2
\$\begingroup\$

I have a USB-A 3.0 hub with ten ports. I've owned it for about 2 years now without any issues, until recently. About a couple of months ago I had a flash drive plugged into one of the ports on the hub and noticed it seemed to die much earlier than I had expected. Thankfully the data I lost I could get back, but it really concerned me when another flash drive in the same port died again a month after. These aren't cheap flash drives, either, they're really nice Kingston ones. They both visually appear to have the same controller, so my initial thought was that it's just a matter of cheap controllers. The flash drives were also plugged in and powered on 24/7, so I thought maybe that had something to do with it.

But then I decided on measuring the voltage output of the port to make sure it's not causing issues, and I discovered something interesting. The port where the 2 flash drives died is outputting about 5.48 volts. Curiously enough, the other two ports I checked output about 5.24 volts. I couldn't check all ten as the rest are occupied with other devices (all USB devices except my keyboard and mouse are storage devices, mostly HDDs but some flash drives).

My iMac outputs about 5.11 volts, a custom-built PC at about 5.08 volts and my Dell monitor at about 5.04 volts (it has USB ports, too).

My question is, how safe are all of these voltages, more particularly, the voltages from the hub? I use a lot of USB devices on my computers and using all ten ports on the hub would be very handy, but I don't want to risk damaging more drives if that's the case.

The most information I can find is "USB uses 5 volts," which doesn't help much here.


Update: I had a chance to disconnect all my drives from the hub. I reran the tests and found out that every port outputs a stable 5.42VDC (stable as in, the number didn't fluctuate for the 5 or so seconds it was connected). The wall adaptor should output 12VDC and outputs at the most 12.12VDC. No devices were connected at the same time during this measurement (update: with 7/10 I get the same results).

With that being said, is it still possible that this hub is usable? Could I potentially use a lower voltage adaptor to bring the voltage down to more acceptable levels? I ask because this hub I am using is really nice, and I'd hate to buy another one (I bought an Amazon basics one before and had problems with that). Of course, if it means risking my devices lifespans, I'll buy another hub.

I would like to note that the other devices on this hub, mostly WD My Passports (hub powered), one WD My Book (wall powered), a SATA Hitachi drive in a 3.5" Rosewill enclosure (wall powered), and a Crucial SSD in a Sabrent USB enclosure (hub powered) are all okay and have been plugged in longer than the flash drives that died were.

\$\endgroup\$
10
  • 1
    \$\begingroup\$ electronics.stackexchange.com/questions/12935/… \$\endgroup\$
    – mhaselup
    Nov 11 '20 at 0:00
  • 1
    \$\begingroup\$ Thats outside spec. Many devices will tolerate it well, but it's not guaranteed. // The manner that you are using the drivers may be an issue. If you are doing large numbers of read and or word you may be exceeding their lifetimes . How are you using then? \$\endgroup\$
    – Russell McMahon
    Nov 11 '20 at 0:27
  • \$\begingroup\$ @RussellMcMahon A bit difficult to understand your question. I do a lot of reading and writing to the drives connected to the hub. I measured the hub with a multimeter and an old USB cable. I cut the end off of it and went to the red and black wires with the multimeter to get the voltage readings. The flash drives were maybe 7-8 yrs. old, both were 32GB and both had about, give or take, 750GB written over time. \$\endgroup\$
    – leetbacoon
    Nov 11 '20 at 0:37
  • \$\begingroup\$ Is the hub externally powered by a power supply? If so, is the power supply itself supplying the correct voltage? Are the two ports with higher voltage by any chance specially marked high output current ports? \$\endgroup\$
    – Justme
    Nov 11 '20 at 5:17
  • 1
    \$\begingroup\$ @leetbacoon I just did a web search on wear levelling on SSDs . Some mention only write wear-out and some specifically say read / write. \$\endgroup\$
    – Russell McMahon
    Nov 13 '20 at 9:02
2
\$\begingroup\$

The USB spec is a maximum of 5.25V. I think that port may be failing and others may follow so I'd recommend replacing the hub at this point rather than risking destroying another device.

\$\endgroup\$
2
  • \$\begingroup\$ You need to follow USB-IF updates to USB 2.0 specifications more closely. Since 2014 the VBUS is allowed to be 5.5V. \$\endgroup\$ Nov 14 '20 at 7:59
  • \$\begingroup\$ @Ale..chenski thanks for the heads up. This increased voltage tolerance won't apply to the OP's devices which pre-date 2014 being 7 or 8 years old. \$\endgroup\$
    – mhaselup
    Nov 14 '20 at 8:37
2
\$\begingroup\$

According to USB-IF updates, "USB 2.0 VBUS Max Limit ECN as of August 11, 2014", the allowed VBUS voltage is now 5.5V. See latest release of USB 2.0 specification with all ECNs. Old poorly-designed devices may have difficulty to handle the 5.5V input.

The ECN mandates the following change to Tab.7-2 of USB 2.0 Specifications, plus corresponiding changes in several related places: enter image description here

\$\endgroup\$
2
  • \$\begingroup\$ Hmmm, quite an interesting tidbit. Thank you for sharing! I assume this also applies to USB 3.0 (and its numerous iterations). So, even though it is a bit higher than preferred, it's still (theoretically) safe. The oldest device connected to my hub is about 10 years old, a USB 2.0 Seagate drive, and it's been working fine this whole time. :-) \$\endgroup\$
    – leetbacoon
    Nov 14 '20 at 8:37
  • \$\begingroup\$ @leetbacoon, this is not a "tidbit", it is a normal legal process of amending aging specification with progress in technology. There are many ECNs, if you have time to download the spec and corresponding ECNs. And yes, the USB 3.x specs does include the change to 5.5V as well. These ECNs even have explicit warning: "New designs adhering to this ECN will need to consider the new 5.5V VBUS max level during the component selection process." Many old designs did use internal linear regulators to 3.3V with max input voltage of just 6V. \$\endgroup\$ Nov 14 '20 at 9:13
1
\$\begingroup\$

Usb spec is 4.4 to 5.25. Or for 3.0, 4.55V to 5.25V. Some devices or chargers put out more to compensate for voltage droop for high amperage charging devices. I've seen some as high has 5.7V. These tend to assume a few feet of cable and 2 Amp draws.

5.5V is not unusual but outside of spec. At 10% higher most devices should handle it fine. Most usb drives work at 3.3V internally as does usb data signals. And typical ICs meant for 5V would have 6V as the absolute max.

While I can't argue against your results with multiple dead flash drives I think them being on 24/7 had more to do with them. Even really nice Kingston are mass produced consumer products and I wouldn't expect them to be much higher quality than a random SanDisk or whatever. Drive failures are a bell curve so you could just have confirmation bias.. Without a scope and logging the signal over time to check for dirty power or surges, or stress testing a bunch of drives, replacing the hub is the easiest solution.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for the detailed answer, much appreciated. \$\endgroup\$
    – leetbacoon
    Nov 11 '20 at 1:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.