[time-nuts] Serial port server .. any interest in a write up on using ?

[Bob Camp]

Hi

What ever degradation the serial stream sees on the LAN, the resulting NTP
output will see once it's on the same LAN. It's unlikely you will see more
than a 2:1 net degradation no matter what is going on. The flywheel in the
NTP algorithm will likely help you in this case to actually improve things a
bit.

Bob

-----Original Message-----
From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] On
Behalf Of Chris Albertson
Sent: Wednesday, May 23, 2012 3:13 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Serial port server .. any interest in a write up on
using ?

On Wed, May 23, 2012 at 9:08 AM, Bob Camp <lists at rtty.us> wrote:
> Hi
>
> If the timing involved is NTP, I'm not so sure that a normal home lan with
> gigabit switches would be a problem. You can indeed saturate the poor
thing.
> Unless you have a very unusual system, it is unlikely you will saturate it
> for very long or saturate it very often. NTP is pretty tolerant of the
> occasional burp of a few ms.

Depends on your accuracy goals.  Of course it would work to some
degree.   But we are talking several orders of degradation if the PPS
second is carried over even a dedicated either net that uses only a
crossover cable.   The problem is not so much the network but the
buffer in the ethernet interface.   This is the same thing that
happens when petiole try and send the PPS over a serial to USB cable.
 Those cables typically cost you two orders of magnitude timing error.

The reason the direct connected PPS is so good is because of the
simple hardware design and the even simpler software interrupt
handler.   There are only four of five lines of code that get
executed.  It is so simple.   Both Either net and USB are "packetized"
where the data arrive in chunks and the entire chunk has to get inside
the computer before it can be looked at.   While PPS to a serial port
is just an edge triggered logic gate that forces an interrupt.
Again you can do it but you take a 10E2 or 10E3 "hit".

If you do need to send a PPS over Ethernet cable you can do it, but
don't use networking.  Cat-5 cable has "extra" pairs that ethernet
does not use.  Those can be used to send a balanced, differential
signal down the wire.  A pair of RS-422 transceiver chips and
terminating resistors would work well for this.   (Don't try to send
an unbalanced TTL level pulse down 100+ feet of Cat-5 cable, it works
very poorly.)

This is all easy to test.  All you need is uSec level test gear.  You
don't need hyper expensive counters.  Compare the transmitted PPS to
your Thunderbolt's PPS or simply look at NTP's log files.
-- 

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.