Spoke with Matty briefly this morning, he and his family are all having or had symptoms of covid 19, doing ok but working through all outstanding consoles.
I'm sure we all wish him and his family a speedy recovery.
There are a few ways to get into uf2 mode. Are you using the samd21 update payload?Hey mattytrog you gave me some advice a couple months back but I am having a hard time finding the post. Awhile back I did a UF2 update on my Rebug chip and since then Windows has been unable to recognize the UF2 in USB and takes over all the USB data connections as well. I know flashing the updated files would work but is there a trick to flash the UF2 again?
After you replied I just did and I get the same response from Windows with the samd21 update payload. I'm not sending it through tegraRCM because it won't recognize RCM mode because the chip is blocking that too. The last uf2 I wrote launches payload.bin off the root.There are a few ways to get into uf2 mode. Are you using the samd21 update payload?
If the chip is blocking tegrarcmsmash, you have a big problem somewhere. All external usb connections override the chip. Thus any PC usb connection overrides anything the modchip will do.After you replied I just did and I get the same response from Windows with the samd21 update payload. I'm not sending it through tegraRCM because it won't recognize RCM mode because the chip is blocking that too. The last uf2 I wrote launches payload.bin off the root.
I tested the last thing you said about if it's connected to the PC and turned on the modchip will fail to send a payload. And that seems correct while plugged in I powered it off and on and it got stuck at a black screen autoRCM I think. But my PC doesn't show it as the APX but whatever state the chip is in. Because no matter what if the switch is powered on and connected if I click the button 3 times I'll get a new device alert sound followed by unrecognized device message.If the chip is blocking tegrarcmsmash, you have a big problem somewhere. All external usb connections override the chip. Thus any PC usb connection overrides anything the modchip will do.
Any chance of a pic of your wiring?
In all the time I've been messing around with these things, I've never. I repeat never seen a samd21 override PC usb. The chip will always fall too send the payload if usb is connected to pc. If this isn't the case, you have a fault.
No problem. I just want to help sort this.I tested the last thing you said about if it's connected to the PC and turned on the modchip will fail to send a payload. And that seems correct while plugged in I powered it off and on and it got stuck at a black screen autoRCM I think. But my PC doesn't show it as the APX but whatever state the chip is in. Because no matter what if the switch is powered on and connected if I click the button 3 times I'll get a new device alert sound followed by unrecognized device message.
I'll go digging for pictures of my chip install if I don't have them I might crack this open. Because I thought I recalled you saying something back then about needing to desolder it in order to reflash it again for some reason. I appreciate the help.
yeah I had a similar issue a few pages back, I did eventually get it sorted, I had used some kynar wire I have used for lots of mod chips etc and never had an issue, but after installing the mod in my own and using a different type it worked fine, so I went back and replaced the ground and 3V wires in my friends switch and now they both work perfectlyNo problem. I just want to help sort this.
I've done all my chips in exactly the same way. Same gnd point, same wire, same length of wire etc and have never had a problem. I run Windows, Unix and even an amiga box, amongst the 5 billion raspberry fookin pi's I seem to have amassed... And usb / uf2 has worked every single time.
My gut feeling is that you have a dry joint or your GND isn't soldered quite right.
Grab some pics, sir. Let's take a look. Don't worry
Thanks for the response I need to find my screwdriver set to get it apart again but I found this picture if it helps. I googled a bit and I saw mention about usb descriptor strings being a thing in some UF2 github I was wondering if there is a trick where i need to temporality disconnect the chip. It responds to being clicked I can put it in different blinking states but the PC still won't recognize RCM or APX device. It's either the failed USB descriptor thing or if I click it Windows makes another acknowledgment sound but neither RCM mode nor UF2 mode with the disk drive will pop up no matter what I've done/ But I'll take it apart shortly to get a look at the actual connections.No problem. I just want to help sort this.
I've done all my chips in exactly the same way. Same gnd point, same wire, same length of wire etc and have never had a problem. I run Windows, Unix and even an amiga box, amongst the 5 billion raspberry fookin pi's I seem to have amassed... And usb / uf2 has worked every single time.
My gut feeling is that you have a dry joint or your GND isn't soldered quite right.
Grab some pics, sir. Let's take a look. Don't worry
They are long mate. Too long I'd say.Thanks for the response I need to find my screwdriver set to get it apart again but I found this picture if it helps. I googled a bit and I saw mention about usb descriptor strings being a thing in some UF2 github I was wondering if there is a trick where i need to temporality disconnect the chip. It responds to being clicked I can put it in different blinking states but the PC still won't recognize RCM or APX device. It's either the failed USB descriptor thing or if I click it Windows makes another acknowledgment sound but neither RCM mode nor UF2 mode with the disk drive will pop up no matter what I've done/ But I'll take it apart shortly to get a look at the actual connections.
They are long mate. Too long I'd say.
Mod the shield and shorten the cables. Your reset button would probably be activating randomly. Best guess it is rammed too tightly between the shield and the cover, making it too easy to press.
Any warranty on patched units is long expired, so you can freely mod the shield.
@mattytrog Should have read the post I mentioned you in, it is relevant to the current issue.
Short version: I just experienced exactly this developing the DMMC. Wire type has a massive impact on USB signal integrity, mostly between the modchip and the Switch, but in really bad cases like mine, it can passively take down the Switch's USB functions even if the modchip is off or has USB disabled. I was using solid core 30AWG and it was nothing but headaches.
@TheZander
Find better wires and make them as short as you can. Stranded, pure copper wire is best, at 26 AWG or better. Many cheap wires look like copper, but are actually just aluminum coated with copper - if the bare wire doesn't really keep it's new shape when you bend it, it's likely aluminum core. Both wires should be the same length. If the problem persists, you can add 20 ohm resistors to each wire as close to the Switch motherboard pads as possible, which should stop any signal reflection.
Which post sir? I didn't see! I'm missing bloody everything just lately.@mattytrog Should have read the post I mentioned you in, it is relevant to the current issue.
Short version: I just experienced exactly this developing the DMMC. Wire type has a massive impact on USB signal integrity, mostly between the modchip and the Switch, but in really bad cases like mine, it can passively take down the Switch's USB functions even if the modchip is off or has USB disabled. I was using solid core 30AWG and it was nothing but headaches.
Find better wires and make them as short as you can. Stranded, pure copper wire is best, at 26 AWG or better. Many cheap wires look like copper, but are actually just aluminum coated with copper - if the bare wire doesn't really keep it's new shape when you bend it, it's likely aluminum core. Both wires should be the same length. If the problem persists, you can add 20 ohm resistors to each wire as close to the Switch motherboard pads as possible, which should stop any signal reflection.
--------------------- MERGED ---------------------------
Which post sir? I didn't see! I'm missing bloody everything just lately.
Signal problems and standing waves occurred to me too.
That means the wires are long enough to have interference go along them.
We normally use ferrite chokes to choke off any common mode current, but if the wires are short enough, nothing can radiate up them as the impedance would be too high for the given frequency.
If you are getting problems because of interference, it isn't actually high resistance causing it. It's high reactance. Either inductive or capacitive.
But I digress.
Like @MatinatorX says, a couple of resistors would help. Or a couple of 10pf caps going to gnd maybe. But that may cause more problems than it solves.
Regarding cable, if they are aluminium cored cable (God knows how you would solder to it, unless it has a copper "veneer" on it - we used it in college for training.) that is bad.
Short wires. Modify the shield. Good quality kynar wire. I used to use awg40 even with no issues. So it isn't the CSA of the conductors, it's the length.
Keep us posted.
Use something from the Panasonic exc24 range. 2nH is the rated inductance on them. They work fine for the pi3usb multiplexer/switch in the erm... Switch! (See what I did there ). You will probably want something with higher current capacity though.Just teasing! Most of it got struckthrough anyways as I slowly figured things out. It'd be my last post on the DMMC thread if you still want to read it.
Would you happen to know of a good SMD common-mode choke to use? Not entirely sure what inductance I should use for USB lines. I wish I knew the inductance of the one on the Switch mobo. I'm already gonna stuff one of these on the lines, it includes the termination resistors as well as some caps to ground, and a bunch of diodes for ESD: https://datasheet.lcsc.com/szlcsc/1809041135_STMicroelectronics-USBUF01W6_C123726.pdf