druck  wrote:
> On 07/01/2020 21:34, Jim Jackson wrote:
>>>> It seems to be this thing:
>>>>
https://www.notebookcheck.net/Review-Iomega-eGO-Portable-500GB-USB-3-0.33106.0.
html
>>>
>>> It's spinning rust drive, that is going to be slower as main OS disc
>>> than any sort of flash based medium. It would be OK for mass data storage.
>>
>> Go with the spinning rust - you might be surprised at the performance.
>> Some people seem to think you always need the very fastest. It all depends
>> on what you are planning to do.
>
> Performance for large sequential files is fine, but for use as a system
> disc you need fast small random access performance, and for that
> spinning rust has always been bad. Hence everyone moving to SSDs on
> their PCs.

If you want super performance matching the latest PCs you wouldn't
start with a Pi in the first place. SD cards are hardly designed to
compete with HDDs for speed anyway, and the OP's issue with them is
long-term reliability not speed.

>>>> I tried to plug it in just now, while the Pi 4B was running with power
>>>> from the wall socket, and it crashed the operating system. The X-server
>>>> crashed, it returned to the text console, and then seemed to lock up
>>>> with some kind of file system errors being printed. It accepted no input
>>>> from the keyboard anymore.
>>>
>>> It probably requires too much power, and caused a voltage dip. You'll
>>> need to use a USB hub to provide the power most types of spinning rust
>>> hard drive require.
>>
>> Quite a lot of modern spinners work perfectly well powered from Raspberry
>> pi's. You do need an adequate power supply. And for the 4B that means one
>> that supply 3 amps.
>
> But this one crashes the OP's Pi.

When he plugs it in while the Pi is already booted. This will be due
to the surge of current drawn by the HDD when it starts to spin up.
When the Pi is running, the dip in voltage is enough to disrupt it.
However when both Pi and HDD are powered on at the same time, the Pi
is presumably in some sort of power-on reset state (a deliberate
delay intended to avoid problems due to power-on voltage supply
instability) so by the time it starts running the HDD has already
got going and is no longer pulling too much current.

So it may not be a problem so long as the HDD is never connected
to a turned-on Pi. If the OP does want to do that, a higher-current
power supply may or may not be the solution. If the problem is due
to the resistance of the 5V lines on the Pi's circuit board causing
a voltage drop, then the beefier power supply will make no
difference. What you want is a direct wired connection from the power
supply output to the USB socket that the HDD connects to. Or
something clever using a capacitor and diodes could work too if done
by someone with knowledge of electronics.

--
__          __
#_ < |\| |< _#

--- SoupGate-Win32 v1.05