Application note
Propagation Delays of BNC, SMA and N
connectors, components and cables
Ivan Prochazka
CTU Prague, Brehova 7, 115 19 Prague 1, Czech Republic, ivan.prochazka@fjfi.cvut.cz
Goals
To determination of propagation delays of various SMA and BNC signal connectors
and converters with picosecond resolution and a few picoseconds accuracy.
1. Delay measurement procedures
The propagation delay was measured using these procedures:
1. The additional signal delay of components under test inserted into test line was
measured. The NPET timing system version “Preliminary ELT Calibration Timing
Device” was used. Each measurement consisted of 500 readings, 3*sigma editing
algorithm was applied. The standard deviation sigma was typically equal to 1 ps. As
a test signal source the signal generated by the timing device was used. It provides
highly uniform pulses with amplitude of 0.38 V and ~ 50 ps fall-time. The trigger
level of the timing device was set to -0.125 V, falling edge. Obviously the inserted
components (or their combination) under test had to be M-F configuration only.
2. Each the individual component under test was measured using a digital scale, the
effective signal path length was determined for each component. The effective
signal path length for each component (whenever meaning full and measureale) is
summarised in Table 1.
3. The measured delay versus eff. signal path length for combinations (M-F) of SMA
only components is plotted in Figure 1. The linear fit to the measured data was
constructed without the point 0.,0.. Nevertheless the fitting line is passing well cross
zero point. It means the effective signal path length was identified and determined
correctly for the components. The data spread is 2 ps rms, corresponding well to the
reproducibility of SMA connectors re-connection and to the precision of mechanical
measurement. The fitting line slope corresponds to 4.5 +/- 0.1 ps / mm.
4. Analogical measurements were completed for BNC only components. The linear fit
to the measured data was constructed. The data spread is 2 ps rms, corresponding
well to the reproducibility of connectors re-connection and to the precision of
mechanical measurement. The fitting line slope corresponds to 4.7 +/- 0.1 ps / mm.
5. Considering the results for SMA and BNC connectors one can conclude that a mean
value of the delay constant is 4.6 +/- 0.1 ps /mm. This value was consecutively used
to evaluate the delay of individual components on the basis of their corresponding
effective signal path length. In cases when both direct delay measurement and
calculated data are available, the mean value is presented. In all these cases the
discrepancy was within +/- 2 ps.
6. The F-M attenuators, inverters, biased T were tested using a pulse generator and
Tex oscilloscope 2.5GHz BW, 40 Gs/s, at a speed of 50 ps/dot. The pulses were
characterised by 50% of corresponding pulse amplitude.
Figure 1 Measured delay versus effective signal path length for combinations of SMA
components.
2. Comments


The losses and BW limitation of the components were neglected, fifed threshold
was used to detect pulses
The “T” components were included. Their use may be considered as F-M
component with a second F port used for “pulse pick-off” purposes.
type
eff. signal path
length (mm)
delay (ps)
25.5
119
19.0
89
20.4
96
17.4
82
17.0
79
NA
158
16.5
76
8.5
39
NA
110
NA
87
BNC-BNC
F–F
BNC – BNC
M–M
BNC – SMA
F–M
BNC – SMA
M–F
BNC – SMA
F–F
BNC. BNC.BNC
F–M
path
SMA – SMA
M–M
SMA – SMA
F–F
SMA.SMA.SMA
F–M
path
SMA – SMA
F–M
Table 1 BNC and SMA connectors and converters, effective signal path length and delays
type
BNC – BNC
attenuator 2x
M–F
BNC – BNC
trafo – inverting
M–F
SMA – SMA
DC insulator
M–F
eff. signal path
length (mm)
delay (ps)
40.8
170
NA
570
24.5
124
~ 23
106
~ 24
111
NA
175
NA
direct path
N – SMA
F–F
N – SMA
M–M
SMA – SMA
attenuator
F–M
SMA – SMA
power splitter
F–F- F
157
SMA – SMA
direct path
biased T
NA
241
F–F–M
Table 2 BNC, N and SMA connectors and components, effective signal path length and
delays.
3. Signal cable delays
The SMA – SMA signal cables for ELT Calibration Device were tested for their
delays. The cable delay was determined in connection to the SMA-SMA F-F interface. Its
additional delay of 39 ps was taken into account.
Cable Id
ELT No. 1
ELT No. 2
ELT No. 3
ELT No. 4
signal delay
7.116 ns
7.114 ns
7.111 ns
7.112 ns
precision rms
2 ps
2 ps
2 ps
2 ps
average ELT
7.114 ns
2 ps
Table 3 Measured signal delay of cables marked “ELT No.x”.
Conclusions
 Measurement technique for signal propagation delay of individual
components has been developed and tested, see 1.
 The test procedure and its results have been verified by and independent
check based on physical dimensions measurements. The results of both
measurement techniques coincided well within +/- 2 ps in all cases.
 The measured delay values are summarised in Tables 1,2 and 3.
 The cables delays are equal within measurement error.
 Applying the delay determination technique and instrumentation
described above the delay of any signal component used in ELT
calibration process may be determined with ps precision and accuracy.
in Prague, April 17, 2015