• TI Thinks Resolved

DLPC150: booting problem

Prodigy 180 points

Replies: 17

Views: 311

Part Number: DLPC150

Hi,

We are developing a system using the DLP2010NIR. We designed our system based on the circuit diagrams provided for the evaluation module. Despite everything else seems to be correct, our DLPC150 chip cannot load the firmware successfully, i.e. we don't get the expected high to low transition on the HOST_IRQ pin, it remains high all the time.

Let me describe our current situation:

To make sure we programmed the flash IC correctly, we soldered it directly on the eval module, programmed it through the TI software, checked that it's working and transferred it to our circuit. DLPA2005 does produce all the correct voltages, PROJ_ON signal is high, DMD is connected, we have a 10k pull-up resistor on the HOST_IRQ pin, there seems to be some activity on the flash SPI lines so the chip is active. When we disconnect the DMD and power-up, DLPA2005 enters the standby mode: Voffset and Vbias = 4.5V, Vrst = 0.45V. When the DMD is connected we get the proper voltages, so the device can switch between Active1 and Standby modes.

Do you have any suggestions?

Thanks,

Kıvanç Hedili

  • Hello Kivanc,

    Welcome to the E2E forum and thank you for your interest in DLP technology.

    Is the firmware that you are loading on the board the base firmware or has it been customized? To rule out the firmware as the cause of the inability to boot, you may wish to flash the basic image and test the startup before any custom images are used.

    Regards,

    Austin

    --------------------------------------------------------------------------------------------------------

    If a post answers your question, please click the "Verify Answer" button.

    Have questions about DLP Pico Chipsets? Check out the DLP Pico Universal FAQ for quick answers: https://e2e.ti.com/support/dlp/f/94/t/946506 

  • In reply to Austin Snyder:

    Hi Austin,

    The firmware we are using is the basic version, we didn't change it in anyway. We downloaded it from the TI website (version 2.2) and we checked that it works on the evaluation module. We are pretty sure that the firmware is not the problem.

    Regards,

    Kıvanç Hedili

  • In reply to Kivanc Hedili:

    Hello Kivanc,

    We will get back to you  in first week of January.  This is a holiday period for our team.

    Thanks for your patience.

    regards,

    Vivek

    --------------------------------------------------------------------------------------------------------

    If a post answers your question, please click the "Verify Answer" button.

    Have questions about DLP Pico Chipsets? Check out the DLP Pico Universal FAQ for quick answers: https://e2e.ti.com/support/dlp/f/94/t/946506 

  • In reply to Vivek Thakur:

    Thank you Vivek.

    As an update: We probed the SPI lines with our logic analyzer, on both the evaluation module and our own board. In both cases we confirmed that the DLPC150 starts the communication by sending 0x0B followed by 4 bytes of all zeros. After that the flash IC sends the exact same bytes on the first 32 bytes of the firmware file. Once the first 32 bytes are sent, the SPI clock frequency increases and our logic analyzer cannot measure the data reliably. Considering that we programmed the flash IC by directly soldering it on the eval module and programmed it through the TI software and it behaves as expected for the first measurable 32 bytes, we believe that the problem is not with loading the firmware.

    I should also note that we followed the PCB guidelines on the datasheet. All differential pairs are of the same length and low speed data lines are length matched as well. The trace lengths are well below the maximum allowed values.    

  • In reply to Kivanc Hedili:

    Kivanc,

    What is the flash part you used for your replacement?

    Regards,

    Philippe Dollo

    --------------------------------------------------------------------------------------------------------

    If a post answers your question, please click the "Verify Answer" button.

    Have questions about DLP Pico Chipsets? Check out the DLP Pico Universal FAQ for quick answers: https://e2e.ti.com/support/dlp/f/94/t/946506 

  • In reply to Philippe Dollo:

    Philippe,

    We used Winbond W25Q16. Before desoldering the chip we checked that it was working on the eval module.

    Regards,

    Kıvanç Hedili

  • In reply to Kivanc Hedili:

    Kivanc,

    If you were able to confirm the flash device is working on the EVM but not on your board design, it does seem more likely that perhaps there is something wrong with the board design you are using.

    You said you based it on the existing DLP2010 NIR EVM board. Are you able to share your iteration of the design?

    Regards,

    Philippe Dollo

    --------------------------------------------------------------------------------------------------------

    If a post answers your question, please click the "Verify Answer" button.

    Have questions about DLP Pico Chipsets? Check out the DLP Pico Universal FAQ for quick answers: https://e2e.ti.com/support/dlp/f/94/t/946506 

  • In reply to Philippe Dollo:

    Thank you Philippe, 

    Sure, I can share our design but how do I do that on the forum? My e-mail is khedili@gmail.com, maybe we can communicate over there.

    Regards,

    Kıvanç Hedili

  • In reply to Kivanc Hedili:

    Kivanc,

    A TI applications engineer will reach out to you shortly with a method for sharing your schematic.

    Regards,

    Austin

    --------------------------------------------------------------------------------------------------------

    If a post answers your question, please click the "Verify Answer" button.

    Have questions about DLP Pico Chipsets? Check out the DLP Pico Universal FAQ for quick answers: https://e2e.ti.com/support/dlp/f/94/t/946506 

  • In reply to Austin Snyder:

    Thank you Austin.

    Looking through our design I just realized that we are using a different sub-LVDS pin configuration than what is used on the eval module. According to eval module schematics "Option3" is used for the differential pairs, whereas we are using "Option1". Could that be the problem? I couldn't find any information on how to change this configuration so I am guessing maybe we need a different firmware.

    Regards,

    Kıvanç Hedili