Teensy 4.1 custom board troubles.

sandvoxel

Member
I am currently working on a custom board for the teensy 4.1, but have been having issues getting the main chip to start correctly. The issue I am having is the oscillator for 24 MHz is not being driven by the chip just getting a flat 400 mV on the pads for the oscillator and the 4 blinks from the bootloader that jtag is not responding. The startup sequence seems to be correct as I am getting PMIC_ON_REQ and that is enabling my 3.3V regulator. I am also getting my 1.5~ V out of the cpu on VDD_SOC_IN. I ask that you forgive if I have left off some relevant information but I am quite new to the whole forum thing.

Here is the schematic for the board.
Schematic_Flight Controller_2022-09-15.jpg
 
Hello I have swapped out the caps for some that are known to be 12pF and here is scope view of one of the pads of the oscillator.

Scope.jpg
 
I guess I might have just busted the chip because I cant see why it would not be trying to bump (Not sure the right term) the oscillator. I'm not sure when I would have done that thought because I did not use hot air I used a reflow oven but who knows.
 
I take back what I said about the chip being broken It seems to be an error with not taking high freq into account when designing my pcb. I believe I have just placed my 24MHz way to far away from the chip it is only a centimeter or so but If I had to guess the trace capacitance is interfering with the oscillator. I can see it trying to start the oscillator but it dies off quickly can kind of falls on its face.

My troublesome circuit.
MyOscillator.jpg

Proper Teensy Oscillator (The voltage scale us also different the signal is much stronger)
TeensyOscillator.jpg

If there is any way to save this circuit I would love to know but if I it is beyond salvaging then so be it.

Here is a snip it of the PCB layout to showcase oscillators distance from cpu.
PCBOscillator.png
 
1 cm distance is probably ok.

But if the ground plane between the crystal and chip has substantial gaps due to traces or too many vias, it could fail. Good ground between the crystal and chip is essential.
 
1 cm distance is probably ok.

But if the ground plane between the crystal and chip has substantial gaps due to traces or too many vias, it could fail. Good ground between the crystal and chip is essential.

Ok so this is a error on my part I made my pcb so that both pads for each side of the resonator are connected together. So the crystal was connected wrong I did some precarious soldering to get the crystal to hover over the board only connecting the 2 crystal pin and now I have a strong clock that looks like one off of a factory teensy. But now I am receiving a 10 blink code. Any pointers on this as I have seen "Serial Number Already Assigned" on the website but I'm not sure the troubleshooting steps after this.

The flash chip I have W25Q128JVSIQ currently and maybe that's a issue because it ends with Q?
 

Attachments

  • PCBGround.jpg
    PCBGround.jpg
    160.7 KB · Views: 34
Last edited:
Here is where I am at as of right now. The chip appears to start up correctly but gets begins flashing with boot code 10 (10 blinks) and nothing else happening. I have removed the incorrect flash chip but I still only get code 10 is that something that needs the Flash chip on the board? I saw that on the documentation It is labeled as "Serial Number Already Assigned" but that seemed to be a bit vague on what that means or where I would begin troubleshooting. With the flash chip removed I would think I would get a code 3 but I have not received that with the flash chip unpopulated. I did attempt to find a thread on the forum about this issue but if there is one it is buried under other threads. Most other codes have a clear problem area to look at but I'm at a bit of a loss on where to head next.
 
The "M" flash chip is required. "Q" version will not work.

I don't understand why that would give 10 blinks. Something else is likely wrong, but I'm afraid I just can't even guess from the info given.
 
Thanks for the response I believe it is because I messed up and changed the boot-loader to another one and unfortunately the original I used appears to be dead as I cannot get it to send out DC_SWITCH. I’m going to try to assemble another board with a fresh cpu and boot loader and see how that goes.
 
Back
Top