On 4/27/2015 2:43 AM, Paul wrote:


This is from a slightly older WB800. Some of the chip markings
are a bit clearer. And I don't see any overshoot protection
on this one, or on your PCB either...

*******

The seven pin chip, with the soldertab on the opposite end,
is a regulator, and makes 2.6V, 3.3V, 8V, and accepts 5V and 12V input.
No datasheet seems to be available (IRU1239SC). Pin for pin sub
is ST3L01K7R from STMicro, and I could get a datasheet for it.

http://obrazki.elektroda.net/44_1178958806.jpg

(Link to a PDF for the voltage regulator)
http://www.digchip.com/datasheets/pa...L01K7R-pdf.php

There doesn't seem to be any protection on that thing. I
was thinking maybe it had shunt regulation on the input
rails (as the regulator is intended specifically for the
disk drive industry). That regulator can withstand up to
18V on both the 5V in and 12v in pins.

Your PCB still contain the classical LC filters for
5V and 12V. But each filter is treated differently. For
example, on your PCB, the inductors are missing. Which
means a parallel path must exist somewhere to provide
continuity via a zero ohm resistor. In the "elektroda"
picture above, one input inductor is present, and the
other one is replaced by a zero ohm resistor. Rather
a lot of "dynamic" design going on here.

*******

One thing you could consider doing, is using your multimeter,
determine whether the thing you connected, could have done
damage or not. Was the pinout actually wrong or not ? It
could be that the drive has stopped for other reasons.
Maybe your power source actually applied correct potentials.

Yet another test case for you... Connect only your
valid power source to the drive, and see (with nothing
connected to the data cable), whether it will spin or
not. You should be able to spin up and spin down a drive,
without a data connection. I occasionally do that
here on my SATA drives, when testing an external and
not wanting any clone drives to interact. I just unplug
the data cabling and leave the power connected.

*******

Power supply design practices, vary from one industry to
another. In the personal computer industry, the
ATX supplies at least, use a common transformer for all the
rails. And rely on turns ratio to establish the voltage
for each output. All the outputs are monitored, and if any
one output falls, the common transformer input
voltage is raised. Which "makes the heavily loaded rail
feel better", but overshoots on voltage on the rails
that didn't need that boost. Antec made at least one
supply, with separate circuits, or so they claimed. I've
also seen a picture of the inside of a high power ATX
power supply, which used separate circuits. An impressive
beast, and several inches longer than a standard ATX.

The adapter you were using, could be similar to ATX, as
there would hardly be room for a lot of transformers
in there. And they could be using a shared transformer
design. What's supposed to happen, is the circuit is
supposed to be designed to meet +/- 5% maximum variation.
And your 12V was well outside the 5% range.

On some rails in computer circuits, there is actually
a fair amount of headroom. Digital circuits don't
take kindly to excess voltage. The original TTL 5V
chips, could handle up to 7V. Later ones couldn't
take quite as much overvoltage. Whereas, circuits
with regulating elements, say powered by 12V,
they might easily take 18V without frying. There
are even three terminal regulators, that can take
50V or so. So not every circuit fries instantly
if outside the normal range. But if you swapped
the 12V and 5V on that drive, then any 5V logic
would be well and truly fried by the 12V.

The last logic family "that could take it", was 4000
series CMOS. Which ran from 5V normally, and
could accept up to 15V (and would run three times
faster while doing so). They don't make sweet logic
like that any more. Designers used to love that
stuff, and even though it was a digital logic family,
some guys would go overboard with the "analog tricks".
For example, with 4000 series, if you needed a timer,
you could hang an RC circuit on a logic chip input, and
the charging of the capacitor was your timer.

I'm suspecting something else has happened to your
drive. At least give the drive a "power up test" without
a data cable, and see how it goes (listen for spinup).
As spinup proves the motor controller chip is
working. The motor is three phase (for low torque
ripple), and a chip makes three phase signals,
controls acceleration of the spindle and so on.

Paul,
the tests I did WERE done without data load as I did not connect the USB
port to anything. Only the 4-prong Molex connector was connected.
Yet the drive did not stir.

I was wondering what, if any, information could I gain if I measured the
voltages on those 4 finger-like spring contacts that provide the
connection between the HD and its PCB. I could probably slide some
flattened wires between the connectors to provide test leads for the
multimater. I just don't know the polarity on those 4 fingers. Do you?

Thanks again.