Not according to this thread
... in which some report three months of uptime with vectoring? I agree that it might be wonky to set up but the capability is there.
Expect that he's not even using the device?
Anyway, TS clearly stated no hardware modding...
That is not exactly true, there are a number of different vectoring capable firmwares that work reasonably well on the BT HH5a (e.g.: https://forum.openwrt.org/t/bt-home-hub-5a-bthomehubv5a-vdsl-vectoring-connection-problems/37856/8), although I happily admit that quite a number of modem firmwares are pretty error prone. But is any ISP in the UK actually using vectoring, I believe not?
I do agree with the rest of your points though, the BT HH5a is a nice device for what it does, but it is easy to run out of CPU cycles... (I happen to use one as bridged VDSL2-vectoring modem, when I tried to also operate it as my main router I was hitting CPU limits even with a 50/10 Mbps link, since the OP already has a modem, this will only be an alternative if he can find a hh5a for free or maybe even already flashed with openwrt).
That is not a goo argument IMHO, unlike MIPS arm has not yet proven itself as a decent network CPU, or rather those CPUs used in SOCs are not necessarily optimal for a router (or not optimally configured, see R7800 issues with frequency scaling for example).
USB3 also often interfere with 2.4GHz wifi, so IMHO not a clear cut must-have for a router; but that is subjective opinion...
But keep in mind that centralizing such services on your main access server has some security implications (if the router gets hacked you are fully exposed, unlike if the NAS operates on its own hardware).
1 Like
I use that linked firmware blob on a BT HH5a and it works pretty well and stable.
Ebay comes to the rescue:
already modded HH5a for 22 pounds.
But I tend to agree with a modem that allows bridge mode in his posession, the EA6350 looks like the better bargain.
IPQ8 isn't the topic at all? 
Anyhow, in general Marvell is very popular, reliable and performs a lot better than MIPS counterparts and most ARM solutions (within the same price range).
Do you have any real life evidence of that (I do know that Intel published a document outlining potential issues)? Most of it is resolved by shielding and you can disable USB if you're overly concerned about it.
Not personally, but see:
https://forum.openwrt.org/t/xiaomi-wifi-router-3g-18-06-x-wifi-issues-2-4ghz-5ghz/20169/34?u=moeller0
OP here, I just marked the problem as solved. Use the right firmware and it flies.
Thank you all who contributed to this thread. I went for the Linksys EA5350, and I'm pretty happy with it so far.
The one thing I was unsure about was the v3 vs v1/v2 issue, with OpenWRT only supporting the former. Online stores don't tell which version they are selling, and neither does the box. Doing online research, I found this page https://www.linksys.com/gb/support-article?articleNum=141033 It says that only v3 has "Dual-Band Guest Network" (question 8 under "Hardware Information"). The online retailer I used listed "guest access" as a feature, so I thought this might be the clue that revealed this to be a v3 device.
My question now is: did I interpret this correctly? Could it be that "dual-band guest network" is not the same as simply "guest network", and I could have ended up with a v1/v2 device?
I'm curious about this, as it may help other people in the future to clear this doubt if indeed I was right.
I wouldn't rely on this as a distinguishing fact, while it might be that Linksys changed the phrasing between those revisions accordingly - it's still a marketing term and not based on technical differences (and at least sellers are likely to use both marketing blurbs interchangeably, without looking at the h/w revision). On a dual-band device, one would usually expect a potentially existing guest network to cover all available bands and not to be restricted to a single one (the later might exist in the older or low-end sector).
This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.