I suppose you have to rename mt7996_eeprom_233.bin to mt7996_eeprom.bin to have the driver pick that up instead. In the end this should be done using dt-overlay again with distinct dtbo for each wifi module type (BE14 vs. BE19) needing user-supplied EEPROM data.
Are you sure _233 is for this wifi module/frontend and i only need to change eeprom file? There are more _233 files and itās not clear that these are for this wifi board.
Iām just guessing, because those _233 files arenāt used anywhere by the driver and the BE14 module got 2T2R in 2.4 GHz, 3T3R in 5 GHz and 3T3R on 6 GHz, ie. 233 makes sense to be somehow.
thx, tried all 233 files for mt7996, but issue is similar, i see only the first 2 wifi devices (i expected 3 or is this wrong as the new board has only 2 frontends, so 2G4+5G should work over 1,but looking in iw list it seems 5G is missing), but i get no timeout messages
root@bpi-r4-v11:~# dmesg | grep mt7996
[ 14.674304] mt7996e_hif 0001:01:00.0: vgaarb: pci_notify
[ 14.679658] mt7996e_hif 0001:01:00.0: assign IRQ: got 119
[ 14.685090] mt7996e_hif 0001:01:00.0: enabling device (0000 -> 0002)
[ 14.691490] mt7996e_hif 0001:01:00.0: enabling bus mastering
[ 14.697195] mt7996e_hif 0001:01:00.0: vgaarb: pci_notify
[ 14.702687] mt7996e 0000:01:00.0: vgaarb: pci_notify
[ 14.707691] mt7996e 0000:01:00.0: assign IRQ: got 116
[ 14.714133] mt7996e 0000:01:00.0: enabling device (0000 -> 0002)
[ 14.721544] mt7996e 0000:01:00.0: enabling bus mastering
[ 14.866833] mt7996e 0000:01:00.0: HW/SW Version: 0x8a108a10, Build Time: 202a
[ 15.071021] mt7996e 0000:01:00.0: WM Firmware Version: ____000000, Build Tim5
[ 15.118893] mt7996e 0000:01:00.0: DSP Firmware Version: ____000000, Build Ti7
[ 15.160500] mt7996e 0000:01:00.0: WA Firmware Version: ____000000, Build Tim2
[ 15.543104] mt7996e 0000:01:00.0: eeprom load fail, use default bin
[ 15.550618] mt7996e 0000:01:00.0: registering led 'mt76-phy0'
[ 15.559589] mt7996e 0000:01:00.0: registering led 'mt76-phy1'
[ 15.567459] mt7996e 0000:01:00.0: vgaarb: pci_notify
root@bpi-r4-v11:~#
wonder about the Firmware version is blank 
seems not ready yetā¦and eeprom was not loaded, but itās there
root@bpi-r4-v11:~# ls /lib/firmware/mediatek/mt7996/
mt7996_dsp.bin mt7996_eeprom_tm.bin mt7996_wa.bin mt7996_wm_tm.bin
mt7996_eeprom.bin mt7996_rom_patch.bin mt7996_wm.bin
but const is labeled defaultā¦maybe the default bin is the file in lib/firmwareā¦but see no status, that it was loaded successfully.
but i do not see the error āInvalid default binā, so i guess it is loaded successfully
@dangowrt have you got your card?
Hi, Iām currently having trouble finding any options to buy this card.
The aliexpress shop linked above doesnāt have any and seems to be in a state of disarray (no products showing for me). Also the shop that is linked in the wiki is dead and references to the joom shop (I found this here: Elektronik, Marke: BPI ā Preise mit Lieferung von Artikeln aus China im Joom Onlineshop) But neither carry the wifi module.
Is this available anywhere and if so where?
Would really like to have one or any other compatible wifi module for the R4.
The official Wifi NIC is not available for sale yet. Frank (and some others) only received one as a developer preview.
You will have to wait with us others, who are eagerly awaiting the public sale of the module.
Itās getting kind of silly waiting this long. Even if it doesnāt work there should be a way to buy that includes a disclaimer about it not working properly yet. Iād love to help figure the driver issue out. I throughly enjoy working on those types of issues, but canāt help because I canāt yet buy it. I bought my bpi-r4 in January thinking by now the NIC would be out. Iām starting to get worried that it wonāt be released. The BPi-R4 is still fun to play with, but maybe I should have waited for the nic before I bought it. Hope we see it for sale soon.
Or at least if @sinovoip or @simon could be a bit more communicative about this. Like what phase they are in. Are they already mass producing it or do they need to iron out the firmware issues first or even need to design a āv1.1ā, so to say, before it goes into mass production and then public sale.
That would be really helpful.
Agreed. Iām staying hopeful it will all work out though. I have a 7916 in mine for now and itās working great with 6ghz and my quectel rm520n-gl. Very fast.
Sorry Frank, I wasnāt trying to derail your thread. Feel free to delete my posts.
Is there any other NIC available? like is there any example for stick a BE200 or sth on R4 and make it work?
You can use any mt76xx/mt79xx (e.g. mt7615 or mt7916) on the mpci slots. Just be sure sw4 is off
1 Like
I have been busy with BE14 calibration, OpenWRT One testing and R4ās case, so donāt have much time, sorry.
We are modifying the layout of BE14 recently, and the V1.0 version will be produced soon, but the quantity will not be too large, just about 30 to 40. After that, it will be hundreds.
5 Likes
This sounds good!
Please take the time you need for a quality product 
.
PS:
I have no problem to resolder or replace some resistors. But some people have issues with that!
So, my support for a longer development you have 
.
Does code need to be changed compared to the be14 i have? Currently have firmware loading part (233 files) working so far,but cannot start interfaces. If it changes i do not need to put much work into itā¦
So,how about BE19, will it be produced?
Thanks!
The workload of wifi7 NIC is a bit heavy, including test equipment upgrade, calibration, RF matching, factory test program debugging, etc.
There is no need to change the software, we are just optimizing the hardware.
MT76 and MP3.1 are both OK, but they should be kernel 5.4, not the latest 6.1. Porting to 6.1 is really a lot of work. You can ask steven for help
BE19 is currently planned, but it will be after BE14 and OpenWRT One, so there is no specific MP time yet.
3 Likes
Super news, there is new thing to blame - openwrt one nonsense slowing down the main profitable projectā¦