It's a recovery mode, so it's possible that it doesn't care for kernel panics. But it all depends if the switch got the rcm bug.
He tried that already.
We definitely need a low level payload for checking the Switch's hardware.
We definitely need a low level payload for checking the Switch's hardware.
I'll see what I can do when I get home, I'll prolly make a payload that tests the parts and etc, gives as much info as it can; I'll prolly make it use UART-A for dumping the info too in-case MicroSD is not working
The caps around the BCM don't show any shorts. I got it from ZedLabz, a UK eBay seller, but looks legit, well packaged with clean solder balls, so most likely new from the factory.
Hmm maybe? but I can't be anything major as all the functionality is there: screen, touch, audio, buttons, joycons, SD card, I can get into RCM, or boot Linux or Android. Only Horizon refuses to boot.
Yeah the orange screen only appears if the switch is running afaik, for me it only happened once at the very beginning while the switch was running, after which the boot issue started.
Yeah I can definitely boot Linux! As mentioned the WiFi adapter is not detected there, but I don't know what/how else to test there.
Wow thank you that would be amazing! My microSD slot definitely works without a hitchI'll see what I can do when I get home, I'll prolly make a payload that tests the parts and etc, gives as much info as it can; I'll prolly make it use UART-A for dumping the info too in-case MicroSD is not working
-----
I was actually wondering how you can read out UART from the switch
Also do you know if Horizon puts out any logs during boot?
Last edited by GrimDim,
I don't know much about the bluetooth/wifi module. But I bet it's solder on the board? I don't know how experienced you're with a hotgun station incase it's really confirmed it is that chip.
Yeah that was my suspicion, which is why I replaced the chip (using hot air). I botched my first attempt though (shown in the third picture, here the switch would not power on), then I took the chip off again, reballed it (twice until I got it halfway right, here the switch was back to getting stuck on boot), and then I bought another one thinking better use factory balling to make sure. That one went on well I think and produced the first picture.
Hmm... well I can't really help much then. I'm superbad with these kind of things and the peeps I pinged know a lot more. Especially @mattytrog that seem to be the Wizard of making the Switch talk on hardware.
Orange screens are Secure Monitor panics, usually they indicate some kind of SoC hardware failure.
It wouldn’t matter for RCM, correct.
Do you know if ReSwitched got some tool or such to dump the boot process to see where it fails?
Could use UART to dump out the process; unsure if there's a dump to log but maybe there is; just check AMS Git; if you can inject AMS and get a dump during the boot or output to UART.
edit: AMS does have UART support, just checked the git; most likely will need to self compile it.
So it can't be outputted like so due to how things work, sorry for the bad news
Last edited by TariqSoftDev,
Hi,
i have the same issue here. I replaced the BCM4356 today with a new one, but still stuck on the switch logo. I can boot into android but no wifi but blutooth works. So maybe we work this out together.
Im working in a computer manufactory. So we have many stuff like xray and so on.
Tomorrow I check the caps for shorts.
This must be the wifi not working thing!
i have the same issue here. I replaced the BCM4356 today with a new one, but still stuck on the switch logo. I can boot into android but no wifi but blutooth works. So maybe we work this out together.
Im working in a computer manufactory. So we have many stuff like xray and so on.
Tomorrow I check the caps for shorts.
This must be the wifi not working thing!
Hey, awesome to see someone stuck in the same boat as me!
Today I actually came across some threads here where people mentioned that Nintendo might use custom firmware for the chip which could mean that new chips are not going to work. Not sure if that's true or not, but since I have some donor boards lying around I will be trying to use the BCM chip from one of those boards next, maybe tomorrow if I have time.
With my previous attempts I feel semi-confident now to perform a reball (as painful as it is). Will report back with the results!
Hey, awesome to see someone stuck in the same boat as me!
Today I actually came across some threads here where people mentioned that Nintendo might use custom firmware for the chip which could mean that new chips are not going to work. Not sure if that's true or not, but since I have some donor boards lying around I will be trying to use the BCM chip from one of those boards next, maybe tomorrow if I have time.
With my previous attempts I feel semi-confident now to perform a reball (as painful as it is). Will report back with the results!
My guess thats not true, because the Blutooth work. And both are on the same chip.
Maybe there is a tool for Android to read out why its not initializing.
Last edited by acidcoolxxx,
Maybe the WiFi module has different firmware than the Bluetooth module?
---
Actually, I found this section from the datasheet (https://www.cypress.com/file/298796/download) quite interesting:
9.2 One-Time Programmable Memory
Various hardware configuration parameters may be stored in an internal One-Time Programmable (OTP) memory, which is read by the system software after device reset. In addition, customer-specific parameters, including the system vendor ID and the MAC address can be stored, depending on the specific board design. Up to 484 bytes of user-accessible OTP are available. The initial state of all bits in an unprogrammed OTP device is 0. After any bit is programmed to a 1, it cannot be reprogrammed to 0. The entire OTP array can be programmed in a single write cycle using a utility provided with the Cypress WLAN manufacturing test tools. Alternatively, multiple write cycles can be used to selectively program specific bytes, but only bits which are still in the 0 state can be altered during each programming cycle. Prior to OTP programming, all values should be verified using the appropriate editable nvram.txt file, which is provided with the reference board design package.
It is referencing a memory that can only be initialized once, and contains hardware configurations. Maybe this is the exact information missing on factory chips to make them work with the switch?Various hardware configuration parameters may be stored in an internal One-Time Programmable (OTP) memory, which is read by the system software after device reset. In addition, customer-specific parameters, including the system vendor ID and the MAC address can be stored, depending on the specific board design. Up to 484 bytes of user-accessible OTP are available. The initial state of all bits in an unprogrammed OTP device is 0. After any bit is programmed to a 1, it cannot be reprogrammed to 0. The entire OTP array can be programmed in a single write cycle using a utility provided with the Cypress WLAN manufacturing test tools. Alternatively, multiple write cycles can be used to selectively program specific bytes, but only bits which are still in the 0 state can be altered during each programming cycle. Prior to OTP programming, all values should be verified using the appropriate editable nvram.txt file, which is provided with the reference board design package.
That would be handy
Last edited by GrimDim,
I found something like this
https://play.google.com/store/apps/details?id=com.dp.logcatapp&hl=en
Not sure if u can install it on the Switch. The Switch have root on android?
https://play.google.com/store/apps/details?id=com.dp.logcatapp&hl=en
Not sure if u can install it on the Switch. The Switch have root on android?
Similar threads
-
-
-
@
Psionic Roshambo:
It's great if you have something anyone else wants, otherwise it's annoying and potentially bad if you lose your keys -
-
-
@
SylverReZ:
@Xdqwerty, Its good if you have any important documents for say a business that you don't want leaking in the wild.+1 -
-
-
@
SylverReZ:
Bad news is that BitLocker was cracked, so somebody could use a Pi Pico and probe the LPC pads on the motherboard to grab the key.+1
-
-
-
-
-
-
-
@
BigOnYa:
My wife's doctor prescribed her a ointment for neck and knee pain, that cost $30. We just read the ingredients, and active ingredient is CBD, found exact ingredients in another off brand ointment online, for only $8. Our healthcare system at its finest. -
-
@
Psionic Roshambo:
I always suggest kneepads with wheels to help with knee pain, then I get slapped
-
-
-
-
-
-
-
