3
\$\begingroup\$

I recently had someone explain to me that he was told he fried a laptop's hard disk by putting it in an external drive enclosure that was not rated to supply sufficient current. (Apparently the disk nominally needed 1000 mA and the power supply was only rated for 500 mA.)

I'm having trouble imagining the chain of events that would result in that letting out some component's magic smoke.

Seems to me that if the drive tried to draw too much current, the power supply's output voltage would sag.

Are there any common modes where an undercurrent results in a fried circuit? My imagination isn't sufficient to come up with any.

\$\endgroup\$
2
\$\begingroup\$

This is just a hypothetical scenario. Don't know if it applies to hard disks.

Suppose there is a control loop in a system, which actively keeps the system out of some harmful condition. Due to insufficient supply current, the supply voltage sags (as mentioned in the O.P.). Due to the sag in supply voltage, control system can't work properly. It can't keep the system out of harmful condition any more. So, the harmful condition occurs and does permanent damage.

I'd say that something like that would be a design flaw.

P.S. Since O.P. is about PC components, I wonder if it's more of a serverfault question?

\$\endgroup\$
1
\$\begingroup\$

There is a somehow famous bug in the AVR Dragon which causes a step-up voltage converter to fail due to overcurrent when the Dragon is plugged into an USB port that does not supply the required minimum voltage of 4.15 V.

\$\endgroup\$
0
\$\begingroup\$

Fried, or just couldn't read the data anymore?

I have a disk which works intermittently off a single USB (500mA) connection. If that were writing some critical bits of filesystem data when it decided to go "intermittent" that could make the drive look fried, but a reformat (or even scandisk/fsck) could potentially resurrect it.

\$\endgroup\$
0
\$\begingroup\$

I recently had this problem with a power supply module.

My first version was simply a third party DC/DC brick on a PCB with some passive filtering.

If the brick didn't have enough input power for the given load, the output would start wildly oscillating between fully on and off, quite quickly. It produced an audible squeal while it did this. Given that it was a 60V output, that probably massively violated the ripple specs on some caps downstream. It never blew the recommended fuse, either.

I recently revved the board to add an MSP430 micro to monitor the input voltage and current. It now cuts the brick power when this occurs, and coordinates with the system controller to organize a soft restart of the load.

Also... Imagine how MOSFETs might behave. A totally closed MOSFET burns no power, and a fully open MOSFET burn very little power due to the resistance being quite low. However, as the gate voltage crosses it's non-saturated region (is that the right term?), the effective resistance can be quite high and it can burn up quite a bit of power. So... if due to some odd quirk a sustained low system voltage produces some intermediate MOSFET gate voltages... I can see how that might burn out some stuff.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.