[time-nuts] measuring multiple gps modules

[folkert]

Hi,

I thought I understood just enough from electronics to be able to these
kind of experiments but apparently not (3 damaged rpi's and gps
modules).

I thought that if signal levels are in range of the device that will
read them and all grounds are connected, that you could safely for
example connect the PPS-out of a gps to both an rpi as well as a micro
controller. I did this with 3 devices (3 rpis + gpses connected to a
"teensy 4.0" microcontroller). The Teensy 4.0 was powered from one of
the rpis.
This resulted in one rpi+gps directly not producing any pps and the
others giving erratic pulses (skipping and enormous jitter).
I'm puzzled how this could happen.

Anyway, the next project is connecting an rpi to a leobodnar gpsdo (
http://www.leobodnar.com/shop/index.php?main_page=product_info&products_id=234
). As you can imagine, I'm a bit weak in the knees now (maybe I should
stick to software) after destroying 6 devices.
So my question now is, if I:
- configure the leobodnar gpsdo to emit 10MHz
- connect the bnc to a picdiv running at 3.3v
  ( http://leapsecond.com/pic/picdiv.htm )
- connect the out-pin & gnd to an rpi
- take care of electrostatic discharges while setting this up

Will I not destroy anything?

(the idea I have: let the 2 signals (reference & dut) be timestamped by
an rpi (pps-gpio kernel module) and then draw allan deviation plots)


regards