We have an old 20GB Maxtor here, which is really a Quantum drive. The PC-3000UDMA Quantum/Maxtor utility correctly identifies it as a Maxtor 6L020J1 but gives CP errors when it starts (see attached photo)and no tests can be performed. There was a loader for this model in the database which we used (not being sure if it was completely compatible with this drive) and after the loader loads and the utility starts, the CP errors are gone, but the model is shown as the wrong one (NP40), and the size is shown as 209 MB (see attached photo).
When we switch to Data Extractor, the drive can read (with ECC errors) and the data looks OK, but it only goes to 209 MB.
With the loader still loaded we ran a system area structure test, and almost all the firmware modules could not be read-- just 5 were successfully read. It seems the head is OK, since we can clone the first part of the drive.

What do you think is wrong with the drive? What should we try next?

Did you tried to open a task when the drive was detecting with the original size?

What were the original symptoms?

Yes-- in the PC-3000 Data Extractor (without the loader-- correct model, firmware and size shown) two partitions were detected, labeled 1FD and 3FD -- I guess this means it doesn't recognized the filesystem-- trying to expand the partitions gives "catasrophic error" message with nothing shown in the partitons.
The drive is fairly noisy (it is 11 years old or so) but no clicking, just a high pitched whine. When hooked up to the Deepspar, it shuts down right after Rescan.
In PC3000 Data Extractor, after loader is loaded, if copy is selected it will copy until it gets to the identified size of 209 MB and then stop. It copies extremely slowly if default Data Extractor settings are used, and sectors are read with ECC error. If I read with CRC off, it copies the 209 MB in about two hours, and the data looks OK. Without the loader, nothing can be done with the drive in the PC300UDMA utility or in Data Extractor.

If drive can not read SA properly, it won't let you access user area.
You have to diagnose exact problem with that drive.

_________________
www.datarecoveryne.com

But we were able to clone the first couple of hundred megabytes with the loader working, so at least partial access to user area is possible.
Where would you go from here?
Is it possible that the loader we got from PC3000 database is nt complete, and a more complete loader would help? Any chance a hot swap would work?
What would you say our next step should be?

Detecting that king of partition probably means the prive have a proprietary file system, maybe the drive was from a industrial machina or something like that.. But at this point you were able to copy the sectors and image the drive? did you try it? what was the sector reading error?

That old drives are usually noisy, that´s not a problem for sure.

When the loader is loaded, the drive shows two NTFS partitions, which is correct, according to the owner. The problem is, with the loader in place the drive does not ID properly, and shows only 209 MB, which is all we can image-- it stops at that point. It is a 20GB drive, so we are getting only about 1/100th of the drive.

The drive can't read most of the modules from the service area (with the loader). I don't know exactly what this means-- system area damaged? Ace Labs says that since the utility gave no errors at startup with the loader in place, it must be the right loader for the drive. Do you think I should try writing modules from a similar drive to the system area? I am a bit afraid to do that, as I am not sure how compatible firmware is from drive to drive in these old Quantum drives, and I don't want to kill the drive. Clearly we shouldn't write any modules that are drive-specific (p-list, adaptives, etc) but I don't know for sure which modules would be OK to write from another drive.

You are probably right about the noise-- I have heard old Maxtor drives that worked fine that were about this loud.

What EXACT drive are we talking about ? (label, first characters of serial no.) and what is the PCB 'code name' ?
P.s. Don't write ANYTHING on it !

I believe the loader is bad and/or some of the adaptives has gone wrong. Reading 200MB with ECC errors off in two hours is way too slow to be correct.

I would start by loading some other loader.

_________________
http://www.northwind.gr
SandForce SSD Recovery
Ransomware Reverse Engineering - NoMoreRansom! partners

Unfortunately it's QUANTUM. The problem to me is elsewhere than loader - for AS, 740, xX etc it's the same. That's why I was asking for model and other info...

I am attaching a photo of the label. It's a Maxtor D740x-6L MX6L020J1
The Quantum/Maxtor Utility defines it as family D740x-6L model 6L020J1 -- the photo shows all the numbers.
The PCB # is 10-123301-01 REV D
Using donor exactly matching PCB makes no difference
Is this drive a good hot swap candidate?
I haven't tried reading the P-list and G-list from it yet. If they are stored on the platter on this drive, those modules probably can't be read, but I will check. Only 5 modules could be read from system area when the drive was funtioning using the loader, none without.

Just saw that the serial number is on a small label on the end of the drive. The label says:

S/N 661213627553
TSZXX

What does the second line signify? I don't know much about Quantum drives.

Ha! VIPER, my favorite.
Hope it is firmware not heads.
Check carefully CPs and modules, and be sure the FW was not tampered.

With the loader, all the CPs were read OK (or so it says) but when I try to read the modules from service area it errors (TON) on almost all-only modules 5, 40, 4c, 73, and C1 test at least one copy OK. I tried a hot swap on PC-3000 with similar drive, but recalibration failed-- drive acted same as with original PCB. I suspect the service are is damaged.

I'm not convinced the loader we got from PC-3000 database is the best we can do. Is there a way to make a loader from another drive? I don't know how to make a Quantum loader-- not in the PC3000 menu anywhere.

I don't want to do anything that could lose client's data-- any suggestions at this point?

I tried that PC3000 loader on another similar drive, and the other drive also IDs wrong after the loader-- shows up as a Quantum VP40, 209 MB, just like the source drive does. It can read all the modules from the service area OK though, which source drive cannot do.

Does anyone have a better loader for this drive, or can anyone tell me how to make one? If I could get it to ID as 20GB after the loader is in, I probably could get the bulk of the data from it, since I can get the first 209 MB.

The loader is the one you have like in ISA PC3000, doesn't exist 'a better loader' . But if a critical CP or module is screwed, or cannot be read, the drive behaves incorrectly at start. Same if you have a weak head or (rare) pcb malfunction.

but the loader in the PC3000 database for this drive turns all of the donors into 209 MB drives-- that can't be right. The donors are good drives with no problems, but when I use the loader, they become VP40 209 MB drives.

Do you know how to make a loader for quantum drives? There is no option in the PC3000 Quantum menu for making a loader, only for using one.

There's something wrong.

If the drive doesn't init properly, you use the loader. Then did you try to load the saved CPs in RAM ? After that, in RAM you have the correct model and whatever it takes to at least INIT.
With a loader alone and if cannot load properly ALL the required CPs, the drive IDs as VP40.
In ISA after you load CPs in RAM the drive is restarted so you hear a sort of calibration, and then you see the correct model, capacity etc.
Only at that point, assuming everyhting else is fine, you have correct access to service area, otherwise you have to find out what else is wrong.

To me the problem of your drive may be a corrupted CP or module so the drive does a partial calibration (you just hear some "spokes" not the calibration you are used to hear) but also can be a weak head. Hope it is not the latter, as due to their structure recovery will be a pain in the arse unless you know PERFECTLY how their firmware work.
From the symptom and behaviour, I would suspect #30, SMART subsystem and error lists.
Forget to find a donor, you can test 100 donors to find ONE that PARTIALLY more or less works (I have developed another solution for it, in fact). One of the worst drives to work with, and that's why, depending on what happened, it is an expensive case if you ask me.

Forgot one thing : in PC3000 you see the model but the SN is 00000... suspicious. The SN is stored in one CP.

Are you absolutely sure someone else didn't try to upload a FW or poke with it ?

(latest thing I think is that PC3000 UDMA doesn't support correctly this model, if so we'll think about Plan B !)

I tried loading the saved CPs to RAM, and it now shows the correct name and serial number, but the capacity is still stuck at 209 MB. If I go to HDD configuration changing, it shows correct Max LBA, but the number of logical configuration shows only 425 cylinders-- a good drive shows something like 50000, if I remember right. Also, if I go to set Max LBA, the biggest number I can choose for max LBA is 1, 500,000 or something like that. If I take a good donor using the loader from the database, then load CPs into RAM, I can't get it to show more than 209MB either, although if I go to set Max LBA, it gives the choice to set it to actual value 40,000,000 or so, but it doesn't work-- it times out and the drive is stuck at 209MB. So there seems to be something not right with the loader in the PC3000 database. It turns all drives we have of this model into 209MB from 20GB.

I don't think anyone else did anything to this drive except try to image it with DDrescue or HD Clone.
The PC3000 does have support for this drive. Other drives from same model work well, without the loader. They show the right Max LBA and drive info, and work fine with DE.

What do you think we should try now? I can't get anywhere with the drive without the loader, but with the loader I can only access the first 209 MB. Is there any way that you know of to make a loader from a Quantum drive? Maybe a better loader would work.

Any other suggestions?

By the way, you said check CPs and modules-- when I do structure test, all CPs are good, but all the modules except 5, 40, 4c, 73, and C1 are bad, or can't be read. This is with loader.