[time-nuts] Looking for off-the-shelf device to timestamp multiple PPS inputs

Kevin Rosenberg kevin at rosenberg.net
Wed Sep 28 16:22:48 UTC 2011

Hello esteemed time nuts,

I'm looking for device with an external clock input as well as inputs
for multiple PPS signals that outputs timestamped messages for each
input pulse via RS232. The initial need will be for two PPS inputs.
I've considered a number of options, but I wonder if there a more
prepackaged solution available since the device will for someone less
technical than myself.

A bit more about the device. It needs to be clocked with an external
reference. Ideally, taking 10 Mhz sine from the PRS-10, but I can
easily enough convert the sine wave that to a square wave.  Also, I
can use a TAPR ClockBox to convert to other reference frequencies.

The purpose of the device will be to measure PPS periods for at least
two oscillators over a course of a few months, sending the data via
RS-232 to a logging computer for each input PPS input. The minimum the
serial record will need to have is a time stamp of seconds.nanosecons
and a label of which PPS signal triggered.

About options I've considered,

1) A Soekris 4501 clocked by the PRS-10 to a ClockBox, and then using
the TMR1IN and a GPIO input to time stamp the PPS signal. However,
that'll require one 4501 for each PPS signal, so scaling effort is

2) Writing a firmware for an AVR XMEGA that is clocked from the PRS-10
and uses external interrupts and timers to output RS-232
messages. Certainly less expensive than #1 option, but takes more
time programming.

3) Using some other micro-controller with external clock and times,
perhaps an ARM.

4) Using a dedicated time counter for each PPS. However, reading the
manual for my PM6680/B as well as a Agilent 53230A, that won't work
because in standard gating mode, some time would be missed between PPS
signals. While the 53230A has a continuous mode, that would require
the polling computer to accurately keep time of the requests unless
the 53230A is collecting data under its 1,000,000 sample limit (only
~11 days at a PPS). Also, at $3500 per PPS input, a 53230A is the
most expensive to scale..

5) Using a FreeBSD or Linux box with multiple RS-232 inputs and using
their PPS API.  However, outside of some ARM boards, I haven't found
any boxes that can be clocked by a external reference. And, I haven't
enough experience running FreeBSD or Linux on ARM to know how well
this will work for me.

6) I'm quite familar with using something like a Thunderbolt and its
PPS to make a stable clock with NTP which could be used with an Intel
FreeBSD or Linux box in #5. However, rather than NTP, the device
should be clocked only by the external reference clock. The goal is to
see how the input PPS signals deviate from the reference clock.

7) Use a box like in #5, and capture the PPS from the PRS-10 as well
and using that to scale the other PPS inputs that arrive around the
same time.

I've done web searches, found some data loggers, but I haven't found
something that can run continuously for months collecting data and
meet the other needs. I'd really like to find a more prepacked
solution for the user of this device. I'm surprised by my inability
find such a device for this seemingly straightforward task.

Thanks for considering the issues. I feel I'm overlooking some perfect
device that exists outside of just my imagination.


More information about the time-nuts mailing list