Saving a poor Seagate ST-412.

[ThePresi]

I'm trying hard to revive a ST-412 hard drive, coming from an otherwise happy IBM 5160.
I am seeking advice from seasoned retro-heroes or electronic expert as my knowledge about electronic is very basic.
All I know I have learned from Adrian's Digital Basement and other retro Youtubers.

The problem:
The HD doesn't spin up when plugged to power. I can hear a faint click then nothing.
I assumed a stuck platter/heads, but both spindle and step seem to move freely.
I'm doing all the tests with the interface board disconnected from the hard disk sealed inclosure. I assumed it should still be a valid troubleshooting method while avoiding to damage any possible data from the platters.
The voltages on the interface board seem fine, no obvious shortcuts, or blown capacitors, but I have noticed a burning resistor.
It's R90 from the schematic of the ST-412 model 20110 (attached). It's a resistor by the output of the negative phase A that goes to the spindle. It gets 12V across it.
I have isolated the resistor, and it checks fine off the board. Before the resistor there is a 7406 inverter IC. I isolated it and it seems working fine off the board as well.
Before the 7406 there is the 6500/1. It gets quite warm too. I checked the usual suspects, voltages, reset pin, NMI, CNTR and clock, but all seems OK.
There is no activity on the 4 data channels. They are either high or low. I'm not sure if this is an issue or the CPU is just waiting for a signal (since the HD enclosure is disconnected).

What is next:
I have ordered a couples of 6500/1. I want to try to replace it to see if that's the issue.
I will also isolate the three transistors array's ICs that seem to interact with the burning resistor from the schematic.
They are two PNP and one NPN that drive the positive and negative phase signals to the spindle.

The questions:
Does this approach make sense? Am I missing something obvious?
I wish to get some expert advice before proceeding, as those ICs are a massive major pain in the floppy to desolder.

Thanks!

 

[oldpcguy]

If the disk is unable to spin then the heat build up on R90 could be the result of a higher current draw as it attempts to spin up the drive.

Adrian has a video dedicated to some hard drive servicing. I would recommend watching that video and doing what he recommends. Otherwise it's likely a lost cause.

 

[modem7]

The HD doesn't spin up when plugged to power. I can hear a faint click then nothing.
I assumed a stuck platter/heads, but both spindle and step seem to move freely.
I'm doing all the tests with the interface board disconnected from the hard disk sealed inclosure. I assumed it should still be a valid troubleshooting method while avoiding to damage any possible data from the platters.

Refer to the photo at [here]. +12V being supplied to the small PCB ("Motor Speed Control PCB") is all that is required to get the spindle turning. As soon as the subject PCB gets +12V, it turns the spindle motor. The interface between the subject PCB and spindle motor is as follows, with the "Hall in" signal being speed feedback from the spindle motor, so that the PCB can regulate the spindle motor's speed.



The maintenance manual that I see for the ST-412 shows (in the circuit diagram section) two variations of the subject PCB.

It's R90 from the schematic of the ST-412 model 20110 (attached). It's a resistor by the output of the negative phase A that goes to the spindle ...

Those phases will be going to the stepper motor, not the spindle motor.

 

[ThePresi]

+12V being supplied to the small PCB ("Motor Speed Control PCB") is all that is required to get the spindle turning.

Great info! That calls for some more testing...

Thanks modem7!!!

 

[modem7]

FYI. Related to the spindle is what is shown at [here]; a separate sensor (the silver coloured cylinder in the photo at [here]) and some related circuitry on the main PCB. They do not control the spindle motor in any way. They are simply a mechanism by which the MPU chip on the main PCB can verify that the spindle is turning, and turning at the required speed of 3600 RPM. Failure of that sensor and related circuitry will not stop the "Motor Speed Control PCB" from doing its job of turning the spindle motor.

 

[VintageVic]

Welcome to the forum.

What a coinsidence, my very first post here was about stuck Seagate 412. You can read my results from this thread:

https://forum.vcfed.org/index.php?threads/seagate-st-412-stuck-spindle-not-spinning.73882/

In my case, it was simply stuck (I guess at the bearings) there were no electrical failure!
Stickyness was probably partly caused due to poor storage. The unit had been exposed to dry but cold storage where
it gets -30 celsius during winters.

The solution was surprisingly simple in my case. I opened the pcb and turned the spindle axel with my fingers.
after sufficient rotation the bearing got a bit loose. it did not immediately fix the issue, when I powered up the drive.
I put power on and used my figer to kind of 'kick start' the spindle from narrow hole between pcb and the hard drive chasis.
for very long time the spindle motor had not sufficient power to start turning the disk, but eventually to my own amazement,
it did started spinning. here's a short video of the event, when that happened:

Not sure if you have the exact same issue, but I have had one message thanking me for the hint. There are at least two
drives (mine included) that had the same issue.

 

[ThePresi]

Thanks guys!

Here is an update on the resuscitating of the ST-412.

Last night I focused on the spindle. I have separated the PCB from the rest of the hard drive and followed modem7's insights and schematic. I manually spun the motor like VintageVic did, I deoxited the plug that connects the spindle to the little controller board, applied 12V and it started spinning (I feel it was simply a dirty plug. The drive was filthy.)

The joy of progress was immediately curbed by the horror of the sounds the drive produced. It got to nominal speed, but with screeches, rattles, banging and an eerie noise that sounded like a spring bell from hell. I opened the enclosure, expecting shrapnel and twisted metal, but it looked rather OK on the surface. I may very well have a twisted heads and scratched platters underneath the clean top platter, but I carried on. I sealed back the enclosure and after a few cycles of powering up and down, it could spin fairly well, although inconsistently. I assume dry/bad ball bearings?

After re-connecting the PCB and doing some measurements, I have found no issues with the spindle circuitry. As modem7 pointed out, it is a quite simple circuit after all. I focused back on the burning resistors that are before the connector to the stepper motor (four 20ohm, 1W resistors, R82, R89, R96, R103 - version 20110) they get scorching hot. They are from the same area in the schematic I have attached above. There is clearly an issue in that section. Perhaps a short in the stepper motor? BTW, the stepper doesn’t move during the initialization phase, but it goes back to track 0 during initialization if I manually move it off, which is an encouraging thing.

I checked the transistors arrays (2D) that drive those resistors, but found nothing wrong. Then I measured across those burning resistors to find 4.8V at 25mA (only 0.12W?!) and I went to sleep defeated.

I know these old hard drives are hard to fix and I’m not nearly knowledgeable enough to get decent odds of success, but they are historically important IMO, and I’m compelled in trying to get it in working condition for the sake of preservation, which is the main reason, along with nostalgia, I got into retro computing. So, no, I will not dismantle it and make clocks out of the platters. For now, I set it to the side.
I’m looking online for other broken ST-412, with the hope to combine salvageable parts, and making at least a working unit (holy dollar the price people are asking for those not working drives is insane on eBay!)

Here are today’s questions:
Was anyone successful in dismantling the heads and platters to clean and lubricate the ball bearings?
Does anyone have an idea on what could be the reason for those grilling resistors?

Thanks again for the feedback and suggestions!

 

[bear]

I had an ST-419 with one power transistor (TIP120) out of three bad on the spindle motor driver board. Replacing that allowed the motor to spin again. I believe the ST-412 is arranged similarly.

 

[ThePresi]

I had to share the "bell from hell" sound, it's so weird:

 

[tradde]

That's weird. Never heard a hard drive ring a bell. But the drive does not sound good at all.

 

[VintageVic]

You opened the case where you can see the surface of the plates and the heads?
That is probaly unfortunate as dust in the room air got in and may/will cause issues later.

My drive also gives horrible sound, but still works despite of it.

 

[powerlot]

If the oil in the bearings and stepper is dry the driving circuitry needs to work harder to move it. I once powered up a stuck drive and it smelled like a space heater on the first few power up cycles.

Does the stepper have an exposed shaft where you can drop some oil in? You can also try heating up the spindle bearings (and stepper) with a hair dryer before powering it up.

The bell sound is interesting.. possibly a stuck/broken bearing ball?

PS: Avoid moving the heads when the platters aren't spinning...

 

[Robin4]

If the disk is unable to spin then the heat build up on R90 could be the result of a higher current draw as it attempts to spin up the drive.

Adrian has a video dedicated to some hard drive servicing. I would recommend watching that video and doing what he recommends. Otherwise it's likely a lost cause.

Only dont tap with an hammer or such thing.(on the side). If the head is stuck, then make things even more worse.

Iam always open up the drive, and have a cannister of compressed air with it. So you can blow of the tiny dust thingys on the patters.

So before closing the lit, clean off the platter surface. And hold the lit directly above it. When the surface is cleaned off direcly close the lit again..

You might have to do a LL format again to get the drive going again.

Spots on the surface can be cleaned up with ISOpropanol 99 % (use a cottontip which is blow off with compressed air) and let it vaporate away so no residu is still there.


With my method i already free-ed up 2 stepper SCSI drives 5.25 inch with good working again result (bearing of the platters had a bad spot, because sitting still for circa 25 years)
If de spindle is loose, then dont attach the controller, and leave it turning for a lot hours so the drive can getting warmed up very nicely.

If it still spins after 1 week, then you can couple it back on the controller and see with the drive will do.

If its running again, advise to low level format it again..

 

[Robin4]

You opened the case where you can see the surface of the plates and the heads?
That is probaly unfortunate as dust in the room air got in and may/will cause issues later.

My drive also gives horrible sound, but still works despite of it.

Not if you have air canned compressed air with you.. Before closing again blow off all dust spots. (thats the way it did.)
Those older HDD arent that fragile, in compair with the most newer ones.

 

[Robin4]

I had to share the "bell from hell" sound, it's so weird:

Think the bearing is shot.

 

[modem7]

BTW, the stepper doesn’t move during the initialization phase, but it goes back to track 0 during initialization if I manually move it off, which is an encouraging thing.

Matches the behaviour of my ST-412. See note 8 at the bottom of [here].

 

[modem7]

Was anyone successful in dismantling the heads and platters to clean and lubricate the ball bearings?

Not that I am aware of, but I and some others have 'band-aid' lubricated the spindle via another way. There are a few threads of the subject of spindle re-lubrication (or noisy spindles), and example being the thread at [here].

 

[modem7]

Does anyone have an idea on what could be the reason for those grilling resistors?

Later in the day, I will see how hot those resistors are on my ST-412.

 

[modem7]

Does anyone have an idea on what could be the reason for those grilling resistors?

Later in the day, I will see how hot those resistors are on my ST-412

On my ST-412:

No data nor control cable connected. ST-412 cold. Power applied, then I waited about 20 minutes.

Resistor #1: OV across it, so it is dissipating no power. Cool to touch.
Resistor #2: 3.9V across it, so it is dissipating about 0.8W of power. Quite hot.
Resistor #3: OV across it, so it is dissipating no power. Cool to touch.
Resistor #4: 4.0V across it, so it is dissipating about 0.8W of power. Quite hot.

BTW. Over the 20 minutes, I occasionally measured the voltage. The voltage across resistors #2 and #3 fluctuates. For example, resistor #2 sometimes had as much as 4.5V across it.

 

[modem7]

I had to share the "bell from hell" sound, it's so weird:

Maybe your drive is meant to sound like a bell.
Clearly, the sticker on the top of your drive is showing the Liberty bell.