Post Go back to editing

EVAL-SDP-CH1Z, potentially bricked.

Category: Hardware
Product Number: EVAL-SDP-CH1Z being used with EVAL-AD4020FMCZ
Software Version: Latest SDP driver version available from AD EVAL-SDP-CH1Z product page, 24-07-2023

Hello SDP support.

I am was using the SDP-H1 board with the EVAL-AD4020FMCZ daughter board and was doing what it was designed to do.

The next day I powered up my test setup in the usual way, everything was preconnected ready to go.

The development laptop did not recognize the SDP-H1, even after several attempts following the instructions in UG-502 (just the SDP-H1 & EVAL-AD4020FMCZ were connected to the dev laptop). SDP-H1 said “No” – laptop didn’t register a USB device plugged in.

I changed laptops, reinstalled SW and changed USB leads – still No-Go with the SDP-H1.

So I started to debugged the SDP-H1, the schematic in the current UG-502 is different with regards to my version of SDP-H1 (ver1.4 2017, Serial No 21148415), it made life hard with respects to debugging some parts of my SDP board.

The SW for the AD4020FMCZ continually asks me to “rescan” because it can’t see the SDP-H1 as a registered USB device.

Results of my debugging exercise :-

LEDS lit

BF_POWER, Yes

SYS_POWER, Yes

FMC_POWER_GD, No

FPGA_DONE, No

STATUS, No

LED0, No

LED1, No

LED2, No

CLOCKS Present and correct levels.

24MHz, Yes

100MHz, Yes

 

Power Supplies and enable signals around the SDP-H1

 

+3.3V_SWITCHED, No – No enable signal from U29 (OR gate of VIN_SDP_CON with BOOT_COMPLETE, neither of which were Logic Hi)

VCC_SWITCHED, No – uses same enable signal as above

+3.3V, Yes

+2.5V, Yes

1.2V_BFIN, Yes

VBUS present when plugged into laptop.

Not_POR does what it is supposed to do.

BOOT_COMPLETE = 0V, BF originated signal via an external latch U19.

FMC_POWER_EN = 0V, FPGA originated signal.

3V3_FPGA_VCC0, Yes correct voltage within tolerance.

2V5_FPGA, Yes correct voltage within tolerance.

VADJ_EN = 0V, FPGA originated signal.

VADJ_FPGA = 2.49V, enabled via DDR2_PWR_GD

1V8_DDR2_VDDQ, Yes correct voltage within tolerance.

0V9_DDR2_VTT, Yes correct voltage within tolerance.

0V9_DD2_VREF, Yes correct voltage within tolerance.

1V23_FPGA_CORE, Yes correct voltage within tolerance.

+12_VIN, Yes of course Wink

FPGA_2V5_PWR_GD = ???, probably couldn’t find a component with known RefDes to probe (not all components have and RefDes on silkscreen and no component layout guide in UG-502)

DDR2_PWR_GD = 4.54V

+12_VIN_SWITCHED, Yes – the schematic for this in UG-502 differs from the components used for it on my SDP-H1.

+5V, Yes correct voltage within tolerance.

3V3_FMC, No Frowning2

FMC_VADJ_PWR_GD=No

VADJ_FMC=0.00V

VADJ_INTVCC=5.0V

VADJ_2V5_SET=0.457V

VADJ_EN=0.0V, FPGA originated signal.

FPGA_CORE_PWR_GD=3.0V

The FMC power supplies do not seem to get enabled, because the FPGA never gets programmed.

BOOT Memory U701 activity.

BOOTMODE 0x3, Yes.

BF is trying to boot from SPI memory.

Activity on Not_SPISEL, SCK, MOSI and MISO.

Pulls ups working on the Not_SPISEL and MISO lines.

 

Following the schematic diagram, I have gleaned an understanding of how this SDP boots.

No USB detect, therefore no Blackfin comms and thus the FPGA does not get programmed, and hence no FMC power supplies enabled.

 

FMC_POWER_EN (from FPGA)

No VADJ_FMC as it does not get Enabled by FMC_POWER_EN

Although +12_VIN_SWITCHED is up and it is according to the schematic in UG-502 enabled by the same signal FMC_POWER_EN

If VADJ_FMC is powered up, that would generate FMC_VADJ_PWR_GD

FMC_VADJ_PWR_GD would enable 3V3_FMC

If 3V3_FMC is powered up, that light the FMC_ PWR_GD LED.

 

I believe the BLACKFIN F/W has been bricked somehow, apparently the F/W gets updated every time the SDP is used (info from SDP WiKI).

 

How do I get this board to work? I believe I may need to do a recovery or re-FLASH the F/W via the BLACKFIN JTAG port.

Best Regards

Daniel

  • Hi Daniel,

    Thank you for the debugging info. It does sound like your board is bricked. 

    I'll look into your debug information in more detail to see if the board can be unbricked. There may be alternative solutions depending on whether you are interested purely in evaluation, or if you intend to prototype.

    A review of the reference material, including schematics, is on my to-do list for the weeks ahead - apologies for the complications you faced with the mismatched schematic.

    Regards,

    David

  • Hello David

    This board was the last only one available in the UK back in April, bought from the distributor "Anglia Live" via an intermediary company because the company I work for doesn't buy directly from Anglia (not on our approved distributors list).

    Not having this board has seriously curtailed my development process, I am using it for prototyping a instrumentation card for a new product.

    So I really need it back up and running ASAP.

    An aside question, I'd like to develop S/W on the SDP-H1 (as a basis for another product) and I'm also interested in using ACE with SDP-H1 and AD4020FMCZ.

    ACE doesn't seem to support SDP-H1 with a AD4020FMCZ connected, unless I'm missing something.

    My SDP-H1 is a version 1.4, Serial No 21148415, Date code 2116, Lot ID 50007193 (info for label on board and the shipping box).

    Best Regards

    Danny

  • Hi Danny,

    If your board is not showing up at all in Windows Device Manager then I'm afraid the bricking is permanent.

    The component shortage affected the availability of components for the SDP-H1. We have had a minor boost in stock very recently but I can't say if any of those made it to distributors.

    There are other options for interfacing with the AD4020. The SDP-H1 is a "gets you up and running in less than 20 minutes" type of solution, but is not suitable for prototyping. Some boards use ADI ACE software and there is some scope for automating and looping together some of the functions that feature in the demo. For the AD4020, the software behind the scenes is LabVIEW - you would not have the automation options in that.

    The alternative is more flexible but there is a learning curve. You could use another FPGA board and the IIO and NoOS collateral to develop a prototype.
    To help with any questions about the AD4020 I'll tag  , he may have some other options/advice for you.

    Regards,

    David

  • Hello David

    That screenshot is old too, its not from the current version of software. It is also used on the old of date User Guide for the SDP-H1.

    I did not see that when I first plugged in my SDP-H1, I did not have a development system all I had was an SDP driver.

    It "Beggars belief" that ADI can supply an expensive hardware dev platform and have no way for a customer to recover it from a "bricked mode". I believe I saw a recovery option for another SDP platform in the ACE software suite, (which is installed on my personal laptop at home).

    It also "Beggars belief" that the firmware for this platform gets loaded every time the board is used, "There's many a slip 'twixt the cup and the lip". I believe it was bricked when the USB lead got "tweaked" whilst the SDP-H1 was being powered up.

    I have work with a ADI dev platforms before (GSM/GPRS chipset) back in the early 2000's and there was always a way to recover them even if was via the JTAG port.

    I just need the supplied functionality of the SDP-H1 to monitor the performance of my "front-end electronics" for an fibre optically isolated voltage/current measurement system I am developing. I have a developed a FMC-Optical Fibre-FMC interface that sits between the SDP-H1 and the AD4020FMCZ, the FRU contents have been copied and placed on the SDP-H1 side of this interface in order for the SDP-H1 to still think it is directly coupled to the AD4020FMCZ which actually sits on the other side of the fibre optics, the interface basically takes the SPI bus into optical domain and back out again.

    I am in urgent need of an SDP-H1 board in order to continue my dev work, the distributors in the UK are also showing lead time that end in May 2024.

    Regards, one very upset ADI customer.

  • Hi Danny,

    Thank you for patience on this. I'm consulting our software/firmware team to figure out other options for SPI-over-FMC to get you back up and running.

    Can you please comment on whether the IIO and No-OS drivers that David recommended could meet your needs? I understand that these would require a different FPGA kit than the SDP-H1, and apologies for the inconvenience.

    Do you need a GUI with time- and frequency-domain plots and analysis to do your early evaluation?

    Thanks,

    Tyler

  • Hello Tyler

    I am predominantly a hardware engineer that sometimes "once in a Blue moon" turns his head & hands towards S/W and VHDL, they haven't been my weapon of choice i quite a while. Not really a solution, more of a re-learning curve for me.

    The SDP-H1 and its GUI for the AD40xxFMCZ boards gave me all I required to develop the electronics for the instrumentation card that will eventually live in a suite of telecoms equipment, telecoms equipment that provides a backbone for the internet.

    I have mange to procure a vintage ADI 3V3-5V JTAG pod and PCI card for the Blackfin processor, I am in the throws of tracking down a suitable version of the S/W dev system. I am hoping to find another working SDP-H1 either new/used, that I can either use directly for my dev work and/or aid in the recovery of my SDP-H1 board as I don't believe it is "permanently bricked".

    Regards

    Danny 

  • Hello David and Tyler

    I would like to use this technique in the link below to try and recover my SDP-H1.

    ez.analog.com/.../faq-how-do-i-restore-the-original-sdp-image-to-flash-on-my-sdp-board

    I don't have a "BF527 EZ-KIT Lite" to hand and I'm not too sure if its still available. I have 3 x EZ-ICE-JTAG 3V/5V pods (2 labelled AD and 1 labelled White Mountain), I also have their matching control cards, two PCI and one 8-bit ISA.

    All are a bit long in the tooth but I haven't been all to track down the SW dev tools they came with, we have some legacy PC's to support our products through their predicted 25 year life span - which run 24/7/52.

    Best Regards

    Danny

  • Hello SDP support

    I have manage to secure a New "Old Stock" ADZS-BF527-EZLITE to help in the debug process and hopefully "unbrick" this SDP-H1.

    Some more helpful interaction with ADI support would be appreciated. Possibly starting with the firmware image file to try and re-flash this SDP-H1 would be a good start or a SDP-H1 recovery process in the ACE application software.

    I'd like to tag someone in ACE Software, I'm not sure I can cross post between EngineerZone forums.

    Regards

    Danny

  • Hi Danny,

    Our software team has agreed to ship you a replacement SDP-H1 board to support your immediate needs. They will coordinate shipping details with you offline.

       can provide more details when he's in next week.

    Thanks for your patience,

    Tyler

  • Hello Tyler & David

    Thank you very much for the timely replacement SDP-H1, I wasn’t expecting that at all.

    This will be very helpful in getting my development work back on track.

    Best Regards

    Danny