1. OCAU Merchandise now available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion here.
    Dismiss Notice

MiSTer FPGA (computer/console/arcade hardware simulation)

Discussion in 'Retro & Arcade' started by elvis, Oct 14, 2018.

  1. Jumpingmanjim

    Jumpingmanjim Member

    Joined:
    Sep 20, 2008
    Messages:
    613
    Ok just stopped watching Amiga demos. I checked everything again while they were running and still 5.05V.
     
  2. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    7,966
    Location:
    4109
    just for reference
    if you had access to each side of the switch (or a long run of cable) this is a better check.
    with the device running at max power you measure volts across the switch terminals.
    the voltage dropped multiplied by the current is the power lost as heat in the switch P=V.I
    so as an example the mister might draw 3A
    if the switch was dropping 0.5 volts thats 1.5W of heat.
    doesn't sound like much but in that little switch it adds up quick
    a good switch would drop much less than 0.1 volts.

    ultimately as long as the mister gets its 5v3a and the switch doesn't melt down then it doesn't matter

    edit - youd need to measure the current too
     
  3. Jumpingmanjim

    Jumpingmanjim Member

    Joined:
    Sep 20, 2008
    Messages:
    613
    I've been flogging it pretty hard today so i'll report back if anything happens.
     
    darknebula and MUTMAN like this.
  4. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    7,966
    Location:
    4109
    if youre still see 5+ volts you should be in the clear for instability.
    long term the switch at only 2A is slightly underrated so it'll worth keeping an eye on it
     
  5. shredder

    shredder Member

    Joined:
    Dec 26, 2001
    Messages:
    13,969
    Location:
    New Zealand
    I like to give switches a lot of rating headroom these days. Had a few pushed (even vaguely) close to their rating fail over the years. From my experience in those situations, the switch tends to sort of just 'fuse' into permanently-on status, regardless of switch position.
     
    Last edited: Aug 11, 2020
    MUTMAN likes this.
  6. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    7,966
    Location:
    4109
    the plastic tabs that typically are used to hold the contact against a spring melt/fatigue due to the heat/wear.
    and that releases the pressure on the contact a little bit increasing the resistance, and that increases the heat. makes the situation worse and the whole thing spirals into a switch mech failure/contact weld.
     
  7. Jumpingmanjim

    Jumpingmanjim Member

    Joined:
    Sep 20, 2008
    Messages:
    613
    Sounds cool
     
  8. Andrew_Wong

    Andrew_Wong Member

    Joined:
    May 6, 2008
    Messages:
    119
    Location:
    Perth, Australia
    I have been quiet for a while.. I've got IOBoard v6.1s in the pipeline now, and the 3D printer has started behaving itself again so the custom 3D printed cases are a thing again. Have some interesting stuff on the go in the workshop, but have just got enough components to build more IOboards and SDRAM modules.

    Am working on a plasma cut steel metal case design.. with a 3D printed press brake on the schedule later this year :D:D
     
    flain, elvis, MUTMAN and 1 other person like this.
  9. SCSA187

    SCSA187 New Member

    Joined:
    Jun 28, 2020
    Messages:
    10
    I look forward to seeing the new case design! :)
     
  10. Andrew_Wong

    Andrew_Wong Member

    Joined:
    May 6, 2008
    Messages:
    119
    Location:
    Perth, Australia
    Just on the current draw from a DE10.. here's a plot for a DE10 with 128MB SDRAM, 5v fan, network and USB keyboard plugged in.

    Struggles to go any higher than 1.2A (excluding any high frequency spikes) even when running update_all.sh.

    [​IMG]
     
    elvis and MUTMAN like this.
  11. flain

    flain Member

    Joined:
    Oct 5, 2005
    Messages:
    2,758
    Location:
    Sydney
    Id be interested to see the draw with a BT dongle or wireless keyboard dongle combo as that was a problem for me before i got a powered hub (i assume the usb port is just not able to draw enough).

    Also a good test is to boot a big game in the neogeo core, as that seemed to be causing a lot of issues for people.
     
  12. Andrew_Wong

    Andrew_Wong Member

    Joined:
    May 6, 2008
    Messages:
    119
    Location:
    Perth, Australia
    Haha.. never test after working all day.. forgot I was hooked into a powered USB hub.

    Add 100mA to the chart above for the Apple USB Keyboard. My Rii USB dongle didn't make a noticeable difference. So still under 1.3A peak.

    Even loading and running KOF94 on NeoGeo.. or anything else.

    So unless your USB hub and peripherals draw another 0.7A or 3.5W (which is quite a bit).. you're unlikely to hit 2A.
     
    flain likes this.
  13. Jumpingmanjim

    Jumpingmanjim Member

    Joined:
    Sep 20, 2008
    Messages:
    613
    Should be alright as long as I don't plug in my USB toaster

    [​IMG]


    How does your 3d printed case compare to the official one? Why did you make your own design?
     
    MUTMAN likes this.
  14. Andrew_Wong

    Andrew_Wong Member

    Joined:
    May 6, 2008
    Messages:
    119
    Location:
    Perth, Australia
    A few reasons..

    i. There are 2 things that generate heat on the DE10 - the power regulators and the FPGA. I tested airflows (see my website)
    ii. So I wanted an 80mm fan running at half speed - airflow and not much noise. And much much cheaper than a Noctua 40mm fan
    iii. My case is modular, so depending on whether you run a IOBoard or not, you have options
    iv. I can make it in whatever colour of ABS you like
    v. You can decide to sand it, respray it and finish it however you like - get creative!
    vi. It's roughly half the price of getting an acrylic case landed in Australia
    vii. If you break it I can get you a replacement easily enough for a few bucks
    viii. It was fun and a good project to go from drawing to a part that just works

    btw the fan & case has turned out to be pretty good for a Raspberry Pi4 too.. my Octoprint Pi has no heatsink on it.. doesn't get above
    40C ever :leet::leet::leet:
     
    MUTMAN likes this.
  15. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    7,966
    Location:
    4109
    flain likes this.
  16. gdjacobs

    gdjacobs Member

    Joined:
    Apr 3, 2007
    Messages:
    2,543
    Location:
    MB, Canada
    That looks tasty!
     
    MUTMAN likes this.
  17. flain

    flain Member

    Joined:
    Oct 5, 2005
    Messages:
    2,758
    Location:
    Sydney
    Finally actual 486 performance :)
     
    elvis likes this.
  18. OP
    OP
    elvis

    elvis Old school old fool

    Joined:
    Jun 27, 2001
    Messages:
    42,228
    Location:
    Brisbane
    This confirms that I need 3 MiSTer setups in my life:

    * Console
    * Arcade
    * Computer
     
    MUTMAN likes this.
  19. shredder

    shredder Member

    Joined:
    Dec 26, 2001
    Messages:
    13,969
    Location:
    New Zealand
    How finite are the components in this? Is there likely to be a MiSTer in 2 years? 5?
     
  20. OP
    OP
    elvis

    elvis Old school old fool

    Joined:
    Jun 27, 2001
    Messages:
    42,228
    Location:
    Brisbane
    Been asked a few times and there's no official answer, but here's my thoughts:

    The DE10Nano is at the heart of the MiSTer. It's important specs are the Cyclone V FPGA and ARM Cortex A9 support CPU.

    This is an educational product at it's core, aimed at teaching university level students FPGA programming. It's got big financial backing by Intel, and an intended long tail.

    The cores that are written for MiSTer are mostly ports from other FPGA platforms, many of which use earlier Cyclone FPGAs. The cores are all open source, and hold great value as both working code and good quality technical documentation.

    Should a new platform be considered, it's likely that whatever work exists now can be ported, so it's not ever wasted work. Beyond that, if these cores are developed and are accurate, they stay that way. Say, for example, the DE10Nano ceases production today and all the MiSTer devs vanished, you've still got an incredibly feature packed system that isn't going to grow old or be outdated overnight. As long as screens exist that take HDMI in, you've got a functional system.

    With all that said, there's no timeline on what the future holds for MiSTer. It's not a commercial product, and the hardware it's based on has no EOL announced. It's very much a case of "how long is a piece of string?".

    For me, the dollar investment is clearly good value. Even today without any future expectations, you're getting a system that can accurately simulate dozens of systems that would cost you thousands of dollars to acquire today, many that would require extensive and expensive modding on top of that to be functional on modern displays at all, let alone to a high visual quality.

    For all of these reasons, I don't think the question of "how much longer will this last?" even matters.
     
    Last edited: Aug 15, 2020
    Fitzi, MUTMAN, flain and 1 other person like this.

Share This Page

Advertisement: