USB3 won't power 2,5" external HDD

[Tiedemate]

Hi,

I have two 2,5" external hard drives, which I use for backups. So far I have used them with different machines mostly on USB2 ports, never had an issue. So I thought surely the PBP USB3-port must be able to power them, but that does not seem to work. One won't spin up and sits there clicking, the other one does spin up but starts clicking once I try to mount it. So now I bought a USB-C to USB-A converter and hooked the drives up to the USB-C port, and now the one that wouldn't spin up mounts (the other one does still does not mount). I couldn't find any power specification for the USB-ports in the Rockchip datasheet, but I am a bit puzzled because so far even the cheapest netbooks would power these drives even on USB2...
Is this the expected behaviour?

Thanks!
Björn

[xalius]

The ports have 500mA (USB2) and 900mA (USB3) current limiters respectively, what current does the HDD need?

[Tiedemate]

The ports have 500mA (USB2) and 900mA (USB3) current limiters respectively, what current does the HDD need?

The HDDs have a specified current during power up of 1,2A, so that's probably the issue. One drive will succesfully spin up, and after that the current should decrease, but anyway I'm out of spec. 
What's the spec for the USB-C port?
Would it be possible to tie the USB-C and USB-3 together with something like this? https://www.amazon.de/Delock-Power-Adapt...B001BHVMIU
I know it's sketchy, but maybe the hardware setup of the PBP allows this, even if it is not foreseen in the general USB spec?

[hmuller]

I'm using a Seagate Backup Plus 2TB on the PBP's USB 3.0 port. Haven't experienced any issues with it, so can recommend it for use.

[Arwen]

Their are USB cables that break out the data and power separately. Thus, you can feed the power from a USB power brick that can supply 2amps. Then plug in the USB data cable into the Pinebook Pro. Should work.

That's sort of what I have to do with one of my older 2.5" disk enclosures. It came with a USB Y cable for extra power, and only one supplies the data.

Another 2.5" disk enclosure came with an optional barrel connector to USB, to run to a power brick. (That enclosure has eSATA too, which does not supply power.)

[neilman]

I have just tested a 5TB Seagate Backup Plus drive with my PineBook Pro

Using the stock, updated, out-of-box software...

USB 3 socket : Works perfectly, plays videos with VLC fine, ejects safely.
USB 2 socket : Powers and clicks a lot but does not appear to function properly - no device shows up.
USB C socket : Appeared to power up initially but after a few seconds it just powered off and died.

Retrying in the USB 3 socket worked fine again and was able to eject safely.

Running off a power bank with a power meter connected shows short high starting current of 670mA dropping to a quiescent idling current of 240mA.
(Since my power meter only refreshes its display a few times a second the peak current might be much higher).

[Tiedemate]

I'm using a Seagate Backup Plus 2TB on the PBP's USB 3.0 port. Haven't experienced any issues with it, so can recommend it for use.

In my case it's a seagate backup plus portable 5TB, seems to draw more amps...

---

I have just tested a 5TB Seagate Backup Plus drive with my PineBook Pro

Using the stock, updated, out-of-box software...

USB 3 socket : Works perfectly, plays videos with VLC fine, ejects safely.
USB 2 socket : Powers and clicks a lot but does not appear to function properly - no device shows up.
USB C socket : Appeared to power up initially but after a few seconds it just powered off and died.

Retrying in the USB 3 socket worked fine again and was able to eject safely.

Running off a power bank with a power meter connected shows short high starting current of 670mA dropping to a quiescent idling current of 240mA.
(Since my power meter only refreshes its display a few times a second the peak current might be much higher).

I have a Seagate Backup Plus Portable 5TB and a silicon power armor with the same HDD. None work in USB3. The silicon power works fine in USB-C. The seagate will spin up but won't mount (won't even be shown in the device notifier (I'm running Kubuntu 19.10)). Both work fine on other systems. I guess I'll order one of those USB measurement thingies, to see what's really going on the with the voltage and currrent.

[stealthbanana]

I was actually going to post about this last night, I have found that external NVME drives will not function in the USB3 port, however, by using a USBC to USBA adapter, or a USBC hub they work perfectly via the USBC port.

When using the USB3 port, the system reports it is in a constant state of spinning up.

Code:

[  745.683789] rockchip-dwc3 usb0: USB unconnected
[  759.483321] usb 4-1: new SuperSpeed USB device number 2 using xhci-hcd
[  759.501126] usb 4-1: New USB device found, idVendor=152d, idProduct=0562
[  759.501144] usb 4-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  759.501152] usb 4-1: Product: External
[  759.501159] usb 4-1: Manufacturer: JMicron
[  759.501166] usb 4-1: SerialNumber: DD56419883893
[  759.506278] scsi host1: uas
[  759.509304] scsi 1:0:0:0: Direct-Access     JMicron  Tech             0204 PQ: 0 ANSI: 6
[  769.510904] sd 1:0:0:0: [sda] Spinning up disk...

attempts to spin up for about 30 seconds.

Code:

[  772.283257] .
[  803.215101] sd 1:0:0:0: tag#0 uas_eh_abort_handler 0 uas-tag 1 inflight: CMD
[  803.215125] sd 1:0:0:0: tag#0 CDB: opcode=0x0 00 00 00 00 00 00
[  803.215532] scsi host1: uas_eh_bus_reset_handler start
[  803.215641] xhci-hcd xhci-hcd.3.auto: ERROR Transfer event for disabled endpoint slot 1 ep 4 or incorrect stream ring
[  803.215655] xhci-hcd xhci-hcd.3.auto: @00000000e0ce2340 00000000 00000000 1b000000 01058001
[  803.327768] usb 4-1: reset SuperSpeed USB device number 2 using xhci-hcd
[  803.347666] scsi host1: uas_eh_bus_reset_handler success
[  813.347189] sd 1:0:0:0: tag#0 uas_eh_abort_handler 0 uas-tag 1 inflight: CMD
[  813.347213] sd 1:0:0:0: tag#0 CDB: opcode=0x0 00 00 00 00 00 00
[  813.347226] scsi host1: uas_eh_bus_reset_handler start
[  813.347330] xhci-hcd xhci-hcd.3.auto: ERROR Transfer event for disabled endpoint slot 1 ep 4 or incorrect stream ring
[  813.347345] xhci-hcd xhci-hcd.3.auto: @00000000e0ce2470 00000000 00000000 1b000000 01058001
[  813.459482] usb 4-1: reset SuperSpeed USB device number 2 using xhci-hcd
[  813.477769] scsi host1: uas_eh_bus_reset_handler success
[  813.477797] sd 1:0:0:0: Device offlined - not ready after error recovery
[  813.478061] ready
[  813.478124] sd 1:0:0:0: rejecting I/O to offline device
[  813.478175] sd 1:0:0:0: rejecting I/O to offline device
[  813.478208] sd 1:0:0:0: rejecting I/O to offline device
[  813.478238] sd 1:0:0:0: [sda] Read Capacity(16) failed: Result: hostbyte=0x01 driverbyte=0x00
[  813.478257] sd 1:0:0:0: [sda] Sense not available.
[  813.478280] sd 1:0:0:0: rejecting I/O to offline device
[  813.478308] sd 1:0:0:0: rejecting I/O to offline device
[  813.478336] sd 1:0:0:0: rejecting I/O to offline device
[  813.478361] sd 1:0:0:0: [sda] Read Capacity(10) failed: Result: hostbyte=0x01 driverbyte=0x00
[  813.478377] sd 1:0:0:0: [sda] Sense not available.
[  813.478399] sd 1:0:0:0: rejecting I/O to offline device
[  813.478428] sd 1:0:0:0: rejecting I/O to offline device
[  813.478459] sd 1:0:0:0: rejecting I/O to offline device
[  813.478484] sd 1:0:0:0: [sda] Write Protect is off
[  813.478502] sd 1:0:0:0: [sda] Mode Sense: 00 00 00 00
[  813.478523] sd 1:0:0:0: rejecting I/O to offline device
[  813.478547] sd 1:0:0:0: [sda] Asking for cache data failed
[  813.478564] sd 1:0:0:0: [sda] Assuming drive cache: write through
[  813.478586] sd 1:0:0:0: rejecting I/O to offline device
[  813.478612] sd 1:0:0:0: rejecting I/O to offline device
[  813.478636] sd 1:0:0:0: rejecting I/O to offline device

Leading me to think it is a power issue with high power demand devices, works perfectly via the USBC port.

[Tiedemate]

Hey guys,

after some wrong items delivered I finally received my USB tester, so I ran a few measurements. For a comparison I measured the USB2 und USB3 port of a cheap netbook with Intel chipset (Trekstor Surfbook E11B) and a all in one PC with AMD chipset (Asus EeeTop E2400A). I used this setup (Pinebook -> Adapter USB-C to USB-A -> USB tester -> adjustable load):


I don't know the accuracy of the USB tester so the absolute values should be taken with a grain of salt, but at least the comparison between the three computers should give an idea about the situation.
Here are the results:

(Small brain fart: the surfbook is actually made by Trekstor, not Medion...)

Observations:
The Surfbook and EeeTop USB2 ports allow currents of 1,5A, I did not try to go higher. 
The Surfbook USB3 port shuts off* as soon as the current exceeds 1,47A
The Pinebook Pro USB2 port shuts off* when exceeding 0,63A
The Pinebook Pro USB3 port voltage drops sharply when going from 0.99A to 1.0A. Spot on ;-)
The Pinebook Pro USB-C port voltage drops sharply when goint from 1.17A to 1.18A.
The Pinebook Pro USB-C port voltage is overall lower than on all other ports measured, however this might be due to a resistance caused by the USB-C -> USB-A adapter.

*shut off = voltage drops that much that the USB tester does not work anymore

Since even overloading the Pinebook Pro USB ports did not create smoke I became brave and hooked the USB2 und USB3 ports together with a Y-cable and started measuring. All went well up to about 1.16A, but then the Pinebook started making strange noises so I stopped that experiment.
I did not try to measure USB3 and USB-C or USB2 and USB-C hooked together.
However I connected the external hard drives with the Y cable to USB-C and USB3, and both drives spin up, get mounted and work fine without the Pinebook making funny noises. So I can use this combination with my drives. I did not try to power my drives with USB2 and USB3 or USB2 and USB-C hooked together.


It is a bit difficult to measure hard drive spin up current because the USB tester measurement value display updates slowly und toggles between current and voltage measurement on its own, but as far as I can tell the current during spin up is 0,55A to 0,6A, and when mounted about 0,75A. Yes, when the drives idle after having spun up and being mounted, the current is higher than during spin up, or that's at least what the USB tester displays. It fits to the behaviour when connected to the Pinebook Pro USB-C or USB3 without the Y-cable; they do spin up but won't mount, they just click.
Seagate specifies 1.2A power up current for the hard drive type, which is in fact too much for the Pinebook Pro USB ports according to my measurements.
My guess: The initial movement of the HDD heads draws too much current so the voltage drops too much and they immediately park again. Then the hard drive tries again, thus the constant clicking (once a second or so). The current peak created will not be visible on the USB tester because it updates too slowly.

Final conclusions:
My other hardware seems to be a bit more generous compared to the Pinebook Pro when it comes to current supplied by the USB ports, so that even USB2 ports power my external hard drives just fine. 
If I combine the USB3 and USB-C port of my Pinebook Pro with a Y cable, I can power my external hard drives. However this is out the USB specification (?) and therefore risky. Seems to work fine in my case however.

Best regards!
Björn

PS: Here are the pictures and the file with the raw measurement values if anyone is interested:
Dropbox