WIP Modchip Software: Simple-UF2 with SWITCHBOOT. Save fuses without AutoRCM

Discussion in 'Switch - Backup Loaders & Modchips' started by mattytrog, Dec 18, 2018.

  1. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    With no straps connected, the chip won't even wake up. If other things are being recognised via USB (is joypads) then the chip isn't the issue.

    The issue is the host mode "stack" or the homebrew itself.
     
  2. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    I figured the Trinket couldn't cause this issue but it seems like such a strange coincidence for the volume button to be affected. The issue exists as soon as the console is powered (even in RCM) so I don't think it's caused by software. I think my motherboard must have some abnormal issue.

    After letting it sit overnight, powered off, the issue is gone, although Tinfoil USB install still produces as error. I will try to recreate the error with the Trinket disconnected.
     
    Last edited by JacksonS, Feb 10, 2019 at 4:36 PM
  3. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    Ok. Once the trinket goes to sleep, it is as though it is not even there.

    Like I say, if joypads are working (plugged directly into the switch) and the dock, and memloader come to think of it (SD & emmc ini files), then clearly the issue isn't in the chip or software.

    I normally don't touch the horizon homebrew side. Not my area. But I'll take a look at tinfoils code later and see if I can get the same issue. Keep me posted

    :)
     
    JacksonS likes this.
  4. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    I haven't test Tinfoil with the Trinket desoldered but I believe the issue is purely software. It seems to happen if I use certain install paths on the PC (e.g. if I define the path as relative it will crash, but not if I define it as absolute).

    My issue with the volume button hasn't gone away but it's definitely a hardware issue. I've measured the voltage on the Up button and it does go fairly low without being pressed, around 0.5v or 0.3v sometimes, while the Down button is always 1.8v. So the Up button isn't shorted to ground but it's defective somehow.

    My guess is that the Tegra uses an internal 1.8v pullup on both volume buttons and the Up button pullup is broken. I can only assume this is because of the Trinket; I had the Trinket connected to the button at some point before I finished the install. It probably pulled the button input high to 3.3V while in bootloader mode or something, and I'm guessing the input is not 3.3v tolerant, so it broke. It sucks but I will try putting an external 1.8v pullup on it and hope it works.
     
    Last edited by JacksonS, Feb 10, 2019 at 8:55 PM
  5. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    The problem will be where the 150R resistor is. If that is shorted to GND (it is a pullup. You are correct) then your vol+ is "stuck".

    However, you can safely remove the resistor and link it. The GPIOs can take considerably more current than that.
     
  6. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    That resistor must just be a passthrough, not a pullup. I'd expect a pullup to be above 1k ohms. I've looked at the 150 ohm resistor and it's not shorted to anything.

    Do you know if there is an external pullup resistor anywhere else on the board? If there is, I don't think I ever touched it, so it would be really strange for it to be broken.
     
    Last edited by JacksonS, Feb 10, 2019 at 9:53 PM
  7. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    No the resistor isn't a pull-up. The actual gpio is. The resistor just limits current
     
  8. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    Well, I've looked at the motherboard for a long time and I can't seem to fix the volume button issue. The physical button is fine; I disconnected it and it still thinks it's "pressed". Replacing the 150 ohms resistor with a wire changed nothing. I added an external pullup, but the button is still being pulled low by something else at random times. It looks like my motherboard is just broken after using the Trinket.

    I'd advise anyone else to be very cautious about connecting a microcontroller directly to any GPIO on the Switch.
     
    Last edited by JacksonS, Feb 11, 2019 at 2:57 PM
  9. Dougiejones

    Dougiejones Advanced Member

    Newcomer
    1
    Jan 13, 2019
    United States
    Test_Dotstar_Brightness done :

    Dotstar brightness/fading works for me since Trinket's UF2 flashed. But blackout after.
    Indeed unable to boot the console in CFW (SX OS) just after and so even with a long power pressed button ( 20 seconds ) to reset the console. Had to boot OFW Horizon then reflash 0.9.8r2.
     
  10. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    You just need to find where it is grounded.

    It must be grounded somewhere. There is a resistor shunted to gnd further down the trace. That could be shorted.

    Can you continuity test the trace?
    Yeah. I thought so.

    I missed an #include out :(

    I`m too old and too forgetful for this "programming" nowadays
     
  11. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    It's definitely not grounded. I've mentioned I've tested continuity and it's not shorted to anything. The GPIO is connected just as it should be, but it's being pulled down internally by the Tegra or something else that's broken.

    Can you show me where this other resistor is? I'd like to inspect every external component connected to the GPIO.
     
    Last edited by JacksonS, Feb 11, 2019 at 5:34 PM
  12. Dougiejones

    Dougiejones Advanced Member

    Newcomer
    1
    Jan 13, 2019
    United States
    No no...not bad. Keep going on ;)
     
    Last edited by Dougiejones, Feb 11, 2019 at 6:01 PM
  13. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    The bottom resistor. The end of the trace marked in red. This then goes to a testpoint on the reverse of the board if memory serves.
    volup.
     
  14. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    Thanks for the visual. That component looks like a capacitor to me. I haven't tried removing it but I don't have much hope that it's related to this issue.
     
  15. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    Are you sure? Looks like a resistor. I`m getting 75K ohms across it.
    capacitance in low pF. Basically nothing.

    Can`t see why it would be a capacitor. It isn`t decoupling anything, doesn`t need to smooth anything, doesn`t need to condense anything.

    Looks defo like a shunt resistor to me. Give the 1.8v somewhere to go. Or its a pulldown to drop GPIO from 3v to 1.8. I don`t know.

    Anyway... What readings are you getting? impedance between that trace and GND?
     
    Last edited by mattytrog, Feb 11, 2019 at 7:03 PM
  16. JacksonS

    JacksonS GBAtemp Fan

    Member
    4
    Feb 13, 2016
    United States
    Georgia
    It is not a resistor, you can tell by the shape. With the component removed from the circuit, it reads nearly infinite resistance and 100 pF in capacitance. It would make sense for it to be a capacitor since this is a button and it needs to be debounced.

    Even though the capacitor seems fine, my issue seems to have gone away after removing it... I'm still wary that the issue will come back but it's been almost a day and the volume isn't going up on its own.
     
  17. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    Just tested.

    Absolutely correct sir!

    Are my eyes really that bad?

    I should have (supplied a signal) & tested xC (capacitive reactance) in the first place.

    It is indeed a 100pF cap. Probably Murata. Tested out of circuit this time :blush:
     
    Last edited by mattytrog, Feb 12, 2019 at 8:07 PM
  18. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    OK... 0.9.8r3 coming... No 0.9.9 yet. 0.9.9 will have FW 7.0 support

    Changelog:

    • Hekate updated to 4.6svc - added 77620 rtc control. See hekate git for more information
    • Autochainload now has an option to press vol- during boot and it is displayed (1 second)
    • USB strap changed to just sense falling voltage rather than a change (shouldn`t wake the chip up when plugging things in)
    • Modchip USB connection completely disabled at sleep by setting CTRLA bit to 0
    Coming later.

    Got a PI3USB to change and a LCD connector to do first. :)
     
    Last edited by mattytrog, Feb 12, 2019 at 8:04 PM
    Dougiejones likes this.
  19. Dougiejones

    Dougiejones Advanced Member

    Newcomer
    1
    Jan 13, 2019
    United States
    Hasty to test and give feedback :D

    And good luck for the rest !
     
  20. mattytrog
    OP

    mattytrog Not on my watch.

    Member
    10
    Apr 27, 2018
    United Kingdom
    Thank-you sir! You are my main beta tester!
     
    Dougiejones likes this.
Loading...