Is this a hardware or software issue?

[HellCatCharger]

Hey Everyone,
I wont get into specifics to save time but some long time forum members or people who have attempted USB tethering on various Asus routers know what im about to mention. I have an RT n 66u running the latest Merlin and an AC87U latest stock. Some others across other forums are using like an RT n 16 and a couple of other models and having the same problem. What happens with USB tethering to a phone or modem is webpages hang and some dont even load making it unusable. Changing the MTU and such doesn't help. I have tried tomato and DD WRT with the same results on both so i reverted back to Merlin and stock.

The other day i tried something and it fixed the issue but its not ideal. I removed the ehci_HCD driver thus making the router run on usb 1.0 and my USB 2.0 lte modem worked flawlessly tethered with non of the problems. Obviously the speeds take a big hit but it was an eye opener. There is sometimes I/O failures on both routers also. Is this a USB controller issue among many different models or a software one?

 

[sfx2000]

Sometimes it's drivers (software) or settings (MTU size as many 3G/4G/LTE are ppp based so need to reduce MTU size a bit)

 

[HellCatCharger]

I had taken the MTU down to even like 1200 and it was really bad still. Normally lte runs at about 1428

 

[sfx2000]

Also check/verify if possible the APN values - as that applies to the subscriber profile at the carrier side - and that APN might not be correct...

 

[HellCatCharger]

I checked those and all is good. Also, not really necessary because my Lte modem contains all of that information but I did enter it to be on the safe side. Im leaning towards hardware because if others are experiencing the same thing all with different model modems and myself with 2 different model routers and the same problem occurs my inclination is to believe the USB hardware is faulty from whoever they sourced it from. Very strange it works perfect when the USB 2.0 driver is removed and also because its the same result across all avaliable alternate firmware's including stock. Unless that driver is some how kept in tact across all those platforms it would have to be hardware related wouldn't it?