Saucered Unifi AC LR

It's a shot in the dark, but here it goes.
I bricked my saucer with a faulty usb to serial connection,
but I manage to hook it up with a good one, though it's too late.

Does anyone think it can be revived?

U-Boot unifi-v1.6.7.249-gb74e0282 (Apr 25 2017 - 09:18:46)

ath_ddr_initial_config(278): (ddr2 init)
ath_sys_frequency: cpu 775 ddr 650 ahb 258
Tap values = (0x8, 0x8, 0x8, 0x8)
4 MB
Base:0x80000000, Top:0x80400000, Res logbuf:0xa03fb000, log_magic:0xffffffff kseg: 0xa0000000
Top of RAM usable for U-Boot at: 803fb000
Reserving 247k for U-Boot at: 803bc000
Reserving 192k for malloc() at: 8038c000
Reserving 44 Bytes for Board Info at: 8038bfd4
Reserving 36 Bytes for Global Data at: 8038bfb0
Reserving 128k for boot params() at: 8036bfb0
Stack Pointer at: 8036bf98

I suspect I fried the RAM, or the FLASH or both.
if it's just the flash chip, and I can replace it ( I can work with SOIC ), can it be revive?
If it's the RAM chip I guess I'll just keep it in the bin for now :slight_smile:

You've provided 0 information to assist you.

  • What did you do with the USB-to-Serial connection?
  • What procedure were you undertaking?
  • How is it bricked?
  • Can you see the U-Boot command prompt?

I highly doubt that, why do you beleive you destroyed your RAM!?!?

You provide no information to support that - it doesn't seem [hard] bricked to me.

YES - especially if you have a U-Boot command prompt.

U-boot decides there is only 4 MB of RAM. I think that Atheros chips probe for RAM rather than have it hard-coded in uboot. So if that's the case there's a RAM-related hardware failure. If it is hard coded in uboot maybe your uboot is corrupt.


If the U-Boot help reveals memtest or the like, that might provide more insight.

From a different device (ARM)

mtest   - simple RAM read/write test
(IPQ40xx) # mtest
Pattern FFFFFC9D  Writing...  Reading...

Not sure what it's doing, but might "poke" the SoC in a way to get the memory recognized.

It's a stupid mistake on my part out of being curious.
I had it running well untill I decided to probe what's going on in the background.
I tried to use a usb to serial that has a jumper option for 3.3v/5v , it was set for 5v.
I just realized just now that one pad of the four is probably supply output and not input.

I assume that it could be the RAM because it only sees 4MB out of 128, I saw a similar
output from a wiki post that shows the correct amount of memory.

I don't have access to u-boot, it's not accepting any input.

This is where I saw a similar output. Towards the bottom where it mentions Serial port outputs on OEM firmware[]=unifi&s[]=ac&s[]=lr

Oh my! If you connected to the device with 5 Volts, you may have fried something...

Did you post the entire output that you received in Post No. 1?

That's what I'm afraid of.

The output I posted on No.1 is the entire output of the device, it hangs after this

and no input is being accepted.

Hi guys/gals,

I'm reviving this thread in the hope of reviving my AP. I don't give up that easily :blush:
I've learnt a few things since last post.

Anyone care to share a dump of u-boot for AP AC LR?

I managed to read and write the NOR flash of my AP with the help of raspberrypi 3b.
This is what I've got from the serial port, with a custom u-boot test build:

U-Boot 2019.10-rc2-00001-gdf33f86468 (Aug 16 2019 - 19:53:07 -0400)

Qualcomm Atheros QCA9561 ver 1 rev 0
Model: AP152 Reference Board
DRAM:  13.8 GiB

It obviously is not correct, it is dead but definitely showing some sign of life.

I reflashed the corrupted original bin file I saved and I'm now back to original serial output that hangs.
I'm not inclined to buy another one just yet so I could extract u-boot to save this one, 'will wait and search for now.

This will probably not materialize, as I got me a used UAP-AC-PRO, hopefully it doesn't die before I bricked it again this thing can cook an egg :wink:

1 Like