Michael11

Member Since: May 27, 2009

Country: United States

  • Product GPS-08725 | about 5 years ago

    Is the Blue LED off when you connect the USB cable to the KinetaMap? If not, it might be broken.
    I just looked the code I’m running, which looks just like the SparkFun code as far as the LEDs at boot up are concerned. The sequence should look like this:
    Power Applied
    - Both Blue and Green On
    Initializes serial ports, SD card, and I/O pins
    - Both LEDs Off
    Initialize Interrupts
    - Green On
    Parse the configuration file
    - Green Off and Blue On
    Configure device based on configuration settings
    - Both LEDs Off
    After that there will be some blinking based on what is logged. The SparkFun blink code has some bugs, but it should blink.
    If you delete your config file, does a new default file get written? If so, we know it runs to that point.
    -Michael

  • Product GPS-08725 | about 5 years ago

    1) A few more hours and still no lock-ups. I fixed a lock-up like behavior due to settingsString being to short, maybe your change in buffer size moved settingsString and masked the real cause of the lock-up.
    2) The problem is a bug in the checksum calculation. Any sentence with a checksum containing a character ‘A’..‘F’ was miscalculated.
    3) Still no clue. I’m going to ignore this one. It’s a non issue when writing 50 samples a second or less.
    Is there somewhere that I can send all of my fixes so they get considered for the next release?

  • Product GPS-08725 | about 5 years ago

    Chizat:
    I did have one quick question: Is it possible to use a USB charger for power only and continue logging?

    Remove the if clause that starts
    if(IOPIN0 & (1<<23))
    from the SparkFun code.
    To log with USB power do the following:
    1. Start with KinetaMap off and unconnected.
    2. Turn on KinetaMap.
    3. Plug in USB cord.
    To access KinetaMap SD card do the following:
    1. Start with KinetaMap off and connected to the USB port.
    2. Turn on KinetaMap.
    Warning: This is only an educated guess. I haven’t tried it. If it doesn’t work, you need an SD Card writer to get working software on the KinetaMap.

  • Product GPS-08725 | about 5 years ago

    Thanks for the prompt response.
    1) I’ve collected a few hours of data with the buffer at 512 bytes and I haven’t seen a lock-up. I’ll let you know how things go.
    2) I’ll disable all of the checking as see what happens. If corrupt sentences are being thrown out, I’ll try to fix that. I already fixed one scenario.
    3) Now only files 000 through 002 seem to have slow writes. The slowness can be seen by looking at the time it takes to complete an sd_raw_sync call (that forces the writes).
    If there’s ever another revision of KinetaMap hardware, it would be nice to give the processor and SD card a little extra life after the switch is put in the off position, then all of the syncing can be skipped.
    It looks like I have a lot to investigate. I’ll get back to you and even send you my results if you want them.

  • Product GPS-08725 | about 5 years ago

    Question 1:
    Firmware version 1.2 includes the following comment in the change log: “Lowered MAX_BUFFER_SIZE in KinetaMapPins to 400 to prevent ‘lock-ups’ when trying to save too much data”.
    What do the lock-ups look like? Everything seems normal with a few hours at 512 bytes.
    Question 2:
    Commands are sent to the GPS to report GGA/RMC sentences every second, but they aren’t logged at a 1 per second rate. I fixed some errors in the UART ISR that could result in data corruption, but I still see seconds without a sentence. Does the EM408 really report one sentence per second?
    Question 3:
    I’ve noticed that the maximum rate that I can log data to the SD card seems to depend on the file name. My guess is that has to do with the location of the file in the FAT. Initially just the 000 files couldn’t log 100 accelerometer samples per second. Now files 000 - *005 can’t keep up with the 100 sample per second.
    Has anybody else seen this problem? Is it expected?

  • Product GPS-08725 | about 5 years ago

    I’ve had about a week to hack at the KinetaMap software. In that time I’ve come up with a few bug fixes/enhancements and a few more questions.
    Most of my changes speed up the data logging and prevent log file corruption. I have permission to release the source if there is any interest.
    Not so coincidentally, my questions are all along the lines of data logging. I’m submitting them as responses to this comment to keep within the 1000 character limit.

  • Product GPS-08725 | about 5 years ago

    The new package seems to build fine, but FW.SFE is different from the one in the older package. Is the new FW.SFE the one that KinetaMap actually ships with?

  • Product GPS-08725 | about 5 years ago

    Has anybody tried to compile the V1.2 Firmware?
    I get an error indicating that “TIMER_FREQ” is undefined in main.c. Not only is it undefined in main.c, but it doesn’t seem to be defined in any file. main.c is the only file containing the string “TIMER_FREQ”.
    It’s quite possible that I have a set-up issue, but I tried compiling Francesco’s code and that seemed to build okay.

No public wish lists :(