Boot loader compatibility
- Thread starter jonr
- Start date
- Status
Constantin
Well-known member
I seriously doubt it. For one, the LC bootloader may not make it off the board without damage and onto the next one. By then it will have endured three solder-melting heating cycles minimum, and that's assuming you don't have to clean up the pads after removal.
Next, the LC uses SWD for programming while the MK20 uses EZPort. There are more connections between the boot loader and the MCU for the MK20 than the LC (five vs. three). I have no idea if they are also very different implementations, protocol-wise.
As best as I can tell, Paul, Robin, and the rest of their workshop use a bed of nails jig to program the boot loaders when they have already been mounted on the PCBs, followed by programming the MCUs. Thus, while the 3.x series and LC series boards may end up using the same boot loader chip, they may continue to run very different boot loading programs (i.e. one for SWD, one for EZ-port).
There are standalone boot loader chips available, which are programmed individually. This has not yet started for the LC series boot loading chip.Presumably, they use one of the jigs out there that allow those chips pins to be broken out into a DIP pattern.
I have no idea if Paul will be able to create a boot loader chip which can contain both bootloaders and auto-detect the right chip to target. However, I would not count on it if the programming protocols continue to be different. The MK20 does support SWD programming, IIRC, so there is the possibility.
Lastly, I wonder if the pin 33 problem could not simply be fixed by implementing the new boot loader chip followed by re-assigning one of the the newly freed pins to take the place of pin 26 that currently is connected to the 'D33' pad. The new boot loader chip would be the hint to the Teensyduino environment what it's dealing with and D33 would be a thing of the past. (Paul likely would shake his head, mutter 'hack!' and come up with a much cleaner solution that fixes the D33 problem without any pin changes).
The downside to this approach is that the pre-MKL02Z versions of the Teensy 3.x series would still have the D33 problem. However, I doubt the problem affects a lot of users and those that are affected can then buy the new version of the board and be done with it.
Next, the LC uses SWD for programming while the MK20 uses EZPort. There are more connections between the boot loader and the MCU for the MK20 than the LC (five vs. three). I have no idea if they are also very different implementations, protocol-wise.
As best as I can tell, Paul, Robin, and the rest of their workshop use a bed of nails jig to program the boot loaders when they have already been mounted on the PCBs, followed by programming the MCUs. Thus, while the 3.x series and LC series boards may end up using the same boot loader chip, they may continue to run very different boot loading programs (i.e. one for SWD, one for EZ-port).
There are standalone boot loader chips available, which are programmed individually. This has not yet started for the LC series boot loading chip.Presumably, they use one of the jigs out there that allow those chips pins to be broken out into a DIP pattern.
I have no idea if Paul will be able to create a boot loader chip which can contain both bootloaders and auto-detect the right chip to target. However, I would not count on it if the programming protocols continue to be different. The MK20 does support SWD programming, IIRC, so there is the possibility.
Lastly, I wonder if the pin 33 problem could not simply be fixed by implementing the new boot loader chip followed by re-assigning one of the the newly freed pins to take the place of pin 26 that currently is connected to the 'D33' pad. The new boot loader chip would be the hint to the Teensyduino environment what it's dealing with and D33 would be a thing of the past. (Paul likely would shake his head, mutter 'hack!' and come up with a much cleaner solution that fixes the D33 problem without any pin changes).
The downside to this approach is that the pre-MKL02Z versions of the Teensy 3.x series would still have the D33 problem. However, I doubt the problem affects a lot of users and those that are affected can then buy the new version of the board and be done with it.
Last edited:
PaulStoffregen
Well-known member
No, that absolutely will not work. Do not sacrifice a LC trying. The L02's shipping now are hard-coded to only work with the MKL26 chip on Teensy-LC.
- Status