Sign in with
Sign up | Sign in
Your question

USB Disk restarts after memory testing

Last response: in Motherboards
Share
August 11, 2009 10:05:38 PM

Hi everyone. I am very desperate about the problem I have faced. I could not find out the reason of the problem.

After many controlled experiments, I can explain the problem with the following steps:

1. usb disk is connected & vista is running
2. shut down vista & computer (there is still electricity on the system)
3. press power button on the case
4. memory test is done in black screen
5. computer restarts while checking storage devices in black screen
6. goto 4

BUT, problem does not occur in the following steps.

1. usb disk is connected & vista is running
2. shut down vista & computer (there is still electricity on the system)
3. turn off PSU (there is no electricity on the system)
4. turn on PSU
5. press power button on the case
6. no problem & vista is running

1. usb disk is connected & vista is running
2. shut down vista & computer (there is still electricity on the system)
3. un-plug and re-plug usb disk (electricity goes out for a while on the disk)
4. press power button on the case
5. no problem & vista is running

As a conclusion, when I shut down the system and don't cut the electricity on the usb disk, my computer restarts in an infinite loop.

NOTE:
NO PROBLEM WITH DIFFERENT COMPUTERS AND THE SAME USB DISK.
THE USB DISK DOES NOT HAVE ADDITIONAL POWER CABLE. (ONLY USB CABLE)

What kind of thing can lead such problem?

Thanks for your help.

August 12, 2009 7:19:16 AM

Thanks bilbat.

I have read your post. The problem mentioned there is like my problem.
But interesting thing in my problem is that when there is electricity from previous shutdown, restart loops occur; otherwise no problem. And disabling "Legacy USB storage detect" seems to solve the problem.
However, this behaivor of GB motherboards can be a protective structure or a really annoying defect.

a c 177 V Motherboard
August 12, 2009 3:15:09 PM

Quote:
However, this behaivor of GB motherboards can be a protective structure or a really annoying defect

I hesitate to call these problems a defect - I lean toward the fact that the boards are simply 'finicky' - they demand absolute adherence to the actual USB specifications. The reason I believe this to be the case is that the problems seem to occur in 'batches' - some device hits the market, and a number of people have troubles with the same device. The latest 'batch' (three reported episodes, so far) has been WD external backup drives, both 1 and 2TB models; the previous spate was a Rosewill external DVD drive - and that one was eventually fixed by a later firmware release from Rosewill. I can't remember whether USB made its debut during Windoze 3.11, or 95, but whatever - back when it first appeared, one was lucky if every third USB device purchased would actually 'hook up' and work!

The issue here, like for video cards, is 'design by reference'. In the case of video cards, ATI releases a new GPU, and, with it, a 'reference design' for a card to utilize that GPU. There are three issues at stake with video cards: 1, the reference will work; 2, the reference enables ATI drivers to work, for the whole family of cards; 3, accessories (like water cooling blocks) will fit on the 'cards to reference'... I love Sapphire's implementations of ATI cards, but they don't understand the concept of 'reference' worth a damn - at least half their cards are not 'to reference'. With USB, what happens is that a small number of foundries (and as an aside, isn't it interesting that the term 'foundry' has been in nearly every language for literally millenia, as the Greeks and the Chinese have been 'pouring bronze' since about the dawn of time - and now it's been expanded to include people 'growing' and etching silicon wafers!) make USB controller chips, and offer reference designs (and firmware) that ensure the chip will work - but, as I say in my business (industrial systems design), "nobody's happy with the flavor 'till everyone's 'peed in the soup'!" So, they start out with something working, and then 'make improvements' until it's 'semi-broken'. Thus, all in all, I don't think the blame for this can be laid solely at Gigabyte's feet...
Related resources
!