• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 1274 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

What is the correct pinout for RM48L952 ZWT package?

Jason Work
Genius 10830 points
Other Parts Discussed in Thread: RM48L952, RM48L950, HALCOGEN

Why do the datasheet terminal functions description for the RM48L952 ZWT package pinout not match what is listed in the user's guide for the TMDXRM48HDK Hercules Development Kit?  For example, the EMIF data bit 0 is pin K15 according to the datasheet, but it is pin K16 according to the users's guide.  Which is correct?

Reference:
Datasheet RM48L952 16- and 32-Bit RISC Flash Microcontroller
Find, "EMIF_DATA[0]"
RM48 Hercules Development Kit (HDK) User's Guide
Find, "EMIFDATA[0]"

customer said:

We use the EMIF on the Hercules dev board to connect to our FPGA. The communication link works fine on Rev D of the HDK, but it isn't working on Rev E of the HDK. The only difference we can see is that Rev E comes with an L952 processor where as the Rev D comes with a L950 processor. Is there something we need to do to handle that change from one processor to another? Or is there some other change on Rev E of the HDK board that is impacting us. We can confirm that the FPGA is still working properly, but the link between the two isn't when the HDK is a Rev E board.

Also, I think I found an error in your documentation when I was looking into this. SPNU508B lists on page 15, table 2-7, that EMIFDATA[0] [1] [2] and [3] are connected to pins K16, L16, M16, and N16 on the Hercules BGA package. This does not correlate with other documents that indicate they are K15, L15, M15, and N15. The ?16 pins are shown as no connects on SPNS174 figure 2-4 (ZWT package pinout.

Please look into the rev E issue for us.

What are the differences between Rev D and Rev E of the TMDXRM48HDK?  The changes listed for both Rev D and Rev E are identical.  If the same change was made twice, wouldn't they just be the same (no change)?

Reference:
RM48 Hercules Development Kit (HDK) Board Design Files
spnr031\rm48\schematics\rm48_hdk_schematics_reve.pdf
Find, "Changes on RevE"

  • 0
    TI__Expert 4225 points

    Jason,

    i assigned this thread to HDK owner.

    He will get back to you.

  • 0
    TI__Guru**** 196996 points

    Hi Jason,

    On revD HDK, the footprint of Q2 for nTRST is pin-reversed. The 2MCU_nTRST signal  is connected to pin3 of Q2.

    For your product using revC silicon, Q2 is not required.

    regards,

    QJ

     

  • 0 in reply to QJ Wang
    TI__Genius 10830 points

    QJ Wang said:

    On revD HDK, the footprint of Q2 for nTRST is pin-reversed.

    On what revisions of the TMDXRM48HDK Hercules Development Kit is this connection correct?

    QJ Wang said:

    For your product using revC silicon, Q2 is not required.

    Please explain in more detail.  When you refer to revC silicon, are you referring to the revision of the RM48L952 microcontroller itself or the TMDXRM48HDK Hercules Development Kit?

    Jason Work said:

    ...the EMIF data bit 0 is pin K15 according to the datasheet, but it is pin K16 according to the users's guide.  Which is correct?

    I believe this question was overlooked.  Would you please address this?

    customer said:

    The [EMIF] communication link works fine on Rev D of the HDK, but it isn't working on Rev E of the HDK.

    Would you provide some suggestions as to why this might be the case?

     
  • 0 in reply to Jason Work
    TI__Guru**** 196996 points

    Hi Jason,

    Attached please find the SDRAM test case. Please try it on your board.

    Regards,

    QJ4274.RM48 HDK SDRAM Test.zip

  • 0 in reply to Jason Work
    TI__Genius 10830 points

    Jason Work said:

    On revD HDK, the footprint of Q2 for nTRST is pin-reversed.

    On what revisions of the TMDXRM48HDK Hercules Development Kit is this connection correct?

    [/quote]

    I understand now that on the TMDXRM48HDK Rev D, a hand soldered fix was needed to for reference designator Q2.  On Rev E this correction for Q2 was made on the schematic.  The customer has observed that moving from Rev D to Rev E the RM48l952 was used in place of the RM48L950.  What else has changed?  What is the complete list of changes betewen Rev D and Rev E?  We need some help to determine why the External Memory Interface (EMIF) works on Rev D but does not work on Rev E of the TMDXRM48HDK.

  • 0 in reply to Jason Work
    Prodigy 50 points

    We ran the test project here that exercises the EMIF. It works fine on Rev D boards, but not on Rev E board. On rev E boards it ends up showing the same word value written to the entire RAM contents instead showing the correct pattern only for the area that was written by the program. Please run the test program yourself on your own set of boards and confirm that there is a problem with the hardware.

    Jeff

  • 0 in reply to Jeff Fieldhouse
    Prodigy 50 points

    I may have spoke too soon. The SDRAM enable jumper was not in the correct position when we did the test. The sample project seems to be okay with rev E boards now that that is corrected. We will keep looking to see what may be causing our issues.

  • 0 in reply to Jeff Fieldhouse
    Prodigy 50 points

    We found this paragraph in the user guide:

    • Control for other EMIF Signals:
    Bit 31 of the system module control register GPREG1 at address 0xFFFFFFA0 is used to gate off the
    EMIF module outputs: EMIF_ADDR[0], EMIF_ADDR[1], EMIF_ADDR[6], EMIF_ADDR[7], EMIF_ADDR[8],
    EMIF_BA[1], EMIF_nCS[0], EMIF_nCS[3]. These 8 signals are multiplexed with NHET2 signals. By
    default, these terminals are tri-stated and pulled down. Any application that requires the EMIF functionality
    must set GPREG1[31]. This allows these 8 EMIF module outputs to be driven on to the assigned balls.

    The HALcoGen file we used did not set this bit previously, and the code worked fine with the Rev D boards that had the RML950 part on it (but did not work on the Rev E boards with the RML952 part on them). After updating our code to set this bit properly, our EMIF communications are working properly again with all versions of the HDK board.

    Jeff

  • 0 in reply to Jeff Fieldhouse
    TI__Mastermind 49120 points

    Jeff,

    There was a change made to the control of these EMIF signals from silicon rev A to rev B. Your HDK revD must have the older silicon rev, which did not have this additional control required.

    You can read the silicon revision as part of the device id register at address 0xFFFFFFF0. It is also marked on the parts themselves. The part marking looks something like:

    RM48
    L952ZWTT
    YFB-23AGPSW

    The highlighted letter shows the silicon revision.

    Sorry for the trouble caused.

    Regards, Sunil

  • 0 in reply to Sunil Oak
    Prodigy 50 points

    Thank you Sunil. I don't see this change listed in the silicon errata data sheet on your website. It just lists known problems with the silicon. Is there other changes that we should be aware of? Can you have the document updated to contain such information?

    Thank you.

    Jeff