Edge Computing Division Users Forum

Hi,
I have a 2TB SSD (using this:http://www.startech.com/nl/en/HDD/Adapt ... 5SAT22MSAT dual-msata bay with 2 1TB mSata SSDs) that I want to connect to my fitlet via an external enclosure.
However, upon insertion the fitlet immediatly reboots. After booting, the disk is detected and working. Initially I tried this via the eSatap port, but the same thing happens using usb 3.0 (or 2.0). The same disk does work on my laptop (both with eSatap and usb 3). And other 2.5" external drives do not exhibit this behavior when inserted into the fitlet. Leading me to conclude that it's due to a combination of the constructed 2TB SSD and the fitlet.

The most likely suspect to me seems that the fitlet can't draw enough current to power both mSata disks at the same time, while also otherwise fully powered.

I can think of two possible solutions:
1. Try to use a sata -> usb 3 + usb 2 cable. This would distribute the load over two usb ports. Not sure if it would help though. But this would be the easiest solution for me.

2. Power the SSD externally with 5V. This should work but I'm not sure if mixing power from different sources could become a problem.

Any ideas on the matter would be appreciated.

Can you please explain how do you connect this mSATA to 2.5” SATA adapter to the fitlet? Is there a SATA/USB3.0 bridge?
The adapter itself can be a reason for such behavior as it impose additional electric element in the network. Maybe it doesn't support hot swap protocol and reboots under such conditions.

Your assumption is logical as well, fitlet is a low power system, so that powering two SSD devices, when one of them is really large in size - 2TB can be an overload conditions. Current consumption can be checked in this 2TB storage datasheet.

The best way to verify that is as proposed, to try connect this storage via other adapters. There are adapters that features connection to 2 USB ports. In terms of power sourced from 2 different USB ports, should not be a problem.

Thanks for the reply.

The SATA->eSATAp I used was this thing:
http://www.startech.com/HDD/Mobile-Rack ... 510U33RUSM
Which should really support hot-swap. This one failed very consistently.
The SATA->USB3.0 were various external harddisk cases, which did work properly with regular 1 TB SSDs.

I've since also tried external cases which featured a cable with miniUSB -> 2 USB 2.0 connectors and one with a USB 2.0 and eSATA connector. Both worked rather inconsistent depending on which USB ports (USB 3.0 generally caused reboots more often) were used and in what order they were inserted. There were also some cases where attaching during boot up didn't cause a reboot while doing so when in windows did.

I ordered a SATA->USB3.0+USB2.0(for power), which should hopefully solve the problem and provide the desired speed. I'll post an update when I've tested it.

Unfortunately the sata->usb2 + usb3 solution didn't work, same issue.
An obvious workaround is to connect the usb2.0 to an external power source, but that's not really an option for me.

Luckily Samsung released 2TB SSDs this month, so I'm going to look at those.

I have the exact same problem. When I plug in https://www.google.de/shopping/product/ ... 2443568595 (VIA USB3 hub and RealTek 8153 USB3/Ethernet converter) into one of the USB3 ports, the fitlet-XA10-LAN resets immediately.

I also suspect that the voltage drops in that moment, but I have no scope to verify. It works when I plug it into the USB2 ports.

Not great, but I can deal with this for now... just wanted to share my experience.

Would a more robust power supply solve this problem? What about a powered usb hub? (Are they still a thing with usb 3.0?)

I've experienced a lesser problem on fitlet-i and MintBox Mini... With a usb flash drive connected, I then have to use a little trial and error to find working ports for the keyboard and mouse.

We acknowledge fitlet sensitivity to USB current draw. We will redesign the USB overcurrent protection in the next fitlet revision.
It does not appear to be a problem that can be resolved by a different fitlet PSU.