Performance Data for my 4 MHz Collection

Graph showing rates in Seconds Per Year for each of my 4 MHz watches

graph

Table showing deviation from Standard Time in seconds for each of my 4 MHz watches

Date H102 H103 H103G Crystron Exceed Omega
26th Feb. 2016 -0.1 -0.1 0 -0.5   0.4
15th Mar. 2016           -0.5
17th Mar. 2016 3.5 0.2 3 -0.1    
11th Apr. 2016         -5  
14th Apr. 2016         -5.8  
18th Apr. 2016         -6.6  
27th Apr. 2016 11.2 0.6 8.3   -8.8 -1.2
3rd May 2016         -8.7  
6th May 2016       1.3   -1.1
10th May 2016         -8.6  
18th May 2016         -8.4  
24th May 2016 16.3 1 10.9 2 -8.2 -0.7
1st Jun. 2016 17.8 1.2 11.6 2.3 -8.1 -0.1
10th Jun. 2016 19.5 1.3 12.4 2.3 -7.9 -0.2
16th Jun. 2016 20.5 1.3 12.8 2.7 -7.9 0
22nd Jun. 2016 21.5 1.5 13.2 3.1 -7.8 0
29th Jun. 2016 22.7 1.5 13.5 3.4 -7.7 0.3
5th Jul. 2016 23.7 1.6 13.8 3.6 -7.7 0.5
12th Jul. 2016 24 1.7 14.7 3.8 -7.6 0.8

Graph showing the deviation from Standard Time (in seconds) and rate (in seconds per year) for the SP-410

h102 graph

Graph showing the deviation from Standard Time (in seconds) and rate (in seconds per year) for the SP-400

sp400 graph

Graph showing the deviation from Standard Time (in seconds) and rate (in seconds per year) for the SP-400G

sp400g graph

Graph showing the deviation from Standard Time (in seconds) and rate (in seconds per year) for the Crystron 4 Mega

crystron graph

Graph showing the deviation from Standard Time (in seconds) and rate (in seconds per year) for the Exceed 4 Mega

exceed graph

Graph showing the deviation from Standard Time (in seconds) and rate (in seconds per year) for the Omega

omega graph

A Note on the Accuracy and Reliability of Timing Results

There is always going to be an unknown margin of error in any method of timing, and I estimate that each timing event could be off by as much as 0.1 second (though in practice I believe my results to be somewhat better than this). As a consequence of this potential margin of error, results over short periods of time are unreliable. Ultimately, the only result that matters is that which comes at the end of a full 365 days. Where interim results show consistent timings or steady trends, however, then it is safe to assume that those results are fairly accurate.