Flashing LED on old Quantum HDD PCB

Do you hear the disk spinning when power is applied to it? Many old drives, particularly those that have been sitting for a while, fail to spin due to the heads sticking to the disk surface. I believe the term is stiction or something like that. If it is not spinning apply power and tap the edge of the drive sharply against the palm of your hand. Sometimes that will break the heads loose.

A drive not spinning will not be detected.

I'm pretty sure that is a code for something - but you'll likely need to be a tenacious googler to find out what.

"Spinning up, hold on ..."

"Error."

Old drives did that, as they weren't SMART capable and had no way of communicating any kind of detail as to the error. The "blink code" will tell you exactly what's wrong, if you can just decode what it means.

The drive may not be spinning up, or may not be achieving full/stable speed quick enough.

Yup sounds like a stuck drive. Those bigfoots did that a lot, as did the earlier Seagates.

As mentioned by another, just rap the drive on it's side, and it should let go. Be aware however, that the stuck head may pull off the magnetic material when it breaks loose. Normally we did this as a last ditch effort to get the data off of the drive before tossing it out.

- Tim -

On Thu, 1 Dec 2011 10:54:41 GMT, snipped-for-privacy@Onetel.net.uk.invalid (Windmill) put finger to keyboard and composed:

Does the same thing happen if you power up the drive without the interface cable?

If you disconnect one of the RAM buffer pins, does the LED pattern change? Doing so should result in a Sector Buffer Error (error code

- Franc Zabkar

You mean tap the drive into the side of a trash can.

Oddly I have a 12GB bigfoot on my desk, I use to keep a pile of papers from blowing away.

I think some drives had a routine for head stiction, maybe as simple as repeatedly trying to start up a set number of times, maybe trying alternately trying to start bacwards a few times.

Could try repeated power cycling (start up routine) before clonking it one to unstick the heads.

An approach I find more effective in dealing with "stiction" problems in older hard drives (and somewhat gentler on the drive):

- Set it down gently, flat, on a smooth table-top.

- Grasp it on both longer sides (i.e. across its short axis) with one hand.

- Rotate it sharply, without lifting it from the table, by "snapping" your wrist. Often, the inertia of the platters will "break free" whatever is stuck (head-to-platter or shaft-to-bearings).

I agree with others, though... a drive which has stictioned itself once is probably not to be trusted. If you get get it to spin up, make another set of backups immediately. If you want to keep using it, leave it spinning... it may stick again if it's powered down for some time.

Had a drive once I had to use longer cables and park it on the desk so I could clout it with a screwdriver handle immediately after pressing the on button.

Took a fair bit of skip raiding to find a replacement that started without 'help'.

In the older days, I had a nice big ST506 MFM shoebox sized 5.25" full height drive. It used to fail to spin up with a blink code.

Reason: one lousy transistor that was responsible for lifting the "brakes" solenoid off the motor, used to go pop. Never did find out why. Just replace and continue :(

On Fri, 2 Dec 2011 13:17:46 GMT, snipped-for-privacy@Onetel.net.uk.invalid (Windmill) put finger to keyboard and composed:

There is a product manual in several parts here: ftp://ftp.octek.com.hk/UTILITY/UTILITY_PATCHES/DISK_MANAGER/QUANTUM/MANUALS/BIGFOOT/BG_CY_AT

There is a block diagram of the electronics in section 5.2 of chapter

2. The DRAM is a 64K x 16 IC. You should be able to locate its datasheet. I suggest you lift one of the data pins.

I have a database of datasheets here:

FWIW, there are error codes listed for the ATA Execute Drive Diagnostic command (section 6.7.7 of the manual). I don't know if they bear any resemblance to your codes, but 8Xh indicates error number 0Xh for drive #1 (slave).

If putting a fault on the DRAM results in a flash code of 83h, then this would be consistent with the table of errors in the manual.

- Franc Zabkar

On Fri, 2 Dec 2011 13:26:36 GMT, snipped-for-privacy@Onetel.net.uk.invalid (Windmill) put finger to keyboard and composed:

If the drive has a stiction fault, then you should see pulses at the motor windings as the motor controller attempts to kickstart the drive. I believe this is called "spin buzz".

Alternatively, there will probably be a current sense resistor, or an array of parallel connected sense resistors, to sense the motor current. You could measure the voltage at this point. Note that there will also be a current sense resistor for the voice coil.

FWIW, here is my database of motor controller IC datasheets:

I don't know if this tutorial will be of any help to you ...

- Franc Zabkar

I have the background & most of the equipment to do things like that but not the technical info nor the ability to easily work on SMDs (need newer reading glasses, steadier hands, and less absent-mindedness). 'Age I do abhor thee', as the man said.

That's a very impressive collation of information. I've kept a copy for future reference, though hopefully I won't need it. Used to work on ancient mainframes long ago, so in theory I can still understand the details. If I can just find my glasses.

I'll bear that in mind if the drive sticks again. Was going to buy a small cheap drive to use for an old PC on which I planned to run my first Windows system (so I could use a couple of old programs which Wine can't handle). Then prices soared, so I dug out this old Quantum drive. It won't be a tragedy if I lose it completely.

Many thanks. I held off on more violent unsticking methods until I knew more, then used your method. Worked perfectly, with little force needed.

That was pretty much what I had in mind when I disconnected the drive a few months back. But the floods in Thailand changed my mind.

Funnily enough the old PC it's going into came from the same place....

On Tue, 6 Dec 2011 12:08:24 GMT, snipped-for-privacy@Onetel.net.uk.invalid (Windmill) put finger to keyboard and composed:

Ditto.

That too.

- Franc Zabkar