hoglet

Members
  • Content count

    38
  • Joined

  • Last visited

  • Days Won

    2

Everything posted by hoglet

  1. hoglet

    Acorn Atom on the Duo

    I've had one more thought (a bit of a long shot...) The AVR processor on the Duo shares lines with both the PS/2 A keyboard and SD card. It's possible that a design previously loaded in there is interfering with their operation. Could you try moving the slider switch on the Duo board is pointing towards the edge of the board. This should disable the AVR. Dave
  2. hoglet

    Acorn Atom on the Duo

    Pascal, Gosh, this is proving tricky! Make sure the keyboard is plugged into PS/2 A not PS/2 B. I've just tried 4 different keyboards with my hardware and they all worked fine. One some of the keyboards, I did see all 3 LEDs flash once on power up. But I don't expect numlock/caplock to do anything, as this is not implemented in the keyboard interface on the FPGA. In fact, no commands are sent from the FPGA to the Keyboard. Dave
  3. hoglet

    Acorn Atom on the Duo

    Hi Alvie, Thanks for that info. The code that AtomMMC uses to initialize the SD Card is here: https://github.com/hoglet67/AtomFpga/blob/master/atommc2fw/mmcio.c This is not my code, and I have not really delved into it at this level before, but I believe that mmc_initialize sends 88 clocks with data=1 and CS=1: for (n = 11; n; --n) XFER_SPI(0xff); /* Dummy clocks */ There was some evidence in a stardot thread that not all cards support SPI mode: http://stardot.org.uk/forums/viewtopic.php?f=3&t=9467 It was a 512MB transcend in this case. Edit: there is more here: http://stackoverflow.com/questions/955587/initialization-of-a-microsd-card-using-an-spi-interface It would be great if one or two more people can try out this design, and see if the problems that Goos is having can be replicated. Dave
  4. hoglet

    Acorn Atom on the Duo

    I'm not sure what the problem with the bit file could be, as it works when written to FPGA RAM. I've been able to program it, as has one other person. Very strange. I think it would help to have a couple more data points. Would anyone else be able to test this out? Dave
  5. hoglet

    Acorn Atom on the Duo

    Hi Goos, Try pressing just F10 for about a second, then releasing it, then *CAT. If this still gives ERROR - NOT READY, wait a few seconds and try again a couple more times. If this doesn't work, try doing the same but with the Reset button on the classic computing shield. Occasionally we have come across newer SD Cards that can't be put into SPI Mode (which is what AtoMMC uses). You said the brand was Scandisk - did you mean SanDisk? They are normally very good. Regarding your other problem - not being able program the design to FLASH. Are you able to successfully program other designs? I can't think of any reason this design would be different, especially as you have verified the bitstream is good by programming it into RAM. I wonder if anyone else has any suggestions on the programming issue. Dave
  6. hoglet

    Acorn Atom on the Duo

    Hi Goos, OK, lets tackle these one at a time.... When you say the SDCard isn't working, can you be a bit more specific? What actually happens. Photos always help! Dave
  7. hoglet

    Acorn Atom on the Duo

    Lak, I think I know what happened - I had added the wrong .bit file into git - one without the Soft AVR8 firmware included. Can you try the one I've just pushed? Dave
  8. hoglet

    Acorn Atom on the Duo

    Hi Lak, Can you post a photo of what you are seeing (a picture is worth 1000 words!) Does the behaviour change if you remove the SD Card? Dave
  9. hoglet

    Acorn Atom on the Duo

    Jack, Thanks for the reply. I'll post here if I get any further with using the ATmega32U4. Dave
  10. hoglet

    Acorn Atom on the Duo

    Jack, I have a quick question about the Classic Computing Shield / ATmega32U4 wiring. It looks like the MicroSD card is connected to the wrong pins on the ATmega32U4. According to the ATmega32U4 datasheet, the SPI interface uses pins B0, B1, B2 and B3. But on the the classic computing shield, the MicroSD card is wired to pins B5, B6, C6, C7. Why the difference? I guess I can used the FPGA to re-route these, but only if I then disable B5, B6, C6, C7 in the ATmega32U4, which loses me valuable I/Os. Also, why's ATmega32U4 PD5 not connected? It would have been nice to be able to use Port D as a byte wide data bus. Is this correct, or am I mis-reading something? Dave
  11. hoglet

    USB Port in Papilio Duo (Technical Question)

    Unless it really is a private conversation, please keep this discussion on the forum. I, and lots of others are finding it very interesting and informative. Many thanks Dave
  12. Does anyone have any info/updates on the Duo availability for those who missed the Kick Starter round? It's kind of hard to do a port like this without the hardware. Dave
  13. I'd be more that happy to port the Acorn Atom FPGA implementation to the Duo: https://github.com/hoglet67/AtomFpga http://www.stardot.org.uk/forums/viewtopic.php?f=44&t=6313&hilit=atomfpga&start=150#p88030 This is currently a DIY wing for a Papilio One (which includes the ZPU core running the SDCard stack) Jack, I wasn't part of the Kick Starter - when are Duo's likely to be available through distributors? Dave