partially correct. your logic is sound, your experience and lack of parts to try out is holding you back...lets start
I don't care about what other people write on the internet.
you should read what the correct people have to say. the tricky part is finding the correct people.
I tested the same drive (5400rpm) on the same notebook with three different enclosures (yes, I have done this), get the following results:
1. Enclosure A spins up, spins down and doesn't do anything further.
2. Enclosure B spins up, spins for a few seconds, spins down and spins up again. Repeat repeat repeat.
3. Enclosure C spins up, stays spinned up but preduces nothing but errors in my logs. After a few seconds the errors stop, and after this I notice my notebook's USB is screwed.
Enclosures A, B, and C all work normally on a PC. Plugging either of the above enclosures with the 5400rpm drive into my Mac, OS X gives me an error telling me that my USB port cannot produce enough power for the attached device.
Replacing the 5400rpm drive with a 4200rpm, allows each of the enclosures to work normally on my notebook and on the Mac.
Plugging the 5400rpm drive, in any of the enclosures, into a powered USB hub, enables it to work properly.
your testing is thorough, given the parts you have to work with, however you do not have a broad enough range of parts to draw your conclusions form with 100% accuracy
This leads me to the following conclusions:
1. Some machines don't give enough power through a single USB port to power a 5400rpm hard disc.
partially correct. if we ignore the 5400rpm part of your statement I wouldve given you full marks. failing chipsets are via and sis (older versions...), they very often do not supply the proper 500ma per root hub the spec demands
2. Plugging in another USB cable for power (espcially when it's a notebook with only two USB ports) might be pushing the limits.
again, partially correct. for most notebooks plugging in a second usb cable for power will achieve 100% buggerall, as it is 500ma per root hub, not per port. And most laptop usb ports (the external ones) ar eon the same root hub...so the 500ma is shared amongst the external ports
3. Pushing the limits might damage your hardware (as is often the case with electronics).
the spin-up spin-down, cluck cluck cluck spin-up spin-down effect you noticed is the drive not getting enough power to spin the platters up. In that state, it is trying to request "more power" than the root hub can handle. however, I am sure you noticed the usb port is still working, smoke didnt come out.
4. Some hardware is designed to cope with a scenario where a device needs more power than the port can supply, by shutting down the port or the device or whatever is necessary.
correct. in fact, windows xp will handle this all by itself, and when it happens you get a message on your taskbar letting you know it has happened. if you dont get the message, then it didnt happen
If you have a better explanation, I'm all ears.
your conclusion was drawn along the lines of 4200rpm vs 5400rpm. the hassle is, the spindle speed of the hdd (in isolation) has got little to do with current draw. A hard drive demands its maximum current draw at spin up time..to keep the platters spinning requires little, but to get them spinning requires a lot. If you google the specs of the hard drive you will find pdf on the manufacturers site stating the current draw at spin up. You made the mistake of assuming a 4200rpm drive obviously draws less current than a 5400 rpm drive. This is incorrect as a blanket statement, it depends on the individual drive model. Modern drives draw less pwoer than older drives, so conversely I can show you a 7200rpm drive that works fine in external enclosures, yet a 4200rpm drive that I may have wont work. As stated before, dont make blanket statements based on two individual ahrd drives you happen to have tried.
An externally powered usb hub will solve all these problems.
P.S. to the firewire dude, ...firewire negotiates the transfer speed AND the voltage required, and can go up to 40v per port.
I damaged my HP's motherboard in the same way. It still works but the USB bus is no longer reliable, which comes down to the same - I have to replace the motherboard.
