Time Stability

Discussion to talk about software related topics only.
SeeCwriter
Posts: 605
Joined: Mon May 12, 2008 10:55 am

Re: Time Stability

Post by SeeCwriter »

On bootup the value of the TimeOffset was 1495649838.
After an NTP resync an hour later, the value was 1495649836.

I am using a MOD54415. The RTC is stable and not drifting. For the NANO we have changed the NTP resync rate to every 30-minutes to keep the system time within reason. And even then the system clock is typically 4-7 seconds fast.

Correction: We reset the system clock from the RTC, not the NTP, every 30-minutes.
SeeCwriter
Posts: 605
Joined: Mon May 12, 2008 10:55 am

Re: Time Stability

Post by SeeCwriter »

pbreed wrote:Actually is this a MOD54415 or MOD54417, for this they are different internally, and it might matter....
Also what is the board revision, we recently made a board hardware change to reduce EMI...


Paul
The board is rev 1.9.
User avatar
pbreed
Posts: 1080
Joined: Thu Apr 24, 2008 3:58 pm

Re: Time Stability

Post by pbreed »

OK on a netburner carrier board my 54415 is loosing 0 seconds started 13,500 seconds ago...
Also rev 1.9.

One of the things about clocks is that if the power supply is noisy or marginal, or you are in big EMI fields it will cause extra transitions in the crystal circuit.
This will make the clock run fast and is also a symptom that your on the very edge of having power supply noise reliability issues.

Can you run a test using only the MOD54415, and NTP offset on one of our carrier boards to eliminate this kind of potential nose as an issue.
User avatar
pbreed
Posts: 1080
Joined: Thu Apr 24, 2008 3:58 pm

Re: Time Stability

Post by pbreed »

Also do you connect to the clk line from the module with any external hardware?
Could that hardware be driving noise into the clock output pin?
SeeCwriter
Posts: 605
Joined: Mon May 12, 2008 10:55 am

Re: Time Stability

Post by SeeCwriter »

Nothing is connected to the clk output of the module. I will try some tests on the NB carrier board as you suggest and get back to you.
User avatar
dciliske
Posts: 623
Joined: Mon Feb 06, 2012 9:37 am
Location: San Diego, CA
Contact:

Re: Time Stability

Post by dciliske »

Out of curiosity, do you get any network interruption issues on large frames (~1500 bytes)? Basically, if it's an issue of clock noise on the input, I'd expect you to see some frame loss on maximal frames.

-Dan
Dan Ciliske
Project Engineer
Netburner, Inc
User avatar
pbreed
Posts: 1080
Joined: Thu Apr 24, 2008 3:58 pm

Re: Time Stability

Post by pbreed »

Ok its been running for 22 hours and according to NTP has not gained or lost a second yet.
This is a rev 1.9 MOD54415 running 2.8.3
SeeCwriter
Posts: 605
Joined: Mon May 12, 2008 10:55 am

Re: Time Stability

Post by SeeCwriter »

My module has been running for 22 hours since its last NTP resync and it's 2 minutes and about 20 seconds fast. In 2 hours it will resync.

We are also going to do some power supply testing today. I found out we are using a new switcher on the board to provide power. So that will be looked at.
User avatar
pbreed
Posts: 1080
Joined: Thu Apr 24, 2008 3:58 pm

Re: Time Stability

Post by pbreed »

Is your test in your product, or in a netburner carrier board?
SeeCwriter
Posts: 605
Joined: Mon May 12, 2008 10:55 am

Re: Time Stability

Post by SeeCwriter »

The module is on our board. I haven't found the NB carrier yet.
Post Reply