Locutus wrote on 2023-07-29, 14:05:

I kind of 'hit the wall' as 15Mhz clock is missing at PLC 11741-502...

Any ideas ?

• Verify that 11741-502 gets power, likely intended to be +5V on the top right pin at the 35V capacitor, and possibly GND at the pin on the bottom left connected to the big trace.
• Verify that the "3011" resistor upwards/right of the crystal isn't shorted
• Verify that the cap at the low right end of the crystal isn't shorted.

I expect the 11741-502 to have in integrated oscillator circuit that should start up as soon as power is applied. If that is the case, the oscillator not starting means that there either is no power, something connected to the oscillator pins (the two top pins at the right) is broken or the 11741-502 itself is toast.

What system are you testing this drive with?

Locutus wrote on 2023-07-29, 14:05:

So far I've pinpointed and replaced 7 bad 'resistors' on PCB.

Did you put 0 ohm jumper resistors all over the PCB? Because frankly a lot of these don't make any sense to me...

EDIT: I found a somewhat decent photo of unmodified PCB here: http://hugold.pl/gratym0121/st157a1.html
The resistor next to 15MHz crystal is not 0 ohm. Nor is that one near L2 cap (or is an inductor) on the other side a jumper either. Nor the two that form some voltage divider in lower left corner (although there is a zero ohm in place you seem to have 100 ohm) , and same thing in the upper left corner. Was this PCB perhaps a playground for someone to learn how to remove SMD parts, maybe?

mkarcher wrote on 2023-07-29, 14:59:

I expect the 11741-502 to have in integrated oscillator circuit that should start up as soon as power is applied. If that is the […]
Show full quote

• Verify that 11741-502 gets power, likely intended to be +5V on the top right pin at the 35V capacitor, and possibly GND at the pin on the bottom left connected to the big trace.
• Verify that the "3011" resistor upwards/right of the crystal isn't shorted
• Verify that the cap at the low right end of the crystal isn't shorted.

I expect the 11741-502 to have in integrated oscillator circuit that should start up as soon as power is applied. If that is the case, the oscillator not starting means that there either is no power, something connected to the oscillator pins (the two top pins at the right) is broken or the 11741-502 itself is toast.

Thanks for an answer.
Voltages are OK. Resistor, Capacitor are too.
I suspect that 11741-502 itself is gone 🙁

Deunan wrote on 2023-07-29, 17:07:

Did you put 0 ohm jumper resistors all over the PCB? Because frankly a lot of these don't make any sense to me... EDIT: I found […]
Show full quote

Locutus wrote on 2023-07-29, 14:05:

So far I've pinpointed and replaced 7 bad 'resistors' on PCB.

Did you put 0 ohm jumper resistors all over the PCB? Because frankly a lot of these don't make any sense to me...
EDIT: I found a somewhat decent photo of unmodified PCB here: http://hugold.pl/gratym0121/st157a1.html
The resistor next to 15MHz crystal is not 0 ohm. Nor is that one near L2 cap (or is an inductor) on the other side a jumper either. Nor the two that form some voltage divider in lower left corner (although there is a zero ohm in place you seem to have 100 ohm) , and same thing in the upper left corner. Was this PCB perhaps a playground for someone to learn how to remove SMD parts, maybe?

Hi.
All SMD components appear to be factory fitted (glued and 'same' soldering). I started by replacing them with identical parts (all '0 resistors' were open - question is why...).
Here is a photo of PCB before replacements:

I've also noticed this bypassed inductor 😉
In newer ST-157I, this crystal is 'loaded' witch capacitor through 300 ohm resistor...
https://content.invisioncic.com/r322239/month … cf7e029126.jpeg

Locutus wrote on 2023-07-29, 19:23:

All SMD components appear to be factory fitted (glued and 'same' soldering). I started by replacing them with identical parts (all '0 resistors' were open - question is why...).

I've never seen anything like this. I can only offer some wild theories:
1) these acted like fuses and went open during some serious power event (12V on 5V line, or reverse polarity)
2) went open due to age and manufacturing defects
3) actually were meant to be open, 0 is misleading

1 - That would destroy the electronics and I would think such big resistor and thin traces (like near J6 conenctor) would result in damaging the copper before them going open, or at least there would be some traces of such event (the PCB has plenty of intact tantalum caps)
2 - All of them? Seems... unlikely. I mean it's not exactly difficult to make a 0 ohm resistor of that size, and there is no visible mechanical damage either
3 - That would be misleading but such resistors might be put on to pass automated inspection that looks for missing parts, and yes I realize I'm grasping at straws here

Locutus wrote on 2023-07-29, 19:23:

In newer ST-157I, this crystal is 'loaded' witch capacitor through 300 ohm resistor...

Yes, but it's also like that on the photo fromt the page I found (BTW that page opens the image scaled down, download it or copy to some image viewer program and you'll get much nicer photo).
The value is "3010" so 301 ohms I take it. And the date code on that PCB is very close to yours (9027 / 9029). Both cary "106" stamp in the upper right corner where the model/rev print is. BTW the master/slave jumpers there, have you tried removing those? It looks like your current confing is master+slave and the early ATA drives really do not like that if the slave is absent. Weird things happen in such cases, no-spin is one of them.

These 0 ohms are very suspicious, I realize this is partly analog stuff but I have never seen them used like this. Since the PCB doesn't work as-is, why not try putting some resistors that match the other HDD? At least around the crystal chip and maybe the upper-left corner too (it's a stepper servo driver I think?).

Deunan wrote on 2023-07-29, 21:49:

I've never seen anything like this. I can only offer some wild theories: 1) these acted like fuses and went open during some ser […]
Show full quote

Locutus wrote on 2023-07-29, 19:23:

All SMD components appear to be factory fitted (glued and 'same' soldering). I started by replacing them with identical parts (all '0 resistors' were open - question is why...).

I've never seen anything like this. I can only offer some wild theories:
1) these acted like fuses and went open during some serious power event (12V on 5V line, or reverse polarity)
2) went open due to age and manufacturing defects
3) actually were meant to be open, 0 is misleading

1 - That would destroy the electronics and I would think such big resistor and thin traces (like near J6 conenctor) would result in damaging the copper before them going open, or at least there would be some traces of such event (the PCB has plenty of intact tantalum caps)
2 - All of them? Seems... unlikely. I mean it's not exactly difficult to make a 0 ohm resistor of that size, and there is no visible mechanical damage either
3 - That would be misleading but such resistors might be put on to pass automated inspection that looks for missing parts, and yes I realize I'm grasping at straws here

Locutus wrote on 2023-07-29, 19:23:

In newer ST-157I, this crystal is 'loaded' witch capacitor through 300 ohm resistor...

Yes, but it's also like that on the photo fromt the page I found (BTW that page opens the image scaled down, download it or copy to some image viewer program and you'll get much nicer photo).
The value is "3010" so 301 ohms I take it. And the date code on that PCB is very close to yours (9027 / 9029). Both cary "106" stamp in the upper right corner where the model/rev print is. BTW the master/slave jumpers there, have you tried removing those? It looks like your current confing is master+slave and the early ATA drives really do not like that if the slave is absent. Weird things happen in such cases, no-spin is one of them.

These 0 ohms are very suspicious, I realize this is partly analog stuff but I have never seen them used like this. Since the PCB doesn't work as-is, why not try putting some resistors that match the other HDD? At least around the crystal chip and maybe the upper-left corner too (it's a stepper servo driver I think?).

🤣, I have to wake up very early today, but you gave me REALLY intriguing theory I didn't even consider .
In more modern HDDs I've seen 'zero resistors' acting as fuses/croses.
But, I had to try theory #3...

I've almost completely removed 'my job'...
Left only one '0' resistor' paralell to inductor because when I was examining PCB, I've noticed that this inductor 'is open' (but I thought to myself 'why bother it is shorted by '0' resistor anyway).

Some tinkering with non SMD 1k resistors and I've 'clock' !!!
One step further - THANKS !

BTW. Jumpers are set for 'Stand alone drive'.

Locutus wrote on 2023-07-29, 23:10:

Left only one '0' resistor' paralell to inductor because when I was examining PCB, I've noticed that this inductor 'is open' (but I thought to myself 'why bother it is shorted by '0' resistor anyway).

I would desolder it and check in LCR meter. If it shows a standard capacitance (*) then I'd assume it's a cap and not a choke. I really can't tell from the photo and Seagate apparently decided to save some pennies on silkscreen layer.
*) This PCB seems to using 1% resistors so I would assume the capacitors are also pretty close to nominal values (though some might be chosen to have exact, but "werid" capacitance)
In general after seeing these "0 ohm" resistors I would assume nothing about the rest of the parts unless I could easily tell and measure to be sure. Seeing how the zeros and blue resistors seem to be different between the PCBs I'm thinking these might be selected to match particular chips (or mechanical parts, like say bias points for the heads, etc).

Locutus wrote on 2023-07-29, 23:10:

BTW. Jumpers are set for 'Stand alone drive'.

My bad, I misunderstood the jumper table I saw on the net.

I think I mentioned this already but make sure the 12V power you are applying to the drive is not current-limited too hard. I have older (5.25") drives that need 4A or so to overcome the inertia of the still platters (and somewhat sticky main bearing balls at this point in life). A well-worn 3.5" drive of that era migt still pull spikes of 2A or more to start. At first there are some timed high current spikes applied to the motor to get it going, and only then it goes into closed-loop rpm regulation. If that first phase does not impart sufficient angular momentum on the platters then the second phase will fail and on some drives might not even retry until power-cycled.

Deunan wrote on 2023-07-30, 10:59:

I would desolder it and check in LCR meter. If it shows a standard capacitance (*) then I'd assume it's a cap and not a choke. I […]
Show full quote

Locutus wrote on 2023-07-29, 23:10:

Left only one '0' resistor' paralell to inductor because when I was examining PCB, I've noticed that this inductor 'is open' (but I thought to myself 'why bother it is shorted by '0' resistor anyway).

I would desolder it and check in LCR meter. If it shows a standard capacitance (*) then I'd assume it's a cap and not a choke. I really can't tell from the photo and Seagate apparently decided to save some pennies on silkscreen layer.
*) This PCB seems to using 1% resistors so I would assume the capacitors are also pretty close to nominal values (though some might be chosen to have exact, but "werid" capacitance)
In general after seeing these "0 ohm" resistors I would assume nothing about the rest of the parts unless I could easily tell and measure to be sure. Seeing how the zeros and blue resistors seem to be different between the PCBs I'm thinking these might be selected to match particular chips (or mechanical parts, like say bias points for the heads, etc).

Hi there.
'L2' indeed must have been an inductor (it is 'open now ').
With bypass '0 resistor' the drive spins up, can hear 'seek' but:
- it is not detectable
- it makes really bad noises (screeching, scratching...).
Looks like this drive is also mechanically damaged, probably opening it can't make more damage 🙁

Locutus wrote on 2023-07-31, 21:43:

With bypass '0 resistor' the drive spins up, can hear 'seek' but:
- it is not detectable
- it makes really bad noises (screeching, scratching...).

That sounds like crashed heads. Could also be the magnetic material flaking off but the results is the same and it's very hard to tell afterwards what happened first.
You can try running it upside-down, if the noise goes away (or there's less of it) it means only top surface (of platters) is damaged, and not very badly - but such pits can also damage the head. There's going to be bad sectors even in the best case scenario.

Go ahead and open it - these older drives can deal with a little bit of dust contamination pretty well, no need for special hood, though usually there's nothing to rescue. I've yet to see a HDD that could be repaired by opening it...

Deunan wrote on 2023-07-31, 22:22:

That sounds like crashed heads. Could also be the magnetic material flaking off but the results is the same and it's very hard t […]
Show full quote

Locutus wrote on 2023-07-31, 21:43:

With bypass '0 resistor' the drive spins up, can hear 'seek' but:
- it is not detectable
- it makes really bad noises (screeching, scratching...).

That sounds like crashed heads. Could also be the magnetic material flaking off but the results is the same and it's very hard to tell afterwards what happened first.
You can try running it upside-down, if the noise goes away (or there's less of it) it means only top surface (of platters) is damaged, and not very badly - but such pits can also damage the head. There's going to be bad sectors even in the best case scenario.

Go ahead and open it - these older drives can deal with a little bit of dust contamination pretty well, no need for special hood, though usually there's nothing to rescue. I've yet to see a HDD that could be repaired by opening it...

I don't know whether to laugh or cry...

If this was MFM drive and some heads were still good the disk could be used at reduced capacity, but it would require a mod, or a PAL/PLD to "remap" the existing heads to 0-N range. No such luck with IDE or SCSI though... I've got 5 or 6 damaged HDDs myself (early SCSI and IDE) and frankly very little hope of ever fixing any of them. I need parts from donor drives and these days everthing old and "collectible" is expensive, even when clearly damaged, and I'm not paying these prices. I'm about to attempt PCB swap for ST3290A that kinda works but can't seem to init properly most of the time after power-on or reset. Still, it could be an internal short in the voice coil actuator for example, so I'm not expecting much. The donor is supposed to spin and mostly work too but has bad sectors - we'll see.

Curiously enough these heads look like servo heads on more modern drives like ST1126N - while the actual R/W heads got smaller to allow higher bit density.