Hi everyone,

I am a student currently working on a personal project for educational purposes. I've recovered an old Seagate Barracuda ST3000DM001 (P/N: 1E6166-570) that is not working, and I am using this as an opportunity to learn basic electronic diagnostics.

To be completely transparent, I have no formal electronics training. I am working with an old multimeter and an old oscilloscope, and I am trying to learn how to troubleshoot a PCB. I have attached a few photos of the board.

Here is what I have observed so far:

Power: 12V and 5V are correctly reaching the PCB via the SATA connector.

Motor: The motor does not spin at all (no sound). When I probe the 3 motor pins (downstream from the SMOOTH chip) with my oscilloscope, the signal remains flat while the PCB is powered.

Components: I tried checking the 12V and 5V TVS diodes with my multimeter. They appear "OK" based on my basic readings, but I am not entirely confident in my methodology. Is there a better way to test them, perhaps using an oscilloscope?

What are the "must-check" points on this specific model?
Are there any common failure points known for this board that I should inspect with my multimeter?
Should I be actively hunting for shorts?

Any guidance or tips to help me progress in this diagnostic would be greatly appreciated. Thank you in advance for your help and for your indulgence with a complete beginner.

Best regards,
Zebananos

What voltages do you measure across each of the two TVS diodes?

_________________
A backup a day keeps DR away.

Next to the SMOOTH chip, there are large gray coils with two terminals. They should show something, about a volt or more. If all coils show 0 volts, it's worth checking whether the voltage is reaching the TVS diodes. Repairing such boards usually ends with removing the diodes shorted by voltage surges and replacing the burnt-out resistors near the power connector with jumper wires. If the drive still refuses to spin up, the board is usually thrown in the trash and replaced with a working one from a donor drive. No one thoroughly examines these boards, so you'll have to be the first.

There's also an 8-pin chip on the board; this is a ROM. One of its contacts should show 2.5 or 1.8 volts. If the contents of this chip are erased or damaged, the drive also won't spin up.

Judging by the obvious mechanical damage to the board (one of the inductors is broken), this board will never work again - perhaps the contacts under the smooch chip were broken by impact.

I'd also like to point out that these boards have very few components. You can check the capacitors for shorts and measure the resistors (ideally, using another similar, working board as a reference). However, they rarely fail. If the SMOOTH chip had accessible pins, you could measure their resistance relative to a similar chip on another, working board. But that's not the case.

I beg to differ. The board will still work if the coil is not open circuit.

The group of 3 coils are associated with buck converters for Vcore and Vio. The 6R8 coil and the associated diode (E0/U4) constitute the negative supply (-5v) for the preamp.

_________________
A backup a day keeps DR away.

Hi fzabkar and others,

Thank you for the guidance — here is my full set of measurements. I am using an old Metrix MX1 multimeter and a 2000s-era oscilloscope, so please treat these values as approximate.

Coils continuity (unpowered, diode/ohm mode): All 4 black coils beep, indicating continuity on all of them, including the visibly damaged one. Electrically, the windings appear intact (not open).

Coils voltage (powered): 0V on the two unmarked coils and on the "6R8" coil; only ~300mV on the "3R3" coil.

TVS diodes (unpowered): OL in one direction, ~200Ω in the other, identical for both TVS (see attached TVS_1 / TVS_2). When powered, I measure approximately 5.02V and 12.4V across them respectively.

ROM (8-pin, powered): Two pins around 600mV, one at approximately 400mV, the rest at 0V (with noise). None show the expected 1.8V or 2.5V.

Short circuits: No continuity beep between ground and 12V, or ground and 5V.

E0/U4 diodes (unpowered): Open circuit in one direction, ~50Ω in the other (see attached E0_U4_1 / E0_U4_2).
Note: My multimeter beeps in diode mode for these E0/U4 diodes, but does not beep for the TVS diodes. Is this difference normal or does it indicate a specific issue?

What do you think of these results? I hope it's not the SMOOTH chip that's malfunctioning. I've attached a close-up photo of the SMOOTH chip and coils.

Furthermore, I have the impression that the varnish has overheated on the trace connecting the SMOOTH chip to the 3R3 coil. Does this seem suspicious to you?

Thank you in advance for your time and expertise.

Best regards,
Zebananos

It does look like a SMOOTH failure. The most expedient solution would be to move the 8-pin ROM IC to a donor PCB. However, be aware that if you damage this IC, data recovery will become extremely expensive, if it is possible at all. PCB vendors such as hdd-parts.com include a free firmware transfer in the price.

_________________
A backup a day keeps DR away.

Overheated varnish looks different - it takes on a dark yellow or brown tint, depending on how hot it was. I can search my archive for photos of what this looks like if you're really interested, but what you see in the photo isn't evidence of overheating, but corrosion - moisture has gotten under the varnish. Scrape it off with something sharp and metallic, and the varnish will easily peel off, revealing oxidized copper underneath.

I think the board is completely destroyed, beyond repair. Why is that? The ROM chip should have 1.8 volts on one of its contacts. Why exactly that much? Because next to the ROM chip, there's another rectangular chip without pins, made by Samsung. This is DDR2 RAM. It also operates at 1.8 volts and must be powered; without working RAM, the microprocessor won't work. To reduce production costs, chips were selected that operate at a single voltage, which is standard practice for such boards.

One of the required voltages is missing. More precisely, there are no necessary voltages on the board at all, except for those coming from the power connector. The SMOOTH chip is responsible for all the voltages and all the hard drive components that need to be moved. That's why it's suspected of being faulty. It's possible that it's not the chip that's faulty, but rather a component nearby, but for further diagnosis, we need to measure what's feeding this chip.

This chip doesn't have any pins accessible for measurement. There are small yellow elements near the chip. On some of them, at one of the contact pins, there are 5 volts and 12 volts coming from the power connector. You can also take measurements at the yellow dots with drilled holes if you cannot find the voltage on the capacitors. Considering that there was a tiny voltage on the ROM chip, these voltages were likely coming into the chip.

And now the real problems begin.

This is an internal measurement, so it's difficult to tell whether the diodes are intact or not, but common sense dictates that these diodes don't usually fail, and certainly not two at a time.

The real repair begins with a visual inspection. We've already seen the damage in a place where it shouldn't have been, so the procedure begins by checking all tantalum capacitors for short circuits and resistors for deviations from their nominal resistance.

Since, again, all these measurements are performed internally, some resistors will have unusual resistance values, just like capacitors, even though they shouldn't conduct current. The problem is compounded by the fact that an analog meter isn't very suitable for these measurements. Of course, you can diagnose a board with it, but you'll have to use the same measuring device to take measurements on a working board. I probably have a working board with similar components, but if I measure it with my digital multimeter, the readings will differ from the analog one because these devices supply different voltages for measurement, which will lead you down the wrong path.

In a real repair, the quick and correct step is to replace the SMOOTH chip with a working one, or install the suspected faulty chip on a working board. Further actions depend on the outcome - either the faulty chip was the problem, or the board itself. If the board is the problem, then all components and contacts are checked against the working one, and the discrepancy will sooner or later be found, and with it, the source of the problem.

What do we have in the end? We can't take measurements on the board because we don't have the right equipment, there's no donor board, and there's no way to quickly diagnose the board by replacing the suspected chip. I'm missing the point that this particular chip on this board is difficult to replace because it's soldered onto lead balls underneath.

As a last resort, it would be worth looking at the back of the board for damaged traces.

This was meant to be a learning project, not a case of data recovery or disk operation (yeah, ST3000DM001, the old legendary Grenadas that failed en masse due to low-quality manufacturing), and judging by the condition of the drive and the fact that the sticker where the screw of the magnetic head unit is hidden was torn off, the hard drive had already been taken from a trash heap.

As a final, easy-to-fix cause of the malfunction, you can check the integrity of this trace at the locations marked in the image. Also, look for the hole on the back of the board and measure whether there's contact between the two sides of the board. Test your luck; the problem might simply be a slightly rotted trace or a corroded via.

Thanks for your replies. I was hopeful after cleaning that via—which was indeed full of oxidation—but unfortunately, it didn't change anything.

I'm attaching a screenshot of the measurements I retook a little while ago, though I'm not sure they provide much extra information.

Incidentally, the broken coil is starting to come loose. I initially didn't think it was the root cause of the problem since it still passes the continuity test (it beeps). Still, could a damaged coil be the source of the whole issue despite the continuity?

I'm not really sure how to clean the board any further. That via was the only visibly oxidized spot, and the rest seems minor. At this point, it must indeed be a deeper issue with the SMOOTH board, and I think I've reached the limit of what I can test on my end...

Thanks again for the help!

PS: The points marked with "?" correspond to voltages that dropped from 12V to 0V during my measurement (capacitors discharging?).

Not really. The SMOOTH chip has the necessary voltages near it, but it doesn't output anything. You could also try using an oscilloscope to measure the coils on the contacts facing the chip. If pulses appear briefly there immediately after power is applied to the board, it's still trying to function.

You can try household chemicals, such as a rust and water stain remover (the best solution available for rotted circuit boards), and then wash it with dishwashing liquid. Shake out any remaining water (it will be hidden under the chips) and let the board dry. Considering the board is already completely destroyed, you can freely experiment with it.

Maybe yes, maybe no. Considering the board was rotted and damaged, it was clear without any diagnostics that it was all over. As a bonus, I found an identical working board and measured all the resistances and voltages on it. Keep in mind that my multimeter had a dead battery and the probes were broken, so the readings may be slightly different, especially the voltages. Where the image shows 2.19 volts, it's actually 1.8 volts. And where it's slightly more than a volt, it's actually a volt or slightly lower. You can compare it with your board if you're willing and have the time.