[GPS_Standard] Problems PN2222A Inverter missing occasional 1PPS

[Michal]

Hi Dave and all,

When dealing with any common emitter NPN transistor working as an 
inverting stage we have to keep in mind that on-time is much faster than 
off-time. I don't know how sensitive is PIC input for glitches but when 
we want to get fast rising edge we need relatively small collector 
resistor.  I'm suggesting 330-470ohm.
Relatively small base resistor (4,7k) is also not a good solution in 
this case because transistors goes to highly saturated switching further 
delaying off-time. It is better to use larger value with small speed-up 
capacitor.
I have made few simulations with SPICE-Based Analog Simulation Program. 
Enclosed please find the PDF file with the results.
I highly recommend using this or similar simulation software, it can 
save your time and solve many mysterious things.  The Tina-Ti can be 
FREE downloaded from  TI page: http://www.ti.com/tool/tina-ti

As per extra counts reported by Bob there may be another explanation of 
the problem.
I have met with similar problem in a test circuit when the clock signals 
was fed by the 7cm long wire and it could sometimes pick spikes from an 
air (during switching AC loads).  Twisting it with ground wire was a 
solution.

Best 73,
Michal
sp2iqw



W dniu 2013-11-18 05:07, Dave Platt pisze:
> On 11/17/2013 04:47 PM, Bob Stewart wrote:
>> Michal and all,
>>
>> I'm still getting the occasional rare (maybe 5 or 10 a day?) glitch in my 1PPS pulse.  The symptom is a 2D00 instead of a 9680 in the oscillator count value.  This was puzzling until I realized that it's simply 2 times 0x9680, which means that from time to time I am not getting a 1PPS pulse.  As mentioned, the circuit is a simple inverting amp with 1800 ohms in the base lead to the 1PPS signal from the Adafruit, and 1800 ohms in the collector lead.  So, should I use a different transistor, or different values for the two resistors?  I want to keep this as simple as possible, so I would prefer not to have to add a chip, even if that would be the best solution.  I am not a hardware person, and just blindly designed the amp for low current.
> On my GPSDO "shmoo" board (regulators, MAX232 serial interface,
> and the "glue" connections between the GPS and the controller
> board) I use a two-transistor inverter chain to buffer the
> Oncore UT+ PPS signal and send it over to the controller.
>
> It uses three 4.7k resistors - one in the base lead of
> the first transistor (the drive from the UT+ PPS line)
> and one in each collector lead.  The base of the second
> transistor is fed directly from the collector of
> the first.
>
> I haven't had any reason to believe I'm missing PPS
> signals... I don't think I've seen any of those
> "modulus is twice what it should be" glitches you
> have seen.
>
> I did use fast switching transistors rather than
> general-purpose one... I had a bag of nice
> JAN-rated NPN fast switches sitting around and
> grabbed a couple of those.
>
>
>
> ______________________________________________________________
> GPS_Standard mailing list
> Home: http://mailman.qth.net/mailman/listinfo/gps_standard
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:GPS_Standard at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>