Author Topic: Agilent N9320B stuck at boot: accessed boot prompt and need advice  (Read 1679 times)

0 Members and 1 Guest are viewing this topic.

Offline Renate

  • Regular Contributor
  • *
  • Posts: 194
  • Country: us
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #25 on: November 27, 2019, 07:37:54 pm »
I am pretty sure the boot loader (which I believe is sitting on CPU chip)...
Well, what is the chip?

ROM bootloaders are all different and they also have boot options.
Some always pause and check if there is some USB thing out there (like OMAP 36xx).
Others only appear if they have reason to believe that things have failed (like iMX6).
If things were working well enough you might never have seen a lower-level boot attempt.
It could be on a serial port, on USB, on SPI, on I2C ROM, maybe network (I don't think so, but who knows).
Also, it would be more usual to expect a USB peripheral port, rather than a host port.

There's also JTAG. (You've got a lot of test points there.)
Or it could be SD card. (You've got a lot of test points there. I've added SD cards to things on test points.)
How does manufacturing load an image, huh?

Still, SRAM is worth a shot.
 

Offline cgz2001

  • Newbie
  • Posts: 2
  • Country: us
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #26 on: November 28, 2019, 12:27:50 pm »
One comment - I have an Agilent DSOX2002A scope, first put into production in 2011 which has a failure mode in that the boot-up sequence does not finish, and that it endlessly loops on the Agilent splash screen.  See the forum below for details:

https://www.eevblog.com/forum/testgear/agilent-dsox2024-won_t-boot/?all

Keysight Daniel has indicated this repair is free to resolve, and I have PM him to see if this offer still stands.  Anyways, the community seems to believe this is a bootloader memory corruption issue, and I am starting to wonder from this post if it's widespread with equipment first into production during that time frame.

Keysight seems to have recognized the problem and implemented a feature which allows the ability to reflash the bootloader.  I wonder if the later firmware revisions offer the same for you on your instrument.  There are forum posts from the community that indicate a hardware fix has occured (new boards appear to be installed once repaired by keysight) as well as firmware updates to address the problem.

I am also intrigued by the information from another user indicating bootloader code resides in two locations, and disabling one location allows the other to get the machine up and running.  Can anybody provide detailed instructions on this procedure?
 

Offline analogRF

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: ca
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #27 on: November 28, 2019, 01:07:02 pm »
I am pretty sure the boot loader (which I believe is sitting on CPU chip)...
Well, what is the chip?

ROM bootloaders are all different and they also have boot options.
Some always pause and check if there is some USB thing out there (like OMAP 36xx).
Others only appear if they have reason to believe that things have failed (like iMX6).
If things were working well enough you might never have seen a lower-level boot attempt.
It could be on a serial port, on USB, on SPI, on I2C ROM, maybe network (I don't think so, but who knows).
Also, it would be more usual to expect a USB peripheral port, rather than a host port.

There's also JTAG. (You've got a lot of test points there.)
Or it could be SD card. (You've got a lot of test points there. I've added SD cards to things on test points.)
How does manufacturing load an image, huh?

Still, SRAM is worth a shot.

the cpu and the flash names were shown in my first post (in the boot log)

network booting is always there, and shows up along with tffs booting. I cannot attempt it anymore because the instrument
has stopped giving me the option of stopping the auto boot

let me receive the SRAM and see what happens but if that fails, and I can use your trick to get
my network boot option back, my question is what file(s) can I feed into it for booting? I mean the only thing I have is
the firmware files downloaded from keysight. I don't have the .bin image of the flash for example...there is a VxWorks (no extension)
file in that firmware which I am hoping will be the boot file....
 

Offline analogRF

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: ca
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #28 on: November 28, 2019, 01:15:33 pm »
One comment - I have an Agilent DSOX2002A scope, first put into production in 2011 which has a failure mode in that the boot-up sequence does not finish, and that it endlessly loops on the Agilent splash screen.  See the forum below for details:

https://www.eevblog.com/forum/testgear/agilent-dsox2024-won_t-boot/?all

Keysight Daniel has indicated this repair is free to resolve, and I have PM him to see if this offer still stands.  Anyways, the community seems to believe this is a bootloader memory corruption issue, and I am starting to wonder from this post if it's widespread with equipment first into production during that time frame.

Keysight seems to have recognized the problem and implemented a feature which allows the ability to reflash the bootloader.  I wonder if the later firmware revisions offer the same for you on your instrument.  There are forum posts from the community that indicate a hardware fix has occured (new boards appear to be installed once repaired by keysight) as well as firmware updates to address the problem.

I am also intrigued by the information from another user indicating bootloader code resides in two locations, and disabling one location allows the other to get the machine up and running.  Can anybody provide detailed instructions on this procedure?

my N9320B is not in a boot loop, it stops because of some kind of memory failure (either flash corrupted or the SRAM is bad)
I dont think this is a wide spread issue for these instruments. These units were designed by a totally separate group in Agilent
I mean their oscilloscope group in Colorado is separate as I understand and also these units use a completely different architecture, different cpu and flash etc....than DSOX

I have asked Keysight tech support for help (as Daniel also told me to do so in a private message) and they never answered me back, not even saying "sorry you should only send the unit to us to fix it for $5000"
« Last Edit: November 28, 2019, 02:21:03 pm by analogRF »
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3105
  • Country: ca
  • Living the Dream
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #29 on: November 28, 2019, 05:26:03 pm »
The scope issue isn't caused by a bootloader issue, it is NAND memory corruption in the main OS memory. The fix was to make two copies of firmware and to check it extensively at boot and correct any errors.
I believe it has already been stated the N9320B uses linear (NOR) flash so any failure would be different.
The N9320B would likely be a lot easier to troubleshoot with the boot log from a good unit. I have no experience with them and don't own one or I'd get a log.
VE7FM
 
The following users thanked this post: cgz2001

Offline Dacke

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #30 on: November 28, 2019, 06:04:18 pm »
With regards to stopping the autoboot... Just by chance,  have you tried something like this - http://support.motioneng.com/Hardware/eXMP/VxWorks/exmp_qs_vx_05.htm  -  Apologies if I missed it and you already tried.

Maybe it would work for the version of VxWorks on the SA.  I noticed there is an enter key on the front of the analyzer,  I would just try pushing it repeatedly the second you power it up.

 

Offline analogRF

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: ca
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #31 on: November 28, 2019, 06:22:02 pm »
With regards to stopping the autoboot... Just by chance,  have you tried something like this - http://support.motioneng.com/Hardware/eXMP/VxWorks/exmp_qs_vx_05.htm  -  Apologies if I missed it and you already tried.

Maybe it would work for the version of VxWorks on the SA.  I noticed there is an enter key on the front of the analyzer,  I would just try pushing it repeatedly the second you power it up.

yes I have seen that site...I did have the option "Press any key to stop auto booting..." and I could stop it
but the problem is later on it disappeared and now it goes to auto boot right away, tried anything to stop it .... what Renate said
by intentionally making the flash not responding (correctly) i might be able to stop the bootloader and get a boot prompt .
still I am puzzled by why this happened. The only thing related is the boot flag which I never changed it...

Steve is right, only if I had a boot log of a working unit  :'(
 

Offline Renate

  • Regular Contributor
  • *
  • Posts: 194
  • Country: us
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #32 on: November 29, 2019, 12:21:30 am »
The cpu and the flash names were shown in my first post.
Well, okay.
It's a Sharp LH7A404, now taken over by NXP.
This is the manual: http://www.keil.com/dd/docs/datashts/philips/lh7a404_um.pdf
I think your best bet is to look for boot jumpers and see if you can get XMODEM on UART2.
Still, it's going to be an uphill battle.
 

Offline analogRF

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: ca
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #33 on: December 13, 2019, 05:19:04 am »
finally the memory chips arrived. I replaced them carefully and nothing...no change whatsoever...same symptoms and same boot log... |O :-// :-BROKE

I also tried to reflow the ASIC that is near the CPU... not 100% sure if it actually reflowed but i think it did. anyways, no change...still stuck at the same memory allocation errors...now I need to force it to boot from something else and i have lost the option of stopping autoboot...
 

Offline analogRF

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: ca
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #34 on: December 13, 2019, 04:03:02 pm »
the flash eeprom has a hardware reset pin: https://www.cypress.com/file/219926/download
see page 15 and 35 for example

can I try to pull it LOW and then reset the processor and see if it gives me the boot options back again?
or maybe it comes up with USB boot option, too as Renate had experienced in another instrument
 

Offline analogRF

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: ca
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #35 on: December 14, 2019, 06:30:46 pm »
well, after spending more than an hour trying to find a spot that is connected to the RESET pin of the flash and soldering a wire to it (a tiny <0.5mm via pad) it turns out that pin is also connected to the reset pin of the CPU and when I pull it to ground the whole system goes to reset and does not start at all....so that was a waste of time....I don't want to pull any of the CE or OE pins of the flash to Vcc because it might damage the
chip that is controlling them...so i don't know how to prevent the system from booting? pulling an address line to ground maybe?! :-//
 

Offline YetAnotherTechie

  • Contributor
  • Posts: 44
  • Country: 00
Re: Agilent N9320B stuck at boot: accessed boot prompt and need advice
« Reply #36 on: December 15, 2019, 11:47:59 pm »
Could you post the new boot log that doesn't let you stop it?
What's stopping you from cutting that trace or lifting the pin?
There is a risk of breaking the pin, so i recommend you desolder the chip and make a backup.
Don't pull address or data lines to ground, there might be writes going on anyway and you'll corrupt the flash even more.

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf