[ntp:questions] Fwd: Re: NetBSD GPS/PPS using 4.2.6p3

[A C]

On 8/23/2011 08:49, Terje Mathisen wrote:
> A C wrote:
>> The duty cycle is about 10% (100ms on 900ms off). I'll try tossing in
>
> This is pretty much perfect.
>
>> the pulse stretcher and/or inverter just after the GPS receiver and see
>> what happens.
>
> Won't help probably. :-(

At the very least I'll toss an inverter in there and see if the pulse is 
just upside down.  The timing I did quickly by eye and stopwatch (I have 
an LED attached to the PPS pin) but I don't remember if I had to wire 
the LED to V+ or ground to have it stay off for most of the time.  If 
it's wired to V+ then the pulse is inverted and have to try the inverter.

If that doesn't work then I've run out of possibilities other than very 
broken code somewhere.  I still don't understand why the debugging code 
pps-api supplied in ntpd's own sources can find this pulse via the 
kernel but ntpd can't.