[LinuxPPS] ktimer works with the PL-2303 USB to serial converter, but the ppsldisc utility doesn't

[Cirilo Bernardo]

On Sun, Nov 23, 2008 at 1:52 PM, Ron Wright <logiconcepts819 at gmail.com> wrote:
> On Sat, Nov 22, 2008 at 7:03 PM, Cirilo Bernardo <cirilo.bernardo at gmail.com>
> wrote:
>>
>> On Sun, Nov 23, 2008 at 11:24 AM, Ron Wright <logiconcepts819 at gmail.com>
>> wrote:

[snip]

>> Does your USB gadget genuinely support the extra lines?  Some don't
>> and in fact only provide Tx, Rx, GND.
>
> I may not know about the other lines, but I do know for sure that the
> adapter genuinely supports the DCD line, since I tried it out already with
> statserial and gpsd.  Here are outputs from statserial when there is a PPS
> signal and when there is no PPS signal.
>

[snip]

> Notice that the DCD status changes, as seen in the output from statserial.
> Therefore, I have no problem detecting the PPS signals with other programs.
>

[snip]

>>
>> What 'clear file' are you talking about?
>>
>> - Cirilo
>
> The "clear" file I was talking about is the one under the directory
> /sys/class/pps/pps0.  The two files I referred to were "assert" and "clear."
>

OK, so your USB serial adapter has DCD; that's a good start.  The
challenge now is to hack the usbserial module to create a timestamp
and report it via LinuxPPS.  With the patches in LinuxPPS so far, PPS
is not checked on a USB serial device.

The "clear" is the time that the PPS signal (actually the DCD pin) was
'cleared'; whether you use 'assert' or 'clear' really depends on the
nature of the timing edge of the PPS (high-to-low or low-to-high
transition). Typically only *one* edge of the PPS pulse is guaranteed
to give a decent timing reference; the width of the pulse is typically
not guaranteed, is variable (mostly due to temperature), and only a
nominal width is quoted in the data sheets.

- Cirilo