[time-nuts] Lowest Power NTP Server

[Bob kb8tq]

Hi

I’m sitting here getting things set up to start digging into the whole wired / wireless / whatever
side of this. The “real” application will be purely WiFi. More or less, the “real” budget is an 
equal jitter from the server, the client, and the router. If that all comes in under 10 ms, this
is wonderful. If it comes out to 30 ms 90% of the time that still works. 

====

If indeed “something else” turns out to be the answer, going back to the idea of broadcast NTP
would seem to be more attractive than doing something from scratch and plumbing it into a 
variety of operating systems. 

====

Sitting here right now with a quick lash up on an RPi 4, I’m getting jitter numbers in the single 
digit ms range. I need to get a couple of issues worked out on the local lan before I really trust 
the data. At the very least it suggests that an RPi 4 client can do pretty well. 

Bob

> On Dec 2, 2019, at 3:48 PM, David Kern <david at mju.io> wrote:
> 
> Bob,
> 
> Will all the devices and the NTP server be connecting to an already existing Wifi AP - or is there a possibility that the Wifi AP itself provides the NTP service?
> 
> When I did more testing, I discovered that it isn't just ESP32s that do weird things on WiFi - when pinging from something on wifi to wired, I got consistent fast pings.  But when pinging from something on wifi to something else on wifi then I got the strange latency issues where it would spike and end up all over the place.  I'm a bit stumped as to why at the moment, maybe it is something weird with my network.
> 
> But at any rate - if the AP in your setup could handle timekeeping, then it eliminates one of the hops and part of the jitter.
> 
> There are a bunch of off-the-shelf wifi routers that can be flashed with a more capable linux environment and are already DC powered.
> 
> (As an aside, there was mention of adding time distribution over wifi by hacking the protocol a bit.  Does anyone recall how this would work - like doing something with the beacon frames or sending some special frame on the channel?)
> 
> -David (AD7WZ)
> 
> 
> 
> 
> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
> On Monday, December 2, 2019 11:51 AM, Bob kb8tq <kb8tq at n1k.org> wrote:
> 
>> Hi
>> 
>> Wired is out for this particular setup. It needs to be some sort of wireless. It all would
>> be much more simple with wires.
>> 
>> Bob
>> 
>>> On Dec 2, 2019, at 12:46 PM, Robert LaJeunesse lajeunesse at mail.com wrote:
>>> If wired Ethernet seems to be the way to go consider the Orange Pi Zero - about the cheapest wired Ethernet board available that runs Linux. Ethernet is via on-chip MAC and phy, so no USB path delays. http://www.orangepi.org/orangepizero/
>>> Plenty of support exists on the web, for example: https://lucsmall.com/2017/01/19/beginners-guide-to-the-orange-pi-zero/
>>> Bob L.
>>> 
>>>> Sent: Monday, December 02, 2019 at 9:56 AM
>>>> From: "Tim Shoppa" tshoppa at gmail.com
>>>> To: "Discussion of precise time and frequency measurement" time-nuts at lists.febo.com
>>>> Subject: Re: [time-nuts] Lowest Power NTP Server
>>>> Bob, I find that 2.4GHz Wi-Fi UDP latency with ESP8266 will frequently be
>>>> tens of milliseconds and is never/rarely consistent.
>>>> There are specialized non-WiFi 2.4GHz systems for time distribution that
>>>> are far more consistent (possibly even at the tens of microseconds). I
>>>> think several years ago on this list, we were talking about tricking
>>>> commodity WiFi chipsets into doing these but haven't seen anything as of
>>>> late.
>>>> Tim N3QE
>>>> On Mon, Dec 2, 2019 at 8:02 AM Bob kb8tq kb8tq at n1k.org wrote:
>>>> 
>>>>> Hi
>>>>> Indeed, if you get up into the “many tens” of ms, that rules it out in my
>>>>> application.
>>>>> A consistent 90 ms would be ok, you could compensate for that. Random
>>>>> flopping
>>>>> from 4 to 90 … not so much.
>>>>> It seems like that sort of jitter would get in the way of a lot of things.
>>>>> I guess that just
>>>>> shows how little I know about a lot of things :)
>>>>> Bob
>>> 
>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>> and follow the instructions there.
>> 
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.