Thanks for you reply! I wasn't aware that changes in the webinterface overwrite some settings again so now I am doing all changes via the uci command line. I tried your values but it didn't change anything. I set the htmode to VHT80, channel to 37, country both US and DE.

wireless.radio0=wifi-device
wireless.radio0.type=mac80211
wireless.radio0.path=pci0000:01/0000:01:00.0
wireless.radio0.txpower=17
wireless.radio0.htmode=VHT80
wireless.radio0.channel=37
wireless.radio0.country=DE
wireless.radio0.log_level=1

I also set the log level to 1 but I still can't find anything in the logs that seems relevant. I enable the interface in the web gui but the state stays at "Wireless is disabled or not associated". What options do I have to find out what is going on? Is there a guide somewhere on debugging these issues? Like how to see if the device driver can even properly communicate with the hardware? Or is that a given if the 2.4 Ghz interface works?

Thanks for your help!

There's likely a problem with your config as the interface is showing down in Luci, I had this several times in the past, although I would expect something in the logs somewhere (either dmesg or via logread).

I'm not convinced that 37 is a valid channel, as it's not on wikipedia's list of wlan channels and doesn't show when I run iw phy phy0 info on my device, so I'd suggest trying one of the channels that report as available in the output of that command, or in Luci.

It's also worth noting that the TP-Link devices are a bit odd in that the location in the flash (?) is set to US, which unfortunately for us means, even if we set the device to a European country, we can't get anything that isn't permitted outside of the US.
I'm not aware of any way to change the hard-coded location, and the workarounds for the ATH9K don't work with the ATH10K. I did however manage to work around this by updating regdb.txt so that the US settings matched the GB ones and compiling from source - I'm just not sure it's worth the trouble unless you specifically want something that isn't allowed by default

Ah, cool, that made a difference. I can now use the the 5Ghz wifi with channel 40 and 48. I will research if I can somehow manage to enable the higher channels, but for it's fine the way it is. Kudos!

How much is WLAN-LAN throughput on 7500v3 on 5ghz (ac) you have?

My measurements reveals:
20mbps on openwrt,
and up to 120-140mbps on tp-link's firmware.

It seems too slow, any ideas?

(Last edited by andrei.alekseenko on 15 Oct 2014, 17:20)

please, can you provide details about that.

so, i have no any understanding of electronics. i asked a friend and he soldered for me pin-header to J1 on WDR7500V3 main-board. i connected to it cheap "USB To RS232 TTL" adapter i bought on eBay, i can see all the output from WDR7500V3 on Tera Term, but when i type on the keyboard, it doesn't react. it's the exact same as when i disconnect the RX wire between the "USB To RS232 TTL" adapter and RX pin from J1 on WDR7500V3 main-board. basically, RX communication between the "USB To RS232 TTL" adapter and RX pin from J1 on WDR7500V3 main-board is not working.

i would really appreciate if someone can help people like me, that doesn't understand electronics, but who want to be able to update the firmware and give more simple details.

BTW, i see others, have or not have same problem, i want to ask them for details too, e.g. what adapter they use, etc:

many many thanks in advance!!

[EDIT] interesting, when i disconnect the 3.3v wire, then when i type on the keyboard the characters appears on screen, but each character is printed several times and also some "ghost" characters i didn't type appeared. that still prevents to use it normally. so, it seems there is some RX communication, but something is wrong/missing to be normal.

(Last edited by matura713 on 23 Oct 2014, 18:53)

3.3 wire???? Why are you connecting it???

as i told, i'm not technically skilled, there is no any detailed information here, which wires it's needed to connect and which not.

so, all 4 wires GND, 3.3V, TX and RX connected, i see all console output, but i cannot type, nothing appears on screen when i type. it's the same if i connect just 3 wires: GND, 3.3V and TX.

if i connect GND, TX and RX, then i can type, but it's still not normal - for example press "Enter" results in output of "y" on the screen, if i type "m" it's outputted as "mI".

i believe what user "silviop" wrote is the key - "a resistor between vcc and rx", because exactly disconnecting Vcc, i.e. 3.3V wire makes RX work, but "work" with the above bad behavior, which still doesn't allow really normal work.

however, i wonder how other users made it working just with soldering a pin header to J1, which is already done on my WDR7500 main-board.

(Last edited by matura713 on 23 Oct 2014, 19:11)

ok, problem solved and i'm making it separate post for all other technically unskilled folks like me, that will be lost like i was.

so, the main reason for my problem turned out to be picture "88_272402_cf30961c13a0731.jpg", posted here:

https://forum.openwrt.org/viewtopic.php … 36#p233736

and that is floating around Internet, which claims on WDR7500V3 the pinouts on the J1 are different than WDR7500V2 - that's totally not true!! initially i connected the pins according to that picture and ended up with all those troubles i described in the above posts.

so, the pinouts of J1 on WDR7500V3 is exactly the same as on WDR7500V2:

http://wiki.openwrt.org/_detail/media/w … tl-wdr7500

even the board look differently the easy way to identify them is "C40" marking on the board. the pin closest to "C40" is Vcc or 3.3V and from there the order is: Vcc, GND, RX, TX. i needed to connect only RX, TX and GND to get proper and normal work.

that's all about the hardware setup.

i already got 'u-boot' access - to do that, when on the screen appears "Autobooting in 1 seconds", just type "tpl" - at least that what was working for me.

i most likely will update here with easy steps for the firmware update via 'u-boot', but i will hold on for a while to get more familiar with using 'u-boot' and available commands, i don't want to break anything. however, the most frustrating i think was the misleading J1 pinouts information.

(Last edited by matura713 on 23 Oct 2014, 20:53)

it seems anything that can go wrong will go wrong with my WDR7500V3 attempts.

so, now i want to make backup of my flash partitions (i.e. "mtd" devices) before i write anything permanently to the flash and first i tried to use:

http://nosferatu.g0x.ru/pub/wdr7500/fir … _ramimage/

i can boot from ram with it, but no "mtd" devices are present and available when the kernel boots.

then i tried "openwrt-ar71xx-generic-archer-c7-v3-initramfs-uImage.bin" from here:

http://pan.baidu.com/s/1o6mH5qi

it was posted several times in this forum thread:

https://forum.openwrt.org/viewtopic.php … 42#p238442
https://forum.openwrt.org/viewtopic.php … 12#p248012

and it's supposed to work, but all i get is:

"Uncompressing Kernel Image ... ERROR: LzmaDecode.c, 543

Decoding error = 1
LZMA ERROR 1 - must RESET board to recover"

so, please, anyone, how i can boot from ram with WDR7500V3 and backup my "mtd" flash partitions? it seems no of the files posted for that same task are really working.

Do you have root access to the tp-link firmware via serial?

nope, that's the whole point, no one knows the root password of new tp-link firmwares, i read last year there were unsuccessful attempts to reverse-engineer the new hash in "/etc/shadow".

so, now (i mean with newer tp-link models), the steps for backing up the original flash content, to my understanding are:

1. after get serial access, boot openwrt image in ram, that way get root access and then copy the original firmware, what the readme file here says:

http://nosferatu.g0x.ru/pub/wdr7500/fir … readme.txt

but surprisingly there are no "mtd" devices with that image loaded in ram, contrary to what the readme states

2. after copy the "mtd" partitions with the original firmware, there are actually 2 options for a next step, as you already have backup:

2.1. open original files on computer, edit "/etc/shadow" to remove the hash for the root password, i.e. made that way blank password work and using same as step 1., write it back, that will give original firmware without root password

2.2. write openwrt to the flash, as you already have backup of original flash content to which you can revert any time

so, maybe, i should change my questions - since there is already kind of official openwrt support for WDR7500V3 how i can build "ram-image", "vmlinux", "initramfs" or what ever it's called, like "openwrt-ar71xx-generic-archer-c7-v3-initramfs-uImage.bin", but one that works, i.e. image that can be loaded from ram (not need to write it to the flash) and when it's loaded "mtd" devices are present? thanks.

very new to openwrt. i did get it working on my archer c7 v2 and so far everything seems good except i cant get the 450 link speeds back working on 2.4 is there a way to do this and if so how do i go about that. again new to this so if this is a dumb question i apologize.

thank you for any help!!

(Last edited by TheDrizzle on 31 Oct 2014, 17:21)

English firmware was just made available:
http://www.tp-link.us/support/download/ … version=V3

Is the v3 entering western markets?  What's different about it?  Did v2 also hit eastern markets first?

i've become half-developer, but finally built my own ram image from openwrt source code that works for me. so, case closed.

not sure if there is a hardware difference but the firmware is the same version listed for v2 also inside my v2 firmware it was listed as v2/v3 after the version number for the firmware. but i am also highly curious to know if they changed anything hardware wise. hopefully they didnt make v3 that much better then v2 now... doh. tplink it starting to really hit a nerve with me. instead of releasing new firmwares they just release a new version and leave everyone with the old ones out in the cold it seems.

though barrier breaker on the v2 is working AWESOME now. soooo much better then stock even dlna works properly now

(Last edited by TheDrizzle on 2 Nov 2014, 19:22)

The v1 not remaining compatible was really QCA's fault for changing the hardware so much on their AC chips, which is not a huge surprised when things are released to comply with the draft spec. TP-Link has not done this elsewhere with their other routers, so I don't see this as a pattern from them.

Have you seen this?
https://forum.openwrt.org/viewtopic.php?id=53703

wow that is interesting i will have to test this also. thing is so far for me the openwrt 14.07 is awesome on this router. i have not done all the speed tests but so many things are actually working now. dlna was HORRIBLE on this router before and works fantastic now. i have not seen or noticed any slowdowns yet though. in fact the transfers between systems (transferring a video) have done so MUCH better than before with the factory firmware. i used to drop connection while streaming from a laptop that does not happen now. everything just works so good except as i said i have not tested the actual speeds yet. im wondering if he enabled the ac on the 5g band as noted (im sure he did but thought it was worth asking), im very new to openwrt myself but imo even if the speeds are a slower it just works now for me. before i was ready to literally toss this thing in the trash and i can actually use it now.

being new here i do not want to get in any trouble for asking but just being curious if anyone that has tried ddwrt has seen this same result with the speeds in that thread posted? again just being curious being a former ddwrt user (just not on this router model i went straight to openwrt from factory)

as far as revision i understand between v1 and v2 but now they went and made a v3 and as per email conversations direct from them in china they seem to feel there will not be another update for the v2 (this was stated in a email before the v3 was listed on their site from a beta tech (being a beta tester) in china. i just would love to know what v3 changed from v2. i sent a email but will have to wait for the response if i find out before someone else does i will post here.

(Last edited by TheDrizzle on 3 Nov 2014, 01:40)

Your observation about it showing "v2/v3" makes sense given that the links for the latest v2 firmware and the only v3 firmware are one and the same. So while it would seem that there's definitely new hardware out there (why else a v3?), it's very unlikely that it's a fundamental difference like with v1 compared to v2.

That also bodes well for continued v2 updates, since it's tied to v3's updates. Kind of counter to what you heard, which sounded ominous, according to their support, an update to v2 is supposed to happen any time now to correct a bug surrounding (remote) web management introduced in the build from a month ago.

hmm interesting. as i said im going to try to get more info from them if i can.

any idea if a new build of openwrt will be out anytime soon and will it correct the slow wifi you mentioned? i would think if it really was only half speed as shown that would be a major issue and hopefully they are working on that as i said the stock firmware sucked.

I am considering buying an Archer C5 or C7 v2. I see that the C5 is supported in openwrt but there appear some open issues with the C7. Are these resolved or will they be resolved in the near future or do you recommend buying the C5?

I don't see any significant differences between the c5 and the c7 except for the wireless speed. Are there any I should be aware of.

does anyone have a simple revert file to go to stock firmware for some testing (archer c7 v2) or does it have to be done the hard way? a revert file would make it so easy to swap back and forth to test some things.

thanks

Basically the same hardware, so almost anything that effects the c7 will also effect the c5.

edit: got it figured out. finally got it to take the factory firmware and back to open wrt.

(Last edited by TheDrizzle on 4 Nov 2014, 08:36)

so you're basically saying that the 5GHz speed issue probably also affects the C5 only nobody has complained about it yet?