Obihai OBi20x/30x + OBi1000 + OBi50x + OBi2000 firmware mods

charterengr · 2025-01-19T18:36:53+00:00

Can we pickup where we left off here....... naf? OBi1FW? jsolo1? others?

Jargin

Down the rabbit hole. Maybe I can update the d230 firmware with the patched obiapp from the other firmwares. Got the firmware unpacked, headers and bodies, now to figure out how to read the ubifs to update the files.

naf

Darkbotic

Installing the modded fw doesn’t clear the settings, so no relinking necessary. However, the modded fw is going to have the exact same problems as stock since its just a couple of binary mods to the otherwise identical stock firmware.

naf

Jargin I’ve been unable to get a d230 to work with aa2k’s firmware

assuming “unable to work” means “unable to work with GV”, the d230 binary doesn’t contain any of the interface code to work with GV so that would be expected.

a more reasonable method to attempt to get a d230 to talk to GV might be to run asterisk on the d230 as a sorta self-hosted GVGW

Jargin

Oh I thought the d230 did work with GV since it’s on obifirmware.

Timbergetter

I have an Obi-302 that is in a bricked state and I’m wondering if there are any tools out there that might bring it back to life. The current state of the device presents as follows:

• The device powers on and settles in the state where I get an alternating red/green led.  Using the reset pin does not seem to change this behavior.
• There is no phone access to configuration.
• It does respond to pings at IP 192.168.10.1 but generally no web page opens.  Occasionally I do get a firmware update web page but this fails to load any firmware.  
• I connected a console to the J17 serial port and observed the Obi booting up consistently. The booting log does report that partitions 3, 4 and 5 are corrupted.

Are there any resources available that might allow the affected partitions to be refreshed at a low level?

naf

Timbergetter

to quote myself from another forum a long time ago: “flashing red/green indicates that the bootloader detected a discrepancy between the contents of the kernel or root partition read from flash with their expected contents based on the most recent firmware update. if flashing firmware from the recovery mode doesn’t fix that, i would assume that the flash chip is non-functional, but try swapping with a known-working power supply just to be sure”

Timbergetter

You wouldn’t happen to know how I might get my hands on a hex dump from the flash chip of a OBi302?
If I could do that I would be tempted to try a new flash chip. I’ve been waiting for an excuse to fire up a chip programmer that I’ve had sitting on the shelf for a couple of years now. 🤔

naf

Timbergetter

sure.
first you might want to try to read the “OB100 UNIT INFO” section of your flash at 0×40000-0×50000.
its the only thing that is unique per device (mac address, pki keys, etc).

Timbergetter

naf Many thanks for that vital tip. However unless I can track down a donor OBi302 I will still need a hex dump to attempt the upload. It would be awesome if you could link me with a hex dump.

Timbergetter

ie from 0×000000-0xffffff or whatever sub-set of that is necessary for a viable system.

naf

Timbergetter

i sent you a PM with a hexdump from a obi200 with the “OB100 UNIT INFO” section erased.

Here’s an idea of the flash layout:
0×00000000 uboot
0×00040000 OB100 UNIT INFO
0×00050000 kernel uboot image
0×002d0000 recovery kernel uboot image
0×00400000 OBI_PARAM_PT
0×00420000 OBI_PARAM_PT
0×00460000 OBI_PARAM_PT
0×00470000 header of firmware file
0×00480000 RootFS squashfs
0×00b40000 scratch jffs2
0×00c00000 obi squashfs
0×00e40000 recovery squashfs
0×00f00000 bluetooth squashfs

reg/green flashing means uboot is intact and noticed that the checksums written by the last firmware update at 0×470000 dont match the contents at 0×50000/0×480000/0xc00000

Timbergetter

naf “ i sent you a PM with a hexdump from a obi200 with the “OB100 UNIT INFO” section erased.”

Thanks to the hex-dump kindly provided by naf I was able bring my OBi302 back to life by replacing the flash chip which I loaded with the hex-dump contents together with UNIT_INFO data salvaged from my old failing chip. Now that the ATA is working again I am looking at new exciting ways I might be able to brick it again 🙂 But seriously, I do see that the boot loader comes with a tftp client so I am wondering if there might be a reason why the hex-dump couldn’t be loaded into place using that method also. It seems to be expecting a tftp server at 192.168.15.102.

TommyC

Not sure which direction to go next. I have an Obi202 that seems to have bit the dust. Will only flash the bypass-uboot-mod firmware and the web interface is stuck in the upload firmware screen loop. So, I purchased a brand new Obi202 off of eBay. Seems to be fully operational. Had very old firmware which I flashed straight to the arrynrob13.fw successfully. I disabled ObiTalk and other auto provisioning goodies. Followed the Google Voice provisioning page and generated tokens for both of my existing GV numbers (one at a time and either SP1 or SP2) and paste in the correct provisioning XML with System Start and nothing I try has worked. Regardless of which one I try to auto provision after reboot SP1 and SP2 continue to show “Service Not Configured”.

Any insight or help would be very much appreciated. I miss my “free” landlines.

naf

TommyC

“Service Not Configured” makes it sound like the provisioning step is failing somehow.
If provisioning is successful (after a variable delay of up to a minute and a couple of reboots) it will reset the ‘ITSP Provisioning’ Method back to ‘Disabled’, so you should be able to check that to determine if its taking the provisioning xml successfully. If not, its most likely a networking problem between obi and internet (dns, routing, etc) maybe? Looking at syslog might be helpful in that case.

TommyC

Funny you should mention syslog. I just now finally figured out how to run a syslog server locally and the second to last entry on reboot of the obi202 give this, “PROV: Unknown Script Command @ statement 1!<SP1:>”

TommyC

Sometimes it simply takes asking a question to come to an answer. As soon as I posted that log message it dawned on me that the SP1: etc in front of those URLs are simply a guide as to which URL to use, not some magical protocol for provisioning. Removed that and holy guacamole. Thank you taking a look and helping my thought process along.

kenobi

Hi Everyone,

I’ve been admiring and recommending OBI ATAs for years based on how powerful they are but being adequately served by other VoIP gear and being extremely cheap I never had one myself until now which was the result of a generous someone in a forum far far away gifting me an OBi200 out of the blue. Unfortunately, they didn’t also supply an OBiBT dongle and these look to be rare now and likely to be expensive if one turns up (did I mention I was cheap?) I’ve also come across remarks that it didn’t work that well anyway, so I’d be interested in remarks about that. Anyway, that started me looking at what alternate firmware there was out there which is how I’ve ended up here.

Firstly, thanks to those who contributed to the development of the firmware. Now if I may make a pest of myself:

Am I correct to assume that references to non OBi dongles have been removed from the (in-kernel?) Bluetooth driver? I see that is the case with the LKM for the WiFi dongle.
The firmware is pretty old. It would be good to get more recent fixes if there are any (unfortunately, later bugs too 🙁) I’m unable to get GV so I’m wondering whether OBI300 firmware will run on an OBI200? And what about HP / Poly ATA 400 firmware? Has anyone tried? Is the hardware the same? I can’t solder and I don’t have a spare OBI so I don’t want to brick the one I have. The latest 400 firmware has a 4.9.x kernel and IPv6 support which might be useful to some.
Not that I really have much of an idea of how to go about this but it would be really great to run a properly maintained Linux distro. Would porting OpenWRT or another Linux (or even BSD!) be exceptionally difficult given the lack of any source code release - that I know of at least? HP could well be more amenable to releasing source code than Poly or OBI was in the past. Has anyone ever asked? OpenWRT has support for other routers running Marvel SOCs but not the 88E7200-LKJ2 I saw a reference to that some place. My guess is that the SLIC interface might be the only thing that requires reverse engineering and developing a device driver for. Seems that Asterisk already has GV support thanks to some work naf contributed so the only extra thing might be an Asterisk channel driver to drive the SLIC driver. Am I missing something? Other than the skill and determination to see this through myself, that is. Oh, yes some people might like a nice GUI for Asterisk. I’ve no idea whether what’s available for Luci or whatever web GUI platform the distro uses would be decent.

naf

kenobi I’m wondering whether OBI300 firmware will run on an OBI200? And what about HP / Poly ATA 400 firmware?

Judging solely by the firmware’s own restrictions in loading firmware updates: Obi200/300 firmware are interchangeable, but Poly400 is a separate entity.

kenobi

naf Thanks. Of course, you could patch out the check in your sleep… Just sayin’

naf

Timbergetter I do see that the boot loader comes with a tftp client

the most recent boot loader also comes with a “bootdelay=0” setting, so even if you have a serial console you cant interrupt the boot process to go to a uboot shell to launch that tftp client.

Timbergetter

naf I am finding that arrynrob13 allows break out into Uboot shell so I might just try a tftpboot out of curiosity just before my new flash chip order arrives. Just on that subject, I have ordered the newer generation flash chips (JVFQ) to replace the FVFG chip that is likely turning up its toes presently in my OBi302. Do you expect I may have issues with this substitution? I went with with the JVFQ chips because that is what my other OBi200 has.

« Previous Page Next Page »