Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Lawrie Griffiths
    @lawrie
    @Dolu1990 We did eventually run with ramfs in 32MB.
    Dolu1990
    @Dolu1990
    Yes, but the ramfs growed quite a bit ^^
    emard
    @emard
    @pnru_gitlab This is good idea, when I still remember what parts are soldered onboard. This list is at the end of MANUAL.md
    Paul Ruiz
    @pnru_gitlab
    So far I have never played with code that was sensitive to flash/sdram differences but probably I will stumble into it at some point. I have two boards, both of which are "Goran specials" I think,
    emard
    @emard
    Yes, the boards have been bought in skriptarnica cheap, then goran was replacing better parts and adding esp32 manually then sending because there was lot of demand and production was slow
    Paul Ruiz
    @pnru_gitlab
    @emard yes, that is the table I was thinking of. Maybe the production dates are already "batch numbers". So an "old green 85F" is a "Sep-2018 85F", a blue 85F is a "Nov-2019 85F" and a "new green 85F" is a "Jul-2020 85F". I assume each run uses the same sdram and flash, and only the ECP5 differs. Correct? Are there meaningful differences in other parts as well? (e.g. ESP32 hardware).
    Lawrie Griffiths
    @lawrie
    I think "blue 85F" is fairly specific as it refers to one batch of blue 85F boards with 64MB of SDRAM. "green 85F" refers to a variety of green 85F boards with different chips, but I believe they all (apart from perhaps emard specials) have 32MB of SDRAM, which is what is important for SaxonSoc builds. It seems to be the ones with the ISSI flash chips which have the problem with SaxonSoc.
    emard
    @emard
    @pnru_gitlab I'm now pulling my memory and trying to make some table with PN's. There can bi some mistakes even I'm not too shure. But "BLUE" is distinctive, made by watterott and not too many have them
    Goran Mahovlic
    @goran-mahovlic
    Yes, "blue 85F" was small production in Germany with 64Mb SDRAM
    Green boards are made by KonĨar with random mouser available 32Mb SDRAM
    emard
    @emard
    There is 3 different PNs of ISSI FLASH, 2 of 4MB and 1 of 16MB size
    Dolu1990
    @Dolu1990
    Do, can't boot in ramfs, it get out of memory
    emard
    @emard
    Should I replace SD card with ext2?
    Dolu1990
    @Dolu1990
    Yes
    emard
    @emard
    I need 5-10 min
    Dolu1990
    @Dolu1990
    p1 fat, p2 ext2
    thanks :)
    emard
    @emard
    Yes it will be the one that I have been booting 12F: I have some vic20 collection there try not to erase too much :)
    Disconnecting board from rpi
    SD card replaced, p1=fat, p2=ext2
    dtb is from 12f I can write green dtb if you want
    Dolu1990
    @Dolu1990
    Yes please, green dtb (dual corp)
    emard
    @emard
    just a moment
    dtb written
    Dolu1990
    @Dolu1990
    all is setup ^^ ?
    emard
    @emard
    yes, board connected and SD with dtb there
    emard
    @emard
    @pnru_gitlab I tried to update in MANUAL.md the table with mounted PNs from my memory. I still shoud actually look at the boards again which I have and check is that PN exactly so this list may be a bit updated
    Thomas Hornschuh
    @ThomasHornschuh

    https://github.com/emard/ulx3s/blob/master/doc/MANUAL.md#precompiled-opensource-tools-for-all-platforms here is described how to download nightly ready-compiled binaries

    These works better :-) Bitstream generation was successfull, next step will be loading it to the board. But one question: What should I enter for the 85K version on the next-pnr command? Do I need to change it from the default 45K in the blinky Makefile?

    emard
    @emard
    Yes and probably constraints also to v20. Early v17 protypes had 45K this repo is really outdated but I don't maintain it, radiona organized the mantianance of this. Mine is ulx3s-misc
    Lawrie Griffiths
    @lawrie
    The constraint file for that example says it is v20, so that should be OK. So just change -45k to -85k.
    emard
    @emard
    Aha, that should be ok!!!
    Lawrie Griffiths
    @lawrie
    Just trying that example. You also need to remove -basecfg ulx3s_empty.config
    Thomas Hornschuh
    @ThomasHornschuh
    This I had already done yesterday. I'm currently tryiing openFPGALoader, is not able to open the ftdi device, also not with sudo
    lsubs shows the device. But I need to read the docs first :-)
    Lawrie Griffiths
    @lawrie
    And change ..\..\ujprog to ujprog or fujprog or openFPGALoader
    Create this file to avoid sudo:
    lawrie@ubuntu20:~/ulx3s-examples/blinky/OpenSource$ cat /etc/udev/rules.d/80-fpga-ulx3s.rules 
    # file: /etc/udev/rules.d/80-fpga-ulx3s.rules
    # this is for usb-serial tty device
    SUBSYSTEM=="tty", ATTRS{idVendor}=="0403", ATTRS{idProduct}=="6015", \
      MODE="664", GROUP="dialout"
    # this is for ujprog libusb access
    ATTRS{idVendor}=="0403", ATTRS{idProduct}=="6015", \
      GROUP="dialout", MODE="666"
    That requires your user to be in the dialout group.
    I use fujprog rather than openFPGALoader
    Thomas Hornschuh
    @ThomasHornschuh
    @lawrie I'm familar with udev rules and ftdi chips on Linux, was just to lazy to add the rules :-)
    ulx3s suffers from to many possbilties and to many examples, docs, etc. Hard to find the right starting point. But I really like the board from the conecpt, I was backing it on CrowdSupply minutes after I received the anoucement
    Is it ok to compile fujprog from source https://github.com/kost/fujprog
    ?
    Lawrie Griffiths
    @lawrie
    Yes
    Thomas Hornschuh
    @ThomasHornschuh
    Done. Result now:
    Copyright (C) Marko Zec, EMARD, gojimmypi, kost and contributors
    Using USB cable: ULX3S FPGA 85K v3.0.8
    
    Found LFE5U-85F device, but the bitstream is for LFE5U-12F.
    
    Failed.
    At least my board was found, so partial success :-)
    Lawrie Griffiths
    @lawrie
    Are you sure you have --85k
    The examples works for me.
    Thomas Hornschuh
    @ThomasHornschuh
    Solved, I added an idcode yesterday to the ecpack line