Someone with more experience can answer but every computer I have ever used does not treat external hard drives the same and to my knowledge its not something that can be BIOS-adjusted. One PC allows for one USB cable and others need two, almost regardless of what the external hard drive instructions say. Actually, if people bothered to READ those instructions they often say "maybe you need one connection, maybe two" and thats why many of the MFRs included double cables. One hard drive I bought came with an external AC power adapter with the words "maybe you need it, maybe you dont."
Not all PCs come with the same power rating out of each port and so that must be an ongoing headache for the MFRs of hard drives and other components. By the looks of it your NUC is operating fine, it just needs you to use two ports. Had you, like me, had a hard drive with an extra power adapter you plug into the wall, you could use one port; if not, you may or may not have to use two. I dont see that as a fault of the NUC but just how the hard drive MFR made the device.
Thank you for joining the Intel communities.
I have to agree with cwv, there is no firmware update that we can provide; you could try a active hub for you to get better power to the external hard drives.
USB 3.0 ports maximum current is 900 mA for each blue port, 1.5 A for the orange charging port.
USB 2.0 ports maximum current is 500 mA for each port of the white header (1 A total).
It is possible that your hard drive requires more than that and it needs extra power adapter.
thanks for your answer and cwv,
i am however really surprised because my 2 hard drive are designed to work without external power supply. it's the only computer that i use that require 2 usb port.
moreover, I also own a nuc dn2820fykh0 and I can use my 2 usb drive with only one usb port without any problem.
Does the usb power specification changed between the nuc dn2820fykh0 and the nuc5ppyh ?
I know it is not a critical problem as i can use a active usb hub but I am really surprised.
thank in advance for your answer.
Thanks for posting that because my answer didnt take into account or assume any differences between NUC platforms so Intel needs to explain why it varies. Reasonably, it shouldnt. Hmmm.
Meantime in the next few weeks i'll be able to try my 5i3RYH with four different external hard drives, each with different requirements and lets see what happens. By then I will have a GORITE expansion lid with USB 2.0 ports so I will be able to test the drives with 2.0 and 3.0 ports as well.
Without your post I would have stood on my answer which for now I have to take back! Makers of external drives dont specify requirements, at least I have not seen it, they dont say you must have 900 mA (or 500 mA) available. The just include cables to hopefully avoid tech support calls, i guess.
So going back to the original post, is it NUCs fault or that a user has to find a drive that will work with a NUC? When we buy external drives that are inside boxes we dont look under the hood first to see what on box's circuit board, we just assume they work.
Hope im not digging a hole here.
Here is some useful information for you.
"Computer systems that have an unshielded cable attached to a USB port may not meet FCC Class B requirements, even if no device is attached to the cable. Use a shielded cable that meets the requirements for full-speed devices. "
See page 21 on TPS
That is note in the technical product specifications on page 21 that I thought it could be useful for you.
Today I was testing this unit with an external hard drive and I was able to write on hard drive with no problem.
There is a possibility that this is an isolated issue with your unit and it needs to be replaced,however, I would like to ask you if you have tried to enable write caching?
Go to device manager and do right click on your hard drive, Click/tap on the Policies tab, and select enable write caching on the device, you can try by disabling as well if it is already enabled.
To prevent data loss, do not check Turn off Windows write-cache buffer flushing on the device unless the device (disk) has a separate power supply (ex: UPS) that allows the device to flush its buffer in case of power failure.
I’m just following on this thread to ask you if your issue was resolved with the information provided above or if the problem continues.
Here's a set of general rules:
- If you are using a 5" HDD, it absolutely requires a separate power supply regardless of what NUC and what USB port.
- If you are using a 2.5" HDD and are plugging it into a (blue or yellow) USB 3.0 port, no separate power supply is necessary.
- If you are using a 2.5" HDD and are plugging it into two (black) USB 2.0 ports (i.e. special cable that has two USB plugs on one end), no separate power supply is necessary.
- If you are using a 2.5" HDD and are plugging it into a single (black) USB 2.0 port, a separate power supply is usually necessary.
Here's background information:
- Typically, 5" HDDs need 20W whereas 2.5" HDDs need less than 4W.
- 5" HDDs require both 5V and 12V inputs, whereas 2.5" HDDs need only 5V.
- Blue USB 3.0 ports provide 900mA (=4.5W) - which is enough for a 2.5" drive.
- Yellow charging ports (USB 2.0 or USB 3.0) provide 1500mA (=7.5W) - which is plenty for a 2.5" drive.
- Black USB 2.0 ports provide only 500mA (=2.5W) - which might not be enough for a 2.5" drive
Ok, as I mentioned before, i think it depends on the drive, MFR, and whether or not there is any consistency in the amount of power supplied out of the USB ports between PC MFRs. Maybe this is a good or bad example but I use a 2.5 Seagate 1TB HDD, it has one cable and works on any device I connect it to without any help. I also have a Toshiba 2.5 500GB HDD and it requires both cables to be used. I also have an older 250GB HDD (brand forgotten) and it has one cable BUT requires an AC-powered USB hub.
To make it more confusing, or less, did i just happen to say the magic word, that cables and power are related to whether the external drive is 5400 or 7200rpm? Is that the whole problem? Is the power out from the computers port only enough for 5400rpm or have I run the discussion off the rails because my brain is all whacked out?
It also varies by system. System manufacturers must provide "at least" as much current as specified in the USB spec. This is a minimum value, not a maximum value. You must meet the minimum to be USB compliant in your implementation. The maximum safe current for a typical USB port is ~2.0-2.1 amps. But the USB 2.0 spec only requires .5 amps and the USB 3.0 spec only requires .9 amps.
There are some system manufacturers that choose to go above the minimum current in their USB plumbing and implementations. This is why you can sometimes see a difference when you plug the same device in a different computer - and get different behaviors. In every case I have seen, all the computers meet USB spec. It is usually when the external device demands "above spec" current that you end up with varied results.
Sorry for my late answer and thank you all for your help
- for the cable i don't think a shielded cable will solve anything
if the problem was the cable, i would have the problem with my 3 others computer (2 laptops and a nuc DN2820FYKH)
my problem is only on my nuc5ppyh so i will not buy another cable to proceed this test
- I've tested your procedure to enable the write caching on my hard drive. the test once this setting is done is still the same. my hard drive is still unstable when i write data
have you another setting suggestion ?
@N. Scott Pearson
I’ve this problem with only 1 external hard drive plugged into a blue port so it should work without external power supply.
moreover, I've a ssd inside my nuc
my external hard drive are 2,5" 5400rpm so I still not understand the problem
Maybe you are right and it it’s only a system difference. However I've seen that a new firmware 53 is released so I’ll give it a try.
In conclusion my roadmap to maybe solve the problem is :
- to update drivers / firmware as soon as possible to check if my problem magically disappear or not. if it solve my problem, i will post a message.
- Maybe I’ll test later my nuc with a linux distribution to check if my problem is an isolated windows issue or not
- wait for another suggestion on this forum