's PCB randomly failed, yes it's old and that might have played a big part in it, but I need to find out why, that's just how I am. I've got a server running, if one of the drives in it exhibited this exact problem [not a drive dying but the PCB firmware failing] (they're SAS drives so yes, much newer) I would be utterly gobsmacked. Even worse if it happens in a data centre. This is the first time I've seen where a PCB swap from an identical drive will not work. I bought a 2.5" SAS seagate drive years ago and it was DOA, I kept it though because I have a few of the same drives incase something goes wrong with them (I swapped over the PCB to check if the PCB was dead, and it worked). Anyway, I just assumed there would be people interested in hacking things and finding out why problems occur. If I do find out a solution to the problem one day I'll come back and post it in case anyone is interested. In the data centre, you'd only bother wasting your energy if it happened several times on the same server or storage array. Link to comment Share on other sites More sharing options...
+BudMan MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 You will find lots of people interested in "hacking" things - maybe I am wrong, but been on this board a long time and this is the first sort of thread I have seen of this nature.. Now if you were talking somewhat current tech, sure you will find all kinds of interest here. But your talking about something that was never designed to be tinkered with like a old 57 chevy, and to be honest even if working has little use.. For $50 you can buy a modern disk, why would you want to play with stone age tech? If you had a brand new case of those - I don't think you would find anyone here of much interest in them.. When you can go to your local store and for less than $20 pick up a stick that has same amount of space, and way faster in accessing and carry around in your pocket.. Dude it died - it was never meant to last 15 years Link to comment Share on other sites More sharing options...
ToneKnee Posted January 3, 2015 Share Posted January 3, 2015 Budman is right, even Enterprise equipment can die. A lot of Enterprise equipment is guaranteed for 3-5 years and the fact the drive is 15 years old suggests it had a really good run. Many things can affect hard drives and the most common one is mechanical failure. It happens dude, move on. Link to comment Share on other sites More sharing options...
ShadeOfBlue Posted January 3, 2015 Share Posted January 3, 2015 Don't bother trying to fix the drive. It's probably not going to happen. Now I'm going to explain why for some reason (I'm bored?): This is not some kind of serial number checking situation. Every individual hard drive is unique. There's all kinds of information, like defect lists and translation tables, that are stored on that drive and are for that drive alone. Depending on the model, some of this information may be in flash memory on the PCB or in a hidden service area on the platters themselves. All that aside, they make changes to the PCBs all the time. Just because two drive are the same model doesn't mean the PCBs are compatible. A visual inspection may or may not show any relevant differences. Since your drive does not spin up with the original PCB, but does with the other one, the problem must be either corrupted information in flash memory or a bad component on the PCB. The best-case scenario would be that you move the firmware chip over from the bad PCB to the good one and it works. Some drives store drive-specific information in more than one chip, however. And it may even be corrupt firmware that is causing the problem in the first place, so you might just be right back to where you started after doing that. Statistically, firmware corruption is the more common failure. Even if transferring the chip would work, a soldering iron is not the right tool for that job. Some have special tips available, but usually you'd want to use hot tweezers or a hot air gun. Trying to desolder a multi-lead chip like that any other way will just damage the chip and/or the board. If you just want to read the chip and look at the contents, and you have the right hardware to do so, then have at it. But there's nothing you're going to be able to do to fix the firmware short of buying very expensive software/hardware to do it for you. I have some understanding of the non-executable portions of newer Western Digital firmwares. At least there was the tiniest bit of information about them. But that's only after spending countless hours scrounging for it (a handful of paragraphs in total), and hundreds more in front of a hex editor. Good luck finding freely available information about Fujitsu firmware. This is not a weekend project, and most of what you would learn would be useless with newer drives or other manufacturers. Link to comment Share on other sites More sharing options...
+Warwagon MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 I still have a 10,000 RPM 32 GB SCSI Drive. This thing is a monster, and boy does it get HOT! Burn to the touch hot. but man is it fast. Link to comment Share on other sites More sharing options...
+BudMan MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 "but man is it fast." Sorry but fast in terms of what how fast the disks spin? How fast it heats up the room? Here is benchmark of disk talking about - which yes in its day was a rocket ship.. 15 years ago http://www.tomshardware.com/reviews/a-scsi,284-3.html That is seq reads at what 40MBps best speed across the disk.. Here is my my $15 32GB thumbdrive compared to a scsi disk that was 15RPM and 4 years newer.. Do we really want to compare it to SSD? While these disks were lightening in their day - I remember hooking up lots of these.. Matter of fact I have some 16 and 9Gigers in my garage - box of them.. I don't know why I keep them around, but I had grabbed them when we moved on at the company to newer tech.. They wee still someone decent back then.. But when I can get better performance out of a $20 item that I carry around in my pocket - WTF would I spend any time on trying to get an old one of these to work for? Other than some sort of hobby - which is great, more power to you.. But might have more fun just collecting stamps - atleast those would have some value as they got older.. These sorts of disks are just worthless junk.. Shoot found the hd tach used in that old benchmark, version 3 which was like 2007 version had to run in comp mode to get it run even, etc.. goretsky 1 Share Link to comment Share on other sites More sharing options...
+Warwagon MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 I don't actually use it anymore. Its in a box labeled "Old drives" ... but back in the day I use to run XP off it. Now my SCSI cable no longer works... I wonder why... T3X4S 1 Share Link to comment Share on other sites More sharing options...
+BudMan MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 Look at that old school monitor as well Its like a time capsule of antiquated tech...I might be able to find a old 10mbit isa nic in a box No wait I did throw those out last clean up because WTF, when would I ever use them Link to comment Share on other sites More sharing options...
T3X4S Posted January 3, 2015 Share Posted January 3, 2015 I don't actually use it anymore. Its in a box labeled "Old drives" ... but back in the day I use to run XP off it. Now my SCSI cable no longer works... I wonder why... Is that a CRT ?? holy cow. *notices the "odor eaters" bottle .... ewwwww Link to comment Share on other sites More sharing options...
+Warwagon MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 Is that a CRT ?? holy cow. *notices the "odor eaters" bottle .... ewwwww :puke: That photo is from December 7th, 2008 T3X4S 1 Share Link to comment Share on other sites More sharing options...
+Warwagon MVC Posted January 3, 2015 MVC Share Posted January 3, 2015 Speaking of old drives, take a look at this 4200 RPM 9.6 GB Bigfoot drive Link to comment Share on other sites More sharing options...
n_K Posted January 3, 2015 Author Share Posted January 3, 2015 Don't bother trying to fix the drive. It's probably not going to happen. Now I'm going to explain why for some reason (I'm bored?): This is not some kind of serial number checking situation. Every individual hard drive is unique. There's all kinds of information, like defect lists and translation tables, that are stored on that drive and are for that drive alone. Depending on the model, some of this information may be in flash memory on the PCB or in a hidden service area on the platters themselves. All that aside, they make changes to the PCBs all the time. Just because two drive are the same model doesn't mean the PCBs are compatible. A visual inspection may or may not show any relevant differences. Since your drive does not spin up with the original PCB, but does with the other one, the problem must be either corrupted information in flash memory or a bad component on the PCB. The best-case scenario would be that you move the firmware chip over from the bad PCB to the good one and it works. Some drives store drive-specific information in more than one chip, however. And it may even be corrupt firmware that is causing the problem in the first place, so you might just be right back to where you started after doing that. Statistically, firmware corruption is the more common failure. Even if transferring the chip would work, a soldering iron is not the right tool for that job. Some have special tips available, but usually you'd want to use hot tweezers or a hot air gun. Trying to desolder a multi-lead chip like that any other way will just damage the chip and/or the board. If you just want to read the chip and look at the contents, and you have the right hardware to do so, then have at it. But there's nothing you're going to be able to do to fix the firmware short of buying very expensive software/hardware to do it for you. I have some understanding of the non-executable portions of newer Western Digital firmwares. At least there was the tiniest bit of information about them. But that's only after spending countless hours scrounging for it (a handful of paragraphs in total), and hundreds more in front of a hex editor. Good luck finding freely available information about Fujitsu firmware. This is not a weekend project, and most of what you would learn would be useless with newer drives or other manufacturers. 1) The PCBs have a single flash chip with the firmware 2) Under this is potentially an EEPROM, chip marking is C01A, all address pins are grounded so I'm hopeful. 3) The drive info such as bad clusters etc. is all stored on the drive spindle. 4) The drives are the exact same, I have 8 of them in total, they all came from an old dell poweredge, all have the same production dates, firmware codes, etc. except the serial numbers. 5) I do think it's a corrupt firmware on the flash chip, I want to find out why it's got corrupted though. If it's through old age and the chips just degraded - that's fine. If it's something else though that can potentially happen to another drive and is caused by my setup or whatnot then I'll try to narrow that down so I can avoid it in future. 6) The EEPROM is an 8 pin SOIC, I've got tools and tips to desolder it (at my other place though) but I'll take it to work monday and use their better soldering tools to switch it over. 7) I've got all the equipment for reading the flash memory and EEPROMs, ironically this PC was bought because it had a parallel port to use with my willem programmer! Lucky I've got another W7 drive with the willem tools on to read data back if needed 8) If I can get it working, that's cool. If I can't, oh well. At the end of the day it's something to do and experience to gain. And budman, I've got a Link to comment Share on other sites More sharing options...
n_K Posted January 5, 2015 Author Share Posted January 5, 2015 So, I swapped over what looks to be the EEPROM chip to no luck, still gives 'hardware error', so I'm now assuming there's no drive firmware updates for this because it stores the serial number or something actually in the flash memory (512KB), which does in all honesty seem stupid. I've proved that the drive's not dead though, toying with the SCSI adapter settings you can get it to ignore errors and kind of still allow devices without checking them, booted into linux, mounted the drive and what do you know, all data there in tact. Because the drive thinks it's got some huge error though it's all read only, and the only corrupt files were some browser files etc. which is expected seeing as it was open at the time. My next idea is to desolder the flash chip and read it in for the hell of it, I don't have any TSSOP boards though so I might pass on that. None the less, learnt a lot. goretsky 1 Share Link to comment Share on other sites More sharing options...
Recommended Posts