adding a powered usb hub did not fix it. so i modified a usb cable to let me splice in power from an external 5v psu.
*that* fixed it.
@gsuberland @ariadne ah yes, the single-worst unit in electronics.
2 ounces (weight) of copper (volume) per square foot (area) - it's a linear dimension, approximately 70um
My experience is that ASRock sucks, generally speaking, and if they put out any actually good mobo it is by accident.
LOL.
depends. if you're selling directly to the engineering folks, it's perfect. if you're trying to get to a company big enough to have a purchasing department, you might need something a bit more... polished? :)
Get yourself an USB Opto-isolator. The latest ones out of china that I've seen ship with a DC-DC isolator making it unnecessary to splice in an external 5v and it is still isolated up to 10kV or something like that.
I use it for my CNC as the motor causes the USB to drop so low that the host circuit kept detecting it as a reconnect signal. (Also I had to find a 2.0 only host controller as all of the 3.0's when downgraded were not "lenient enough" for this thing...)
the DC-DC isolator is basically a dc-ac converter + transformer + ac-dc converter but all in one small component. Hence it should be good enough to smooth out the unclean power.
And this one here is the cheap one I have. It is good up to 2.5kV (according to the seller): https://de.aliexpress.com/item/1005007254879252.html
@ariadne Maybe
It's a pretty simple device, and cables do seem to wear out faster than you'd expect
USB power sucks ass for ground loop isolation/noise, genuinely.
It took me weeks to get buzzing out of my computer audio stack.
@dadregga though you would think ASRock Rack boards would have clean power distribution... haha nevermind.
I probably wouldn't buy another ASRock Rack board after some of the weird shit my threadripper machine has done.