OWON XSA-1015TG - Spectrum Analyzer - Firmware Update - Boot Failure

[shortfoxyman]

Hello.
I have an OWON XSA-1015 TG - Spectrum Analyzer with built-in Tracking Generator.
Is a great low cost chinese spectrum analyzer and with superb specifications.
A great VFM deal for amateur and professional repairs, tests ..etc.
In past I had the same faulty "boot" issue as many of spectrum analyzer owners had.
((( When Power On, fun working, - One time Flashed Screen, and remains blank.... for ever.)))
I know in some of these cases Owon had send to owners that are in 3 years warranty a working mainboard to replace, and that is a great support from them....but is this a permanent solution? , or....it'll be as long as it goes?? What about the owners that are out of warranty?
My question is if someone had another solution to solve the faulty "boot" problem, or....have the firmware,
and managed to re-wrote it in NAND memory, make the analyzer working again.
Owon company...(for some unexplained reason) did not have the spectrum analyzers firmware file
or any firmware update on their website, but for the other test equipment ,they have the files available.
(That is weird).
I don't understand, why all these years did not offer the firmware or...firmware updates, and a solution with instructions, so the owners that they are in... or out of warranty repair their faulty "boot", on their analyzers?
Many other chinese test equipment companies they offer these services free on their websites.
So strange that Owon left the spectrum analyzer owners with the "hands up"...without any firmware support and service solutions.

[Selectech]

I have one of these.

As you men mentioned I had received a new core from Owon, but it's now in the same situation, so I have a non-functional unit, again.

Probably only needs a reprogram, but no file to be found.

[shortfoxyman]

Oh. that is so dissapointing.
How long (approx. hours) working fine until it appears the same problem again?
What you intend to do with this??
In fact, is someone that he find the really reason of the faulty boot problem??
is the firmware?, or a hardware problem causing this?
I don't have knowledge of these kind of electronics....
I hope someone who is familiar and have knowledge in digital electronics, find the problem
and give a permanent solution, as to read and copy the firmware from a working device, and trying to reprogram a faulty device.
As i discover on interenet blogs, there are many owners with the same problem years ago,
and except the "mainboard replace support" from owon for some customers, no other
solution comes up all these years.

[shortfoxyman]

Owon XSA-1015 TG Power On - Start Failure

https://www.youtube.com/watch?v=UPb1cUDdXnM&pp=ygUNT3dvbiB4c2EgMTAxNQ%3D%3D

[R30_0K]

Hi,
I’m facing the exact same issue with my Owon XSA1015-TG as described earlier in this thread. It seems like a firmware corruption or a NAND flash problem.
Does anyone happen to have a working NAND dump for the XSA1015-TG? I would really appreciate it if someone could share the dump file so I can try to recover my unit.
Thank you in advance!

[shortfoxyman]

That's the Main problem of these analyzers for years....many owners have the same issue and they waiting
for a solution. Also,  Owon didn't respond to the emails for this issue.
I have the same Analyzer with this problem. Seems it is a NAND flash firmware problem and OWON as i know didn't release yet any firmware and the way to re-write it to the NAND flash memory through USB Port with a USB Flash stick.
That really Sucks..!!!! I am waiting 3 years now for a solution and i didn't find anything yet.


PLEASE....to anyone here who has Programming Knowledge, to figure out an easy solution
so to re-write the firmware to the internal NAND Memory through the USB Port, and wake up again the
analyzer.

[alpher]

That's my video a couple of posts above, Owon send me a whole new replacement board set and that fixed the problem so far.
I kept the original boards as they din't wanted it back, here are some pics:
Analog section



[ Specified attachment is not available ]

[ Specified attachment is not available ]

Digital board:

[ Specified attachment is not available ]


The last pis shows what I suppose is the culprit a flash chip most likely holding the firmware, I can dump it if you want.


[ Specified attachment is not available ]

[alpher]

Somehow pics aren't showing, here it is analog top:

[alpher]

analog bottom:

[alpher]

Digital top:

[alpher]

Digital bottom:


I suspect that here lays the culprit, the flash chip 29F4G08 etc...
Most likely holds the firmware, I can dump it if you want?

[tautech]

Somehow pics aren't showing, here it is analog top:

Check the image sizes before you post them as there are forum limits for each attachment and each post.

[coromonadalix]

That's the Main problem of these analyzers for years....many owners have the same issue and they waiting
for a solution. Also,  Owon didn't respond to the emails for this issue.
I have the same Analyzer with this problem. Seems it is a NAND flash firmware problem and OWON as i know didn't release yet any firmware and the way to re-write it to the NAND flash memory through USB Port with a USB Flash stick.
That really Sucks..!!!! I am waiting 3 years now for a solution and i didn't find anything yet.


PLEASE....to anyone here who has Programming Knowledge, to figure out an easy solution
so to re-write the firmware to the internal NAND Memory through the USB Port, and wake up again the
analyzer.

you dont have many choices  if you are still stuck,  grab another and do dumps,  flash dumps, eeprom dumps  etc ...   this is the hard thruth   if OWON does not help, you have to do thing by yourself  or others may chime in

BUT you have to wait and not pressure people ...


Im stuck with GWinstek  MFG 2xxx generator myself, GW Instek  support pffffffff,     and i have to buy another, hoping that i may be able to do fw backup,  i have discovered the memory chip is no longer made, i did not found any good ones, all fakes ... grab a programmer, buy sockets ....  need to decipher the jtag port  ...

[shortfoxyman]

I am  Sorry, but unfortunatelly, i don't have so deep programming knowledge, and i am not familiar with damping...
As i know the "black screen" (failure booting problem) perhaps from NAND Flash Memory (29F4G08ABADA) is the only common problem for OWON XSA-1015TG.
I suppose that some owners that they didn't have any manufacturer support, or other solution to do,
just have the device at the edge of the bench for years...waiting and hope for a future solution, or... they have it on sale as is.
I really  don't  want to pressing others, but I hope someone who has the knowledge to damp and share the files, inclunding and the suitable step by step full instructions to do the Analyzer Working again.

[R30_0K]

I'm uploading images of my motherboard.
While my NAND Flash is the same 29F4G08ABADA as alpher's, the FPGA seems to be different. It appears there are several hardware revisions, which might be why OWON does not release the firmware publicly.
By the way, I believe the NAND Flash data can be dumped directly from the 29F4G08ABADA using programmers like the XGECU T48 or T56. Although the data on my NAND is likely corrupted, I have a T48 and will try to read it when I have some time.
I’m really hoping someone can upload a working firmware image to save our spectrum analyzers...

[shortfoxyman]

Here is the Motherboard of mine OWON XSA-1015TG.
The problem is the usually malfunction that have many owners of this device....
"When Power On, fun working, - One time Flashed light white Screen, and screen remains blank.... for ever."

[R30_0K]

Hi everyone,

Follow-up to my previous post — I have successfully dumped the NAND flash memory from my Owon XSA1015-TG.

I have linked the dump file to this post.

For reference, the mainboard on my unit uses a Xilinx Artix-7 FPGA.

I would really appreciate it if the community could help analyze this dump. Also, if anyone here has a working (or even non-working) XSA1015 unit and can dump their NAND, comparing our dumps could be extremely helpful in pinpointing corrupted files, bad blocks, or potential firmware differences across hardware revisions.

Here are a few details for anyone interested in taking a look:

Dumped File: https://drive.google.com/file/d/1BDekwJ1ojJu5u9Muc08Ol_9Kmd54DDR_/view?usp=sharing

Mainboard Hardware: Xilinx Artix-7 FPGA

Current State of my Unit: Upon powering on, the screen flashes white briefly, and then stays completely black with the backlight remaining on.

Goal: Reverse engineering the partition layout / extracting clean firmware files so we can restore units stuck in bootloops or with corrupted flash memory.

Any analysis, insight, or dump comparisons would be hugely appreciated. Let’s see if we can get a working recovery solution for these units!

Thanks.

[rg58]

This will be a tough one, for several reasons.

As seen on the pictures, the main controller (SoC) is an OMAP-L138 E ZWT from Texas Instruments.

Dual Core SoC: ARM926EJ-S (ARM9 RISC CPU) + C674x (DSP CPU) + loads of peripherals

E = Silicon Revision 2.3 (latest)
ZWT = 361-Pin Plastic BGA, with Pb-free Soldered Balls, 0.8-mm Ball Pitch, 16,00 mm x 16,00 mm

I've done a longer search in your dump. Here's what I found so far:

TIPA YSX - Texas Instruments Application Image Script, generated by AISgen tool

UBIFS - Unsorted Block Image File System

JFFS2 - Journalling Flash File System v2

U-Boot 2010.12 (Oct 19 2018 - 15:50:09)

QNX4 (this should be the OS, although I found mentions of different Windows versions)

Linux-2.6.32-rc6 - kernel version

I wasn't able to extract anything useful from the image, besides some random garbage and PNG files. There are at least 12 GZIP compressed streams inside, but all are mangled, so not possible to extract. This could be because of the way the UBI file system writes to the flash and scattering/distributing/fragmenting the data. I've tried really hard, without success. Also I could not identify the beginning of any file system yet, although I see 3708 + 531 UBI FS blocks. Need to dive into that FS some more...

Now, first thing to do would be to locate the UART port, then we can observe the boot process, AND, if we are lucky, are presented with some boot options (try hitting SPACE or ESC key during boot).

There's a RS232 connector on the digital board:

XP8 on R30_0K's - board revision 2.2 (2018-09-10)
XP3 on alpher's - board revision 1.0 (2017-06-28)
XP3? on shortfoxyman's - board revision 1.0 (2017-06-28) - no bottom pic provided

If you don't know the pinout, check which pin is GND, the other one would be data. Connect them to your scope, one pin on each channel. Turn on the device and watch for the data bursts. This one is the TXD line, the other one RXD. Note the amplitude of the signal, it goes up to +/-15V. You will need a RS232<->USB adapter, then connect TXD of the device to your RXD input and vice versa. Set your serial COM port to 115200 8N1. Not sure, but maybe the USB port provided there could be a virtual COM port already, so you might want to connect a USB cable there and see what comes up in your device manager in Windows.

You might also want to try via web browser:

192.168.1.9 / 255.255.255.0
or
192.168.2.5 / 255.255.255.0

There could also be a TFTP server running, standard port is 69 UDP. Try a READ request, using any random TFTP client. You could do a port scan as well to find other running services like FTP (21), SSH (22) or Telnet (23).

Plenty of things can be tried.

Side notes:

1) I don't own one of these myself (yet!)

2) I am not familiar with TI devices (is anyone here?)

It MIGHT not even be a firmware problem. As OWON previously replaced whole boards without requesting them back for repair, this could also indicate cold (F)BGA solder joints, for instance. The effort of repair would be too high in this case, but a FLASH rewrite should be easily doable for a manufacturer. They know their own programming ports and know how to access them. This should be a matter of 10-20min per unit, worth repairing actually.

Anyway, I'm curious how far we can get. With the combined, unparalleled power of our global experts on here

GANBATTE NE!