[time-nuts] NTP/1-PPS/RS232 question

mc235960 mc235960 at gmail.com
Sun Sep 8 13:15:29 EDT 2013


Le 8 sept. 2013 à 17:08, Chris Albertson a écrit :

> You only need ONE inverter gate.

   true
> 
> The purpose is the flip the pulse.  The MAX 232 chip flips the logic
> once which works fine for the data but you need to flip the control
> lines (maybe if the GPS reciever did not already do this for you.)
> 
> Flipping the PPS has the effect of moving the pulse by the length of
> the pulse so one gate in this case will move the pulse by 100MS.

Not true. The pulse is not moved, just inverted. You just need to have the driver looking for it configured to detect the correct edge polarity. +ve or -ve. That's what the flag is for.  
If you have a piece of kit that outputs PPS, the chances are that it is TTL +ve.  Here is how I see it:-

IF we put an inverter between the GPS(other it) and the rs232 chip)

1 +ve TTL pulse on the line to the inverter 
2 -ve TTL pulse  to the max232 transmitter , inversion
3 mark (+ve) pulse from transmitter out on the line to DB9 pin1 (DCD)
4 same over cable to pin 1 of DB9 on a PC
5 same into the max232 receiver in , another inversion
6 -ve TTL from max232 receiver out to the cd_n pin on the UART (ex 16550A)  
  The cd_n line is active LOW
7 uart MSR register bit 7 set to compliment of cd_n 
   uart MSR bit 3 set to one as cd_n input level has changed.
   uart generates an interrupt
8 interrupt wakes up driver
 etc
 etc
    
So the ATOM driver will get a tick on the +ve outgoing edge of the PPS kit ( plus sone nanoseconds delay in the inverters/uart chain) if there is an inverter between the kit and the max232.
Conversely it will get the tick on the -ve outgoing edge if no intervening inverter is used. This means that the tick will be LATE by the length of the pulse.
For example , with a UT+, it will be around 200ms late and you can see this in the ntpq offset.



> 
> On Sat, Sep 7, 2013 at 9:22 PM, Bob Bownes <bownes at gmail.com> wrote:
>> 100ms is an awful lot of 'inverter gates'!
>> 
>> On Sep 7, 2013, at 23:15, Chris Albertson <albertson.chris at gmail.com> wrote:
>> 
>>> With RS232 the data uses negative logic (low is "1") and the control
>>> lines use positive logic (high is "1".)   That is, a control line is
>>> "asserted" when it is at logic 0, a positive voltage.
>>> 
>>> Some one at one time must have thought this made sense.
>>> 
>>> The good news is that a mistake is easy to detect and fix.  If the
>>> pulse is 100MS long and you have it inverted then your time will be
>>> off by 100MS.   When you are designing the board put in some extra
>>> inverter gates
>>> 
>>> On Sat, Sep 7, 2013 at 7:15 PM, Mark Sims <holrum at hotmail.com> wrote:
>>>> I got in those Adafruit GPS boards.  They are a very nice little GPS.   VERY sensitive.  I could get good lock indoors on my (windowless) kitchen floor.  The house is 2 story,  stucco with wire mesh in the stucco.   I could also get lock in a restaurant that had a tin roof.  We were far from any windows.
>>>> The 1PPS signal is normally low and pulses high for 100 milliseconds.  The RS-232 adapter board that I built feeds this into an RS-232 transmitter chip (MAX3232),  so on the interface connector CD will be at +V and pulse down to -V.  Is this what stock NTP likes?
>>>> Also,  I laid out the adapter board so it an accommodate a Trimble Resolution T or a Crius CN06 receiver.  The Crius receiver uses a U-Blox NEO-6M receiver.  They can be had for around $22 at HobbyKing.   They seem to perform even better than the AdaFruit module.  It looks like you will need to bodge a wire onto the TIMEPULSE pin to use it for 1PPS.
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at febo.com
>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>> and follow the instructions there.
>>> 
>>> 
>>> 
>>> --
>>> 
>>> Chris Albertson
>>> Redondo Beach, California
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
> 
> 
> 
> -- 
> 
> Chris Albertson
> Redondo Beach, California
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.



More information about the time-nuts mailing list