From which router is that pcb?

Hi guys,

I've got a WIFI Router PCB from a friend that found it while dumpster diving.

I've already got an complete WS-AP3915i-ROW and an WS-AP3825i from him, have both converted to an actual openwrt..but don't really use them already.

Now I've got a main-pcb from a router, the two needed wfifi boards (2,4 and 5Ghz), the pigtails and 6 external antennas. There was no housing for the router so I don't know what it may be..
I think it's Extreme Networks too, possibly outdoor, but don't really know.

The CPU is an P1020NXE with 800Mhz (I think).

here are some images:

The board has only two RJ45 connectors ,so I think it's POE only. on the other side of the pcb are two pin headers, 16 pins and 4 pins, think the 4 pins would be the serial console interface. Next to that connector there is a unpopulated footprint "JP2" and it has a plus sign on the left side..but no voltage..what could I connect there? (currently don't have a POE source)

Kind regards,
Addendum: The Board is approx 170x180mm, the distance of the two RJ45 Sockets is 70mm (middle to middle)

The mainboard looks to be a variant or a sibling of the Fortinet/Meru 832e. The antennae are obviously by Extreme Networks ("Zebra"?).

The Antennas are WS-AO-DS02360N4 from Extreme Networks.

2dBi ..nothing special. They have N Connectors at the bottom and in my case there are 6 Adapters from N to SMA plugged in.

Have searched for pics from the Fortinet and found which suggests regarding the type label it is an 832e ..but it has holes all over the surface...??
Is there an outer housing missing? Found another picture where it has two black rubber sea lings at the bottom and 3 Antenna connectors between..that may be possible.

But the PCB I have has the RJ45 connectors in the axis of the board and no USB Connector on the board.
This thing: seems to have exactly that N type connectors for the Antennas that are on my Antennas plugged in ...


Nice. I'm no eco-warrior, but looking through those pictures one is reminded about the vast extent of electronic waste out there, and also the good OpenWrt can do to reduce that to some extent. I watched that video Apple released about meeting mother nature and didn't buy it. Releasing new models so frequently cannot be great for the environment.

1 Like

..let's see if I can do something useful with it. I've needed outdoor Antennas and told that to my friend..that's the result...


In the meantime I've connected two labor supplies in series and both of them (48V) to one of the RJ45 is doing something :slight_smile:
I've connected a scope to ground an fiddeled around with the 4 pin connector..that's J2
Pin1 is 3V3, Pin2 is GND, PIN3 is TX and got it, it's Rx.
So further I've connected an CH3xsomething USB to TTL Converter and looked with a terminal programm..

U-Boot 2010.12.6 (Feb 26 2015 - 19:39:21) (primary)

Power LED ON
CPU0:  P1020E, Version: 1.1, (0x80ec0011)
Core:  E500, Version: 5.1, (0x80212051)
Clock Configuration:
       CPU0:800  MHz, CPU1:800  MHz, 
       CCB:400  MHz,
       DDR:333.333 MHz (666.667 MT/s data rate) (Asynchronous), LBC:25   MHz
L1:    D-cache 32 kB enabled
       I-cache 32 kB enabled
Board: P1020RDB CPLD: V15.15 PCBA: V15.0
rom_loc: nor upper bank
SD/MMC : 4-bit Mode
eSPI : Enabled
I2C:   ready
SPI:   ready
DRAM:  Configuring DDR for 666.667 MT/s data rate
DDR: 256 MiB (DDR3, 32-bit, CL=6, ECC off)
Testing DRAM from 0x00010000 to 0x000fffff
DRAM test phase 1:
DRAM test phase 2:
DRAM test passed.
L2:    256 KB enabled
PCIe1: Root Complex of mini PCIe SLOT, x1, regs @ 0xffe0a000
  01:00.0     - 168c:003c - Network controller
PCIe1: Bus 00 - 01
PCIe2: Root Complex of PCIe SLOT, x1, regs @ 0xffe09000
  03:00.0     - 168c:0033 - Network controller
PCIe2: Bus 02 - 03
In:    serial
Out:   serial
Err:   serial
Net:   eTSEC1, eTSEC3
### JFFS2 loading '/home/config/shadow'
rescan: First time in use
Scanning JFFS2 FS:   .   done.
load: loaded '/home/config/shadow' to 0xfac05d0 (122 bytes)
### JFFS2 load complete: 122 bytes loaded
Hit 'd' for diagnostics, or any other key to stop autoboot:  2  1  0 
## Executing script at 0ff33414
check for watchdog, limit=3...
watchdog count hit previously, running older image...
bSaveEnviron=0 image_order=primary
### JFFS2 loading 'primary.gz.uImage'
load: loaded 'primary.gz.uImage' to 0xa000000 (9330483 bytes)
### JFFS2 load complete: 9330483 bytes loaded
ready to boot kernel... [DTB 1]
## Booting kernel from Legacy Image at 0a000000 ...
   Image Name:   Linux Kernel Image (LZMA)
   Image Type:   PowerPC Linux Kernel Image (lzma compressed)
   Data Size:    9318131 Bytes = 8.9 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK
## Flattened Device Tree blob at 0a8e2f33
   Booting using the fdt blob at 0xa8e2f33
   Uncompressing Kernel Image ... OK
   Loading Device Tree to 09ffa000, end 09ffffff ... OK
Using P1020 RDB machine description
Memory CAM mapping: 64/64/64 Mb, residual: 63Mb
Linux version 2.6.35-- (root@bld-debian1) (gcc version 4.9.2 (crosstool-NG 1.21.0) ) #2 SMP Thu Dec 5 15:46:45 EST 2019
CPU maps initialized for 1 thread per core
bootconsole [udbg0] enabled
setup_arch: bootmem
Found FSL PCI host bridge at 0x00000000ffe09000. Firmware bus number: 0->1
PCI host bridge /pcie@ffe09000  ranges:
 MEM 0x00000000a0000000..0x00000000bfffffff -> 0x00000000a0000000 
  IO 0x00000000ffc10000..0x00000000ffc1ffff -> 0x0000000000000000
/pcie@ffe09000: PCICSRBAR @ 0xfff00000
Found FSL PCI host bridge at 0x00000000ffe0a000. Firmware bus number: 0->1
PCI host bridge /pcie@ffe0a000  ranges:
 MEM 0x0000000080000000..0x000000009fffffff -> 0x0000000080000000 
  IO 0x00000000ffc00000..0x00000000ffc0ffff -> 0x0000000000000000
/pcie@ffe0a000: PCICSRBAR @ 0xfff00000
MPC85xx RDB board from Freescale Semiconductor
arch: exit
Zone PFN ranges:
  DMA      0x00000000 -> 0x0000c000
  Normal   empty
  HighMem  0x0000c000 -> 0x0000ff7f
Movable zone start PFN for each node
early_node_map[1] active PFN ranges
    0: 0x00000000 -> 0x0000ff7f
MMU: Allocated 1088 bytes of context maps for 255 contexts
PERCPU: Embedded 7 pages/cpu @c238e000 s7872 r8192 d12608 u65536
pcpu-alloc: s7872 r8192 d12608 u65536 alloc=16*4096
pcpu-alloc: [0] 0 [0] 1 
Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 64896
Kernel command line: mtdparts=ec000000.nor:62848K(FS),128K(CALIB),512K(BootPRI),128K(NVRAM1),128K(NVRAM2),128K(NVRAM3),128K(NVRAM4),128K(NVRAM5),128K(NVRAM6),128K(NVRAM7),128K(NVRAM8),128K(CFG2),128K(CFG1) BOOT_BOOTROM="U-Boot 2010.12.6 (Feb 26 2015 - 19:39:21) (primary)" BOOT_KERNEL=primary console=ttyS0,115200n81 panic=30 ro
PID hash table entries: 1024 (order: 0, 4096 bytes)
Dentry cache hash table entries: 32768 (order: 5, 131072 bytes)
Inode-cache hash table entries: 16384 (order: 4, 65536 bytes)
Memory: 224772k/261628k available (33960k kernel code, 36856k reserved, 212k data, 153k bss, 30952k init)
Kernel virtual memory layout:
  * 0xfffbf000..0xfffff000  : fixmap
  * 0xff800000..0xffc00000  : highmem PTEs
  * 0xff7d8000..0xff800000  : early ioremap
  * 0xcd000000..0xff7d8000  : vmalloc & ioremap
SLUB: Genslabs=13, HWalign=32, Order=0-3, MinObjects=0, CPUs=2, Nodes=1
Hierarchical RCU implementation.
	RCU-based detection of stalled CPUs is disabled.
	Verbose stalled-CPUs detection is disabled.
mpic: Setting up MPIC " OpenPIC  " version 1.2 at ffe40000, max 2 CPUs
mpic: ISU size: 256, shift: 8, mask: ff
mpic: Initializing for 256 sources
clocksource: timebase mult[5000000] shift[22] registered
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 512
mpic: requesting IPIs...
Processor 1 found.
Brought up 2 CPUs
NET: Registered protocol family 16
mpc8xxx_add_controller RESET button signal is high
mpc8xxx_add_controller hwirq 47
PCI: Probing PCI hardware
pci 0000:00:00.0: ignoring class b20 (doesn't match header type 01)
pci 0000:00:00.0: PCI bridge to [bus 01-ff]
pci 0001:02:00.0: ignoring class b20 (doesn't match header type 01)
pci 0001:02:00.0: PCI bridge to [bus 03-ff]
pci 0000:00:00.0: BAR 9: can't assign mem pref (size 0x100000)
pci 0000:01:00.0: BAR 6: assigned [mem 0xa0020000-0xa002ffff pref]
pci 0000:00:00.0: PCI bridge to [bus 01-01]
pci 0000:00:00.0:   bridge window [io  0xff7ec000-0xff7fbfff]
pci 0000:00:00.0:   bridge window [mem 0xa0000000-0xbfffffff]
pci 0000:00:00.0:   bridge window [mem pref disabled]
pci 0000:00:00.0: enabling device (0106 -> 0107)
pci 0001:02:00.0: BAR 9: can't assign mem pref (size 0x100000)
pci 0001:03:00.0: BAR 6: assigned [mem 0x80200000-0x8020ffff pref]
pci 0001:02:00.0: PCI bridge to [bus 03-03]
pci 0001:02:00.0:   bridge window [io  0xff7d9000-0xff7e8fff]
pci 0001:02:00.0:   bridge window [mem 0x80000000-0x9fffffff]
pci 0001:02:00.0:   bridge window [mem pref disabled]
pci 0001:02:00.0: enabling device (0106 -> 0107)
bio: create slab <bio-0> at 0
Freescale Elo / Elo Plus DMA driver
Switching to clocksource timebase
NET: Registered protocol family 2
IP route cache hash table entries: 2048 (order: 1, 8192 bytes)
TCP established hash table entries: 8192 (order: 4, 65536 bytes)
TCP bind hash table entries: 8192 (order: 5, 229376 bytes)
TCP: Hash tables configured (established 8192 bind 8192)
TCP reno registered
UDP hash table entries: 128 (order: 1, 8192 bytes)
UDP-Lite hash table entries: 128 (order: 1, 8192 bytes)
NET: Registered protocol family 1
fsl-elo-dma ffe21300.dma: #0 (fsl,eloplus-dma-channel), irq 20
fsl-elo-dma ffe21300.dma: #1 (fsl,eloplus-dma-channel), irq 21
fsl-elo-dma ffe21300.dma: #2 (fsl,eloplus-dma-channel), irq 22
fsl-elo-dma ffe21300.dma: #3 (fsl,eloplus-dma-channel), irq 23
highmem bounce pool size: 64 pages
JFFS2 version 2.2. (NAND) © 2001-2006 Red Hat, Inc.
msgmni has been set to 312
io scheduler noop registered
io scheduler deadline registered (default)
Serial: 8250/16550 driver, 2 ports, IRQ sharing enabled
serial8250.0: ttyS0 at MMIO 0xffe04500 (irq = 42) is a 16550A
console [ttyS0] enabled, bootconsole disabled
console [ttyS0] enabled, bootconsole disabled
serial8250.0: ttyS1 at MMIO 0xffe04600 (irq = 42) is a 16550A
ec000000.nor: Found 1 x16 devices at 0x0 in 16-bit bank
Amd/Fujitsu Extended Query Table at 0x0040
cfi_fixup_kern_acc mtd cb8a8600 write and erase f() kern access setup
number of CFI chips: 1
13 cmdlinepart partitions found on MTD device ec000000.nor
Creating 13 MTD partitions on "ec000000.nor":
0x000000000000-0x000003d60000 : "FS"
0x000003d60000-0x000003d80000 : "CALIB"
0x000003d80000-0x000003e00000 : "BootPRI"
0x000003e00000-0x000003e20000 : "NVRAM1"
0x000003e20000-0x000003e40000 : "NVRAM2"
0x000003e40000-0x000003e60000 : "NVRAM3"
0x000003e60000-0x000003e80000 : "NVRAM4"
0x000003e80000-0x000003ea0000 : "NVRAM5"
0x000003ea0000-0x000003ec0000 : "NVRAM6"
0x000003ec0000-0x000003ee0000 : "NVRAM7"
0x000003ee0000-0x000003f00000 : "NVRAM8"
0x000003f00000-0x000003f20000 : "CFG2"
0x000003f20000-0x000003f40000 : "CFG1"
Fixed MDIO Bus: probed
/proc/ptp_1588 created 
IEEE1588: disable on the system.
eth0: Dropping NETIF_F_SG since no checksum feature.
gianfar_pool_init 942: init pNetPool c2182bfc
gfar_probe 2538: bufLen is 2224 rcvbuf len 1980 gChnEthPool c2182bfc
eth0: Gianfar Ethernet Controller Version 1.4-skbr1.1.5, d8:84:66:31:8f:d4
eth0: Running with NAPI enabled
eth0: RX BD ring size for Q[0]: 256
eth0: RX BD ring size for Q[1]: 256
eth0: RX BD ring size for Q[2]: 256
eth0: RX BD ring size for Q[3]: 256
eth0: RX BD ring size for Q[4]: 256
eth0: RX BD ring size for Q[5]: 256
eth0: RX BD ring size for Q[6]: 256
eth0: RX BD ring size for Q[7]: 256
eth0: TX BD ring size for Q[0]: 256
eth0: TX BD ring size for Q[1]: 256
eth0: TX BD ring size for Q[2]: 256
eth0: TX BD ring size for Q[3]: 256
eth0: TX BD ring size for Q[4]: 256
eth0: TX BD ring size for Q[5]: 256
eth0: TX BD ring size for Q[6]: 256
eth0: TX BD ring size for Q[7]: 256
IEEE1588: disable on the system.
eth1: Dropping NETIF_F_SG since no checksum feature.
gianfar_pool_init 942: init pNetPool c2182bfc
gfar_probe 2538: bufLen is 2224 rcvbuf len 1980 gChnEthPool c2182bfc
eth1: Gianfar Ethernet Controller Version 1.4-skbr1.1.5, d8:84:66:31:8f:d5
eth1: Running with NAPI enabled
eth1: RX BD ring size for Q[0]: 256
eth1: RX BD ring size for Q[1]: 256
eth1: RX BD ring size for Q[2]: 256
eth1: RX BD ring size for Q[3]: 256
eth1: RX BD ring size for Q[4]: 256
eth1: RX BD ring size for Q[5]: 256
eth1: RX BD ring size for Q[6]: 256
eth1: RX BD ring size for Q[7]: 256
eth1: TX BD ring size for Q[0]: 256
eth1: TX BD ring size for Q[1]: 256
eth1: TX BD ring size for Q[2]: 256
eth1: TX BD ring size for Q[3]: 256
eth1: TX BD ring size for Q[4]: 256
eth1: TX BD ring size for Q[5]: 256
eth1: TX BD ring size for Q[6]: 256
eth1: TX BD ring size for Q[7]: 256
Freescale PowerQUICC MII Bus: probed
Freescale PowerQUICC MII Bus: probed
tun: Universal TUN/TAP device driver, 1.6
tun: (C) 1999-2004 Max Krasnyansky <>
i2c /dev entries driver
PowerPC Book-E Watchdog Timer Loaded
nf_conntrack version 0.5.0 (3512 buckets, 14048 max)
ip_tables: (C) 2000-2006 Netfilter Core Team
TCP cubic registered
NET: Registered protocol family 10
NET: Registered protocol family 17
Freeing unused kernel memory: 30952k init

init started: BusyBox v1.17.4 (2019-12-05 15:13:42 EST)

starting pid 44, tty '': '/etc/rc.d/rcS'
data plane cores = 1
fast path cores = 1
rcS: watchdogDis 0 serNum 15390642085K0000
Read/Write Crypto Section successfully ! 
Waiting for jffsap_drv: trap Monitor has been suspended.
2_gcd_mtd and flush-mtd to finish, wait time 0 seconds.
ap_drv: trap Monitor will resume shortly.
eth0 hw addr: D8:84:66:31:8F:D4
eth1 hw addr: D8:84:66:31:8F:D5
update_compliance_file: /flash/v3_confData.bin up to date (Ver in Image AP3865e-77 Ver on AP AP3865e-78 format 129)
update_compliance_file:Imported Compliance Table V4.2 dated on 7/10/2020 (AP38XX_Power_Settings_V9.xlsx)

3708+0 records in
3708+0 records out
3708+0 records in
3708+0 records out

MODEL        AP3865e
SERVICEATTRS ac_manager,ru_manager
AC_HOSTNAME  Controller
SERIALNUMBER 15390642085K0000
AP_FLAG      24
AP_MODE      0

starting pid 398, tty '': '/sbi
 (none) ppc #2 SMP Thu Dec 5 15:46:45 EST 2019 (none)

(none) login: ap_drv: trap Monitor will resume in a few seconds.type or paste code here

..So it thinks about itself it is an Extreme Networks AP3865e.

I've found an AP3865e in the Database of supported devices...sorry, I have to read something here first..

Regards holm

How long does router circuitry last? Say a router is in a landfill for twenty years. Would it still work?

:slight_smile: you first have to define what is a router in your eyes.
..a small wifi device for home use?
Or a rack full of professional gear in a computing center?

There are some differences..on money to loan and on years they are usable.

I get FTTH on Nov. 15, have to change my main router because of this. But that Router PCB
was an Enterprise outdoor WIFI router some years before, it should still have enough usefulness to provide a reasonable WIFI net in my Backyard.


...which is wrong. there is something in the csv file, but that seems to have nothing todo with that router on my table.

So what todo next? I've reset the router to factory configuration and "admin" + "new2day" worked for uboot and the system. I'm root..but that thing isn't meant to be configured w/o an external "controller", some software, radius and so on.. but it could do 1,7GBits/s (regarding some website) and could serve open air festivals :slight_smile:
I was a system administrator from a small ISP, but this was twenty years before now.. and I'm not a Linux Guy (but FreeBSD). I have some Ubuntu OS in a VM for the buildroot of tasmota software on the system.. What do you think? It is worth the effort? How difficult is it to get OpenWrt supportet on this device (CPU P1020, periphery..)


It looks like you might be able to get into a uboot shell by pressing 'd' during boot (see message in your bootlog)

Further, The P1020 seems to be well supported already by openwrt as "powerpc_8540", target name: "mpc85xx", subtarget "p1020"


looks similar to yours, you could try ramloading by tftp from uboot the initramfs image for that one:

or other images from that target, and then read through the serial output to determine what exactly works for you on a recent image/kernel

Edit: These look like instructions you could adapt to your case for ram-loading:;a=commit;h=7e614820a89208c4e91a3a5f9de07a5402accdaa


Interesting, it is so similar to the Fortinet mainboard that there must be some OEM situation going on.

Most of these enterprise AP boards were semi-generic hardware designs made by Compex, and later another Taiwanese manufacturer (maybe Lite-On?) that I don't remember now. The market differentiation was in their software.

1 Like

Yes, I've already done that. The factory reset (pressing the button when the OS is up and running) set even the root login to admin and new2day so I can log in into the router. Only looked around a little until now.

I've already flashed an WS-AP3915i-ROW and an WS-AP3825i to OpenWrt with Luci, so that process is already known to me.

But porting it to the 3865e is another thing. I know almost nothing about the flash layout and such things so it may be a little bit dangerous to overwrite things in the flash. On the other side I'm a company of my own and the singe one employee of that company. (Doing repair of electronic devices mostly, no, no radios or TV's..) I'm busy with other things..besides of the fact that I still have to do many things in my house bought shortly before now..

Sure, I want an outdoor ap for the backyard, but maybe it's better to put one of the mentioned above ones into a platic lunchbox or so.. My AP3865 doesn't have any's the plain what...

Yes, I think you are right. Until I see the opposite, I think the AP3865e and the Fortinet/Meru 832e share the same hardware.


Physical hardware:
Is the chassi crushed?
Is the PCB physically broken or beant, if a via or trace is broken the job gets a lot harder.
A BGA over a bent PCB means a totally destroyed board.

Is a capacitor crushed, maybe it can be replaced without destroying the PCB.
The same with connectors.

Operational lifetime events:
Has it during its best days been trough a over voltage event, lightning or ESD event (why was is abandoned to begin with?). Then it can be all from fried and dead to a little strange working from time to time.

Environment storage:
On top of that if we talk about a dumpster or landfill for 20years. Water, oxygen, exposure, heat, cold, salts and unknown chemicals for at least 20cycles.
Well these kinds of electronics use RoHC rules and “green protection chemicals that break down” and are cheep so don’t expect some kind of mil-std-810 survival electronics.

Time and evolution:
But we also have the evolution it self, and that moves fast in electronics. The year 2003…we still used diskettes for smartdrv and to install Win XP back then. CD and DVD was in mid life. No smartphones.
What was the internet speed back then? I don’t even remember if the ADSL had arrived in Sweden back then or if the 52kbit modem (but that was NOT the actual speed! It just looked good) was still in use. But if we had the ADSL it was only at best 500kbit/s.

1 Like

..don't forget other physical effects, like Eprom's, Eeprom's and Flash Chips that are loosing information over the time, since that information is only an electrical charge on a very very small capacitor electrode to distinguish between a 0 and a 1. Technical evolution? .. that means splitting that charge in currently up to 8 different charge levels in a single cell.. must be a good idea, since people keep buying that...
That means: is the firmware is gone, then the device is gone. Someone remembers that Kodak for Example guarantee'd that ther CDROM's have a lifetime of at least 100 years?

Hmm. :expressionless:

FreibergNet was a small ISP born out of the need to sell FreeBSD Daemons and Linux Penguins worldwide. I was the one that triggered a friend of mine, the owner of an advertising company, to make FreeBSD Daemon Plushies

(the real one: )

and to sell them. That was the cause for an Guy named Larry Ewing to design a Plushie for Linux too, that Penguin that we all know.., since the Linux logo until then was something graphical with a pyramid, a ball and so on..nothing that was usable as plushie..
As a consequence the company was getting overloaded from all over the world in making toys and selling them, so my friend hired me out of the University of mining and technology Freiberg, Germany, to "make Internet" and "Webshops". 2000 the internet bubble imploded, that's another story. I'm the last one from FreibergNet, I own the name and still have customers from that time..
Ok.. back to the CDROM's .. the Advertising Company-half made money with typesetting different Newsletters (Paper!) and other advertising jobs like trade fair constructions and so on. Each Newspaper was burnt onto a CDROM, they where never read, only put in some containers for reference and backup. After 2 Years I' taked a look onto this CDROM's ..!!!!! the golden layer from most of them seperated partly from the polycarbonate disc. 100 years can be a very short time..

Hope you had fun.. let's talk about the AP3865 again.. and I hope you can understand why I have slightly different sights when one talks about a router.. A router in my Eyes has a 19 inch form factor and mostly they are labeled Cisco or LevelOne, Ascend, Lucent, Huawai or even Extreme Networks..
I must still have an Ascent Max and some Cisco 261x laying around somewhere in my warehouse.. that are Routers.



BTW: sorry for my poor english, I'm a german and in fact I never learned english in a school (..but russian, almost forgotten now). My english comes only from communications in the internet..

1 Like

It was you that came with the dumpster garbage you called a router and asked what it is.

And AP3865 isn’t even a router. It is a outdoor access point so it doesn’t have any meaningful hardware to ever be a router.

And when we talk about outdoor access points and my last post, water, heat, cold, sunlight, lightning and ESD…
Why was it dumped to begin with?

On the other part about routers, they actually have a definition.
Rack mounted doesn’t even exist in the definitions I have seen.

Labels doesn’t really mean anything, that is only the manufactures name. They most often doesn’t do the hardware in the box. They just sell it as a manufacture.

Businesslike, yea ok if you say so.

So I give you a total network and internet speed of max 15Mbps for your new router, if and when you get it working with OpenWrt. Then we can see afterward what you got?

Seems to me that there are people that are more equal as other people. No wonder, if there is an "Guidance Counsellor" involved that simply couldn't be wrong and is besides of ranting to ppl running amok now.
Please, please, give him a gun..

Hey, do what you want with me (there are things that are much ore important as old WIFI Apps..), nevertheless I'm rolling on the floor laughing...

Best Regards,


You obviously already have one flagged post…
This is your first tread in the forum so I give you a real warning to begin with. You do realize I can flag this whole tread and it is gone when you talk like this. And the chance is that you are gone also.

The only one running amok is pretty much you when you start bringing weapons to your “stick to the subject” tread.

And to be honest you are pretty bad at sticking to your own subject, which you demand all others to do.
And what exactly is the subject and goal of this tread?
I talk about your device all the time but you are to stubborn to realize it while barking to much.

So what performance do you expect from a access point you absolutely want to call a router with a cpu that in the above link have 800MHz and 2cores?
I doubt it has a switch and as you said it needs PoE to run so you at least need a PoE injector to get it running.
The only thing that CPU can do is to run luci and system admin.

So my bet still stands, you will get max 15Mbit/s trough that router no matter how high you scream.

And how do I know this? Well that is because I am a guidance counselor so I have seen this in the forum many times already when switches and access points are run as routers with this kind of cpu. And I can promise that no one seems impressed of that kind of router.
The thing is that the actual design work switches and access points do is hardware offloaded so the data speed they produce ain't a matter of cpu power.
Routers on the other hand are all about raw cpu power, number of cores, memory, and if needed crypto accelerators.

To try to bring this discussion back on topic:

That's why I suggested RAM-loading the initramfs image. That should leave the flash untouched.

You might be lucky, and the suggested image boots right up, only complaining about the absent USB-ports. The hardware details of the two looks similar enough, that it might just work.

The 15Mbit/s figure posted above seems low though for a dualcore 800Mhz PPC. I mean, the single-core 500 MIPS in the RTL838x is supposedly able to route 80Mbit/s, and the MT7621 is also only a DualThread MIPS, yet capable of routing Gigabit, or at least hundreds of Megabits, depending on configuration.

I wasn't expecting this device to be a rocket, but would have a guessed at least 200Mbit/s. The 2 Gigabit interfaces seem to be independent network interfaces. I'd be curious to read it, if you make more progress/experiments/speedtests with this device.