Closed kb1lqd closed 7 years ago
I ran the as last saved (with COM port updated) BSL uploader I wrote and it worked almost:
I'm using the unit that I just took off the external 18pF LFXTAL1 load capacitors (REV B1 SN#3) but that shouldn't affect this. It should use the REFOSC...
Hmmm... It looks like the unit is now PORT 6! I did update the programming on the FTDI chip prior to trying the BSL to make sure the CBUS GPIO's were configured correctly...
Trying COM PORT 6
It's really really slow even at 115200 baud and still failed checksum. Is my checksum calculation ok?
I tried quickly again with the other REV B1 on COM 3 and the unit is still slow:
REV A used fly-wired connections, this is REV B. I bet I'm an idiot and didn't check the configuration of BSL entry CBUS GPIO's...
Datasheet for SSOP package used on REV A:
Looks like I'm using pins 7 (RST) and 16 (TEST) which are CBUS 2 and 3 respectively.
My old code used 0x04 and 0x08 bitmasks which indicate the the MSB's are CBUS 2 and 3.
REV B1 moved to:
RST = CBUS 0 (0x01) TEST = CBUS 1(0x02)
Due to the CBUS pin change the bitmask for CBUS in the FTDI driver needed updating.
RST = CBUS 0 (Bitmask = 0x01) TEST = CBUS 1(Bitmask = 0x02)
class FtdiD2xxCbusControlObject(object):
def __init__(self):
self.BITMASK_IO_OUTPUTS = 0xF0
self.BITMASK_IO_OUTPUTS_RESET = 0x40
self.BITMASK_IO_INPUTS = 0x00
self.BITMASK_RST = 0x01
self.BITMASK_TEST = 0x02
self.DELAY_TIME = 1
self.handle = ''
self.ftd2xxDll = ''
I loaded an old code on with the BSL to REV B1 and to my inspection it looks correct. I can play with LED's to absolutely verify but I feel confident due to the above debugging and known good scripts verifying the matching checksum's.
Closing.
Verify that REV C boards which contain the FTDI traces as tested with flywire correctly load bootloader of TI and program board.