[Thumbgps-devel] Jitter and latency in a linux system might be a problem. Minimal BoM crosspost.

[Eric S. Raymond]

tz <thomas at mich.com>:
> I would think so to, but I think ESRs original post or one of the
> requirements comments said 100nS.
> 
> It would help if we knew how hard the number was.

I've specified 1ms becauser that's a convenient order of magnitude below the
expected 10ms accuracty of NTP time, and the first application of this defice 
will be to check the accuracy of NTP time.

There's evidence from the ZTI Z050 spec sheet that this is an
achievable target - they quote 1ms timing accuracy.

> I've seen this one with the same description from a few searches:
> 
> http://www.chinazrh.com/wholesale-gps-receiver-usb-dongle-for-computers-netbook-laptop-umpc-p-728.html

I've tracked down their support address and asked these questions:

1. The SkyTraq 6 emits 1PPS. Is this signal brought out to the USB bus
where it can be seen as an emulated DCD or CTS change?

2. What USB-to-serial converter does it use?
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>