Skip to content

Iltarastit Postintaival

Sprint-like course in Helsinki central park. 23 legs over 7 km is about 300 m per leg on average.

Three-four smaller and larger mistakes.

  • after a good start at #1 I zigzag towards #2 via the wrong hill. maybe +1min
  • down from #3 I take the wrong path to the bigger road. maybe +1min
  • towards #14 I should have taken the same path as on the #9-#10 leg but instead make a loop to the right. +2min at least
  • Out of #18 I first have the map 180-degrees wrong and take a few steps south before realizing I should go north. Then I read the #6-#5 line for a while before realizing I should go to #19. Crossing lines like this are always dangerous! Maybe +2min and the worst split.


Itärastit Landbo

Mostly walking in quite slow terrain - with at least 3 bigger mistakes...


Meinberg Lantime M300

Following Dave's advice here: "Don't turn it on - take it apart!"

Meinberg Lantime M300 - Multi Reference NTP server. This server works as stratum-1 using a number of different input signals such as GPS, 1PPS, 10MHz, and IRIG/Timecode.

The loop time-constants seem to be very long, since the ADEV plot does not change much (at least not immediately) when connecting or disconnecting e.g. the 1PPS or 10MHz inputs. If I would be more patient I would measure it free-running and with all the different stratum-1 input signals for a few days each - that would maybe show some interesting things about how the internal OCXO is locked to the reference inputs. My assumption would be that IRIG would produce the worst ADEV, GPS second worst, 1PPS quite good, and the 10MHz input should produce an excellent ADEV with the shortest lock time-constant.

Microsemi/Symmetricom 3120A noise floor

This was measured by taking a H-maser 10MHz signal (output from a frequency distribution amplifier) and splitting it in two for the 3120A REF and DUT channels with a ZFSC-2-1+ passive splitter.

The first measurement is with 50 Hz bandwidth and a duration of 5 minutes. The second measurement is at 5 Hz bandwidth and ran overnight for 12 hours.

Here's the ADEV plot with some typical performance-curves added.


Data from:

SRS FS710 noise measurement

Update: changing the main electrolytic capacitors as well as the tantalum caps around the 7805/7905 voltage regulators did not help. There is still about 600 mVpp of ripple on the +/- 13.9 VDC inputs to the regulators.

Here's a noise measurement on an old second-hand SRS FS710 distribution amplifier. I wonder if the excessive noise at 50 Hz and harmonics is within spec or a result of old age?

Measurement setup: 10MHz signal to a Mini-Circuits ZFSC-2-1+ splitter, one branch to 3120A REF-input, the other branch to the FS710 input, and FS710 output1 to the 3120A DUT input. Noise floor was measured with the same setup but bypassing the FS710.

White Rabbit Switch noise measurement

I measured the CLK2 10 MHz RF-output of two White Rabbit switches with a Microsemi 3120A. The results are similar to previous results with ADEV(1s)=1.5e-11.

When phase-locked to an external reference clock the long-term stability is obviously much better than when free-running, but for small tau, or high offset-frequency in the phase noise plot, the free running oscillator is actually better.

Iltarastit Solvalla

Only bigger mistake towards the end on #20 crossing the swamp and then correcting right - maybe +1min.
Otherwise quite ok in about +12C and sunshine.


Intel Atom C255Xs

The Atom C255X is becoming a popular processor in not-so-demanding low power boxes.

Here's a pfsense SG-4860 router/firewall using a C2558 (15W) processor and 8 Gb of RAM. It comes with 4 Gb of on-board flash storage, so the external mSATA drive (lower edge, just above the Ethernet connectors) is not required unless you need to store very big pfsense logfiles. No moving parts, passive cooling only. Although pfsense-branded (running FreeBSD + pfsense), under the hood it's probably an ADI Engineering RCC-VE.

Here's a SuperMicro 5018A-MLTN4 with a C2550 (14W) processor in a 1U 19" rack enclosure. SSD drive and 8 Gb of RAM. I suspect the front fan is quite overkill and could be slowed down a lot or completely removed. The built-in graphics produce not-so-great performance under Ubuntu, but with lubuntu-desktop it's usable as a lab server that you need to configure or check only once a week or so.

Phase noise with AllanTools

I've looked at calculating phase-noise spectra with python - to be included in allantools. Here are some first results.


Iltarastit Luukki

No major mistakes. OK speed until the 3rd control - then quite tired legs. About 24th/64.