Hacking WIP [Trinket] [Rebug] [Others] SWITCHBOOT_UF2 / FUSEE_UF2 modchip software

[FR0ZN]

Not really problems... Wondering why you need to use different UF2 files. That would make no difference whatsoever.

You see the green and yellow lines in the pic? They are USB ready. They can be linked straight to for example a micro-USB connector if you wish.

If you remove the logic chip, you are effectively turning the chip into a mini "trinket" without LED or DotStar. There will be no drain at all from holding the "enable" pin high during sleep.

And because at that point, nothing is connected to D3 taking power, you can leave it high or low. Makes no difference as nothing is taking power from D3 as it is unused.

So, no need to mess around with old software.

And yes. UF2 splicer is good. But I prefer doing things the manual way. I don`t trust anyone else to do it for me. Especially how we chop up and dissect / regenerate the payload on-the-fly via chip firmware

Ok that sounds good - then my understand of everything was completley wrong.

Different plan then:

I want the following:

1) No drain of any power during sleep mode -> USB IC has to go I understand this right
2) Doinga 4 wire install here, as I go with AutoRCM
3) Use one of your UF2s that load my payload.bin during boot from SD card -> which one can I use once I'm done with number 1 ?

I hope this constelation is possible.

 

[mattytrog]

Ok that sounds good - then my understand of everything was completley wrong.

Different plan then:

I want the following:

1) No drain of any power during sleep mode -> USB IC has to go I understand this right
2) Doinga 4 wire install here, as I go with AutoRCM
3) Use one of your UF2s that load my payload.bin during boot from SD card -> which one can I use once I'm done with number 1 ?

I hope this constelation is possible.

Of course possible.

Use any you like apart from Feather (needs a crystal oscillator), Trinket Method3(D2 is capacitive to sense POR from pin 5 on M92T36 - yes we used to do that) and legacy3... The same as method3 but rather than pin 2(D2), it is pin 4(D4) which is capacitive.

Bear in mind if LED fitted, LED probably won`t work.

 

[FR0ZN]

Of course possible.

Use any you like apart from Feather (needs a crystal oscillator), Trinket Method3(D2 is capacitive to sense POR from pin 5 on M92T36 - yes we used to do that) and legacy3... The same as method3 but rather than pin 2(D2), it is pin 4(D4) which is capacitive.

Bear in mind if LED fitted, LED probably won`t work.

Thanks for the information, I actually don't need a LED as I won't be able to see it anyway

But I wonder what you mean with the D2 / D4 capactive pins ... why is that information relevant for me? Afaik I only need the D-, D+, GND and 3v3 pin if I'm not mistaken for the 4 pin / AutoRCM config correct?

 

[mattytrog]

Thanks for the information, I actually don't need a LED as I won't be able to see it anyway

But I wonder what you mean with the D2 / D4 capactive pins ... why is that information relevant for me? Afaik I only need the D-, D+, GND and 3v3 pin if I'm not mistaken for the 4 pin / AutoRCM config correct?

No it isn`t. Nothing is connected to D2 / D4 as we know. However... If nothing is connected to them (or not shunted to GND or whatever), whenever they get a sniff of capacitance, they activate and reboot the chip. Example: Flash method3 and legacy3 then hold your chip in your hand (or touch D2/D4). It will bootloop because it is sensing your finger touching it.

 

[kiwihop]

Hi mattytrog. I've been screwing around to try to solve the issue with the USB IC on RCM x86 on hardware level to not make it drain any battery.

The PCB that I have for the modchip is different from the screenshot that are on the thread.

On this "new" board there are some pads that make it easier to solder directly to the SAMD chip:

With this, as you said, the chip will not drain any battery.

Thanks!

 

[FR0ZN]

No it isn`t. Nothing is connected to D2 / D4 as we know. However... If nothing is connected to them (or not shunted to GND or whatever), whenever they get a sniff of capacitance, they activate and reboot the chip. Example: Flash method3 and legacy3 then hold your chip in your hand (or touch D2/D4). It will bootloop because it is sensing your finger touching it.

Ah ok, so you say when I use:

- Trinket Method 3 I need to ground D2
- Trinket Legacy 3 I need to ground D4

Did I get that right now?

Jesus christ why the f*** did I buy this chip

 

[mattytrog]

Ah ok, so you say when I use:

- Trinket Method 3 I need to ground D2
- Trinket Legacy 3 I need to ground D4

Did I get that right now?

Jesus christ why the f*** did I buy this chip

The chip is fine! It just has its "quirks". Designed to overcome a different set of problems.

Just don`t use method 3 or legacy 3. Use trinket or trinket / rebug.

--------------------- MERGED ---------------------------

Hi mattytrog. I've been screwing around to try to solve the issue with the USB IC on RCM x86 on hardware level to not make it drain any battery.

The PCB that I have for the modchip is different from the screenshot that are on the thread.

On this "new" board there are some pads that make it easier to solder directly to the SAMD chip:

With this, as you said, the chip will not drain any battery.

Thanks!

I`d probably chop the traces going to the logic IC from D-/D+.

 

[FR0ZN]

The chip is fine! It just has its "quirks". Designed to overcome a different set of problems.

Just don`t use method 3 or legacy 3. Use trinket or trinket / rebug.

Thanks for the info !

One more question, is there a Fusee-UF2 that is like very simple?
Simple as in, it always expects a payload.bin at the root of the SD card?
I'm looking a payload.bin loader that doesn't have features like listening to Vol-/+ buttons etc. - doesn't that cause trouble with hekate as payload.bin anyway? afaik hekate also listens to the Volume buttons.

 

[mattytrog]

Thanks for the info !

One more question, is there a Fusee-UF2 that is like very simple?
Simple as in, it always expects a payload.bin at the root of the SD card?
I'm looking a payload.bin loader that doesn't have features like listening to Vol-/+ buttons etc. - doesn't that cause trouble with hekate as payload.bin anyway? afaik hekate also listens to the Volume buttons.

No it doesn't cause trouble. The simplest is fusee v1. No buttons or strap support (apart from usb strap). Loads payload.bin. if payload not found, console powers off. It's on my git

 

[peteruk]

thank you for all the help with everything and for the continued software updates, very possibly the most patient man I know

 

[josete2k]

Drain battery is only in sleep mode?

If so... If we power off the console we wouldn't care about it, isn't it?

 

[mattytrog]

Drain battery is only in sleep mode?

If so... If we power off the console we wouldn't care about it, isn't it?

Correct

--------------------- MERGED ---------------------------

Only while the console is on. It's miniscule amount anyway nothing to worry about

 

[FR0ZN]

No it doesn't cause trouble. The simplest is fusee v1. No buttons or strap support (apart from usb strap). Loads payload.bin. if payload not found, console powers off. It's on my git

Thanks for the info, I didn't know you also had such a UF2 on the menu !
I don't understand this strap thingy you mentioned, what exactly is that?

EDIT: Oh wait, I just found the spoiler

EDIT2: Ok so with "buttons support" you mean that there is no features by pressing any of the volume buttons during boot - and "strap support" means, the software on the chip can't enable RCM mode correct? I have to either use AutoRCM (which I want) or bring it into RCM myself.

And USB strap? Never heard of it.

 

[PSxed]

Currently running the latest firmware which is working great. I'm trying out L4T Lakka which is auto booting perfectly fine using a custom entry in hekate_ipl.ini. There is a no proper shutdown, atleast i can't see it and in their docs they say the following: "the shutdown option in Lakka was removed for safety reasons ; to shutdown, reboot to RCM then press the POWER button for twelve seconds. It is a good idea to boot in Horizon to shutdown properly from here.".

I get stuck in the typical black rcm screen. Would it be possible to implement a way for us to return to hekate and shutdown over there?

 

[mattytrog]

Currently running the latest firmware which is working great. I'm trying out L4T Lakka which is auto booting perfectly fine using a custom entry in hekate_ipl.ini. There is a no proper shutdown, atleast i can't see it and in their docs they say the following: "the shutdown option in Lakka was removed for safety reasons ; to shutdown, reboot to RCM then press the POWER button for twelve seconds. It is a good idea to boot in Horizon to shutdown properly from here.".

I get stuck in the typical black rcm screen. Would it be possible to implement a way for us to return to hekate and shutdown over there?

Hmmm... Doubtful.

The only solutions that spring to mind, are getting on the latest firmware, or at least the firmware that matches the amount of fuses you have burnt... You should be in a position to be able to boot stock without any payload. (ie disable RCM and you can boot)

And enabling dual-boot (holding down vol+ to boot to CFW from a power-off) - flashing the "joycon only" uf2 for your chip - Switchboot part 1...

Then flashing part 2 again (as your chip will be blank after changing Switchboot part 1)

Once you can do this... You can just power off and you won`t have a black RCM screen. To boot Lakka again, just hold vol+ to dual-boot as usual.

............................................................
Alternative plan/idea...

I developed a "live-rcm" feature ages ago for people without straps... This will disable autoRCM and re-enable it as needed. But never put it in. However, for this to work, the payload that has "live-RCM" running (in this case "switchboot") needs to run AFTER L4T Lakka has ran (to re-enable autoRCM).

If in L4T Lakka, they could make a "reboot to payload" feature, this could work. But it would need to be implemented as Atmosphere does it.

 

[mattytrog]

@peteruk

and anyone who would like to try it...

I`ve made a UF2 and a Switchboot.bin which I have attached... Some fairly radical changes underneath. This will only be any good for people with vol+ strap fitted (to change mode).

These are:

• 4 modes now
• Mode 1: Switchboot 1.3.4 - 8 payloads
• Mode 2: Fusee-UF2 2.1 - 8 payloads
• Mode 3: Switchboot 1.3.4 - 8 payloads - autoincreasing on fail(for strapless users - will need hex edit UF2 to change mode)
• Mode 4: Fusee-UF2 2.1- 8 payloads - autoincreasing on fail (for strapless users - will need hex edit UF2 to change mode)
  
• paths_info.txt & straps_info.txt no longer on SD root. They go in a new folder in bootloader/switchboot_info
• A path in fusee-uf2 corrected... Before was Payloads/payload.bin or payload1.bin... Now corrected to bootloader/payloads/payload.bin or payload1.bin
  
• Changing from Fusee to Switchboot and vice-versa will update and regenerate strap information as required
• If you use the chip to choose payload (rather than Hekate/Switchboot/Argon), this will be duplicated across Fusee-UF2 and Switchboot. So, if you are currently on "payload3.bin", launching Switchboot.bin, change mode and "payload3.bin" voa Fusee-UF2 will load

Should I issue Switchboot v1.3.5 onwards like this? Fusee-UF2 will be unchanged. Still flash and forget... Just adding it as an optin to the Switchboot UF2...

This is all about "trying" to make it easier for resellers to set up consoles and for users to keep their consoles updated...

Anyway... Here are the files...

Extract switchboot.bin and put it in bootloader folder (will probably be alongside hekate_ipl.ini & update.bin) if you want strap information. If you don`t need the strap information, forget Switchboot zip below.

 

[peteruk]

1.3.5 looks awesome, great job !

look forward to testing it this weekend - thank you

 

[peteruk]

Hi Matty,

Just a quick feedback regarding my findings, I'm using the switchboot.bin method

It's still placing the straps_info and the other file on the root of the card if I delete it and not inside bootloader/switchboot_info

Apart from that working fine mate

 

[mattytrog]

Hi Matty,

Just a quick feedback regarding my findings, I'm using the switchboot.bin method

It's still placing the straps_info and the other file on the root of the card if I delete it and not inside bootloader/switchboot_info

Apart from that working fine mate

Thank-you sir. Must have forgotten a path or I`ve stitched a wrong file in...

 

[mattytrog]

Hi Matty,

Just a quick feedback regarding my findings, I'm using the switchboot.bin method

It's still placing the straps_info and the other file on the root of the card if I delete it and not inside bootloader/switchboot_info

Apart from that working fine mate

Just checked and it definitely should not be placing the files on SD root still. It is not possible for the UF2 to place straps_info.txt above in SD root. Another Fusee-UF2 must be running.

Just to confirm... You are following these steps to properly test:

1) Flash UF2 above
2) Place Switchboot from above post in bootloader folder

3) You need to change modchip "mode" via your volume button, so it launches Fusee-UF2 to properly test.

So using your volume button, change mode on your chip by holding down volume, as described in OP. Give it a try

Speak soon sir!