Application Note - GBS Elektronik GmbH

Application Note
Influence of very long detector cables on spectrum
measurement with MCA166
Jörg Brutscher
GBS Elektronik GmbH
12.4.2000
Problem:
For some applications, it is desirable to use very long detector cables. Normally, only a short detector cable
comes with a detector and it is not recommended to use longer cables.
This paper shall point out the problems which occur using very long detector signal cables, how they can be
overcome and what compromises have to be made in terms of system performance.
Effects and problems are not specific to MCA 166, but concern interaction of any MCA with a detector.
Experiment
A Co60 source was measured with a GL0310 HPGe detector. At first, the spectrum was measured with the
standard cable (approx. 4m length). Then the spectrum was measured again with a 250 m coaxial cable
(impedance 50Ω) introduced between detector signal out and MCA signal input.
Measuring the Co60 spectrum caused 90% of the high energy part of the spectrum to be cut off! Examining the
preamplifier signal with a scope showed why:
As the MCA 166 has an input impedance of 1 kOhm, the input is not matched to the cable impedance and the
signal is reflected. Also the detector preamplifier has a poor match to the cable and the signal is reflected again.
This can be seen as an attenuated version of the original signal with a delay which corresponds to 2 times the
cable signal traveling time. The cable length of 250 m corresponds here to a delay of 2.7 µs.
Preamp Signal (V)
0
-1
-2
Preamp signal with 250 m cable, terminated by 50 Ohm
Signal
Echo
Preamp Signal with short cable
Preamp Signal with 250m cable
0 2 4
Time (µs)
6 8 10
Fig. 1: Preamplifier signal with short cable, with long cable (250 m) and with long cable and a 50 Ohm
terminating resistor parallel to the MCA input.

The point is now that the pile up rejector (PUR) interprets this echo as a close following event and this way
reject everything. The time limit for the PUR to distinguish between two close following signals is about 500 ns.
So for cable length >50 m this effect should be seen.
If the spectrum is measured with PUR switched off, a "normal" can be seen. However, the resolution seems to be
a bit decreased.
Another possibility to get rid of the echo signals is to place a 50 Ohm termination resistor in parallel to the MCA
input. This eliminates the echo signals, but also reduces significantly (66%) the signal amplitude. So with this
method the amplification has to be increased; also here a slight degradation in resolution can be found.
Fig. 2: Co60 spectrum measured with short cable, with long cable, with long cable and PUR off, with long cable
and 50 Ohm termination.
Condition Resolution at 1333 keV Drift
short cable 1.73 keV
250 m cable, PUR off 1.90 keV 0.833
250 m cable, 50 Ohm terminated 2.77 keV 0.34
250 m cable, 50 Ohm terminated, amplification increased 1.87 keV (1)
Conclusions
-Without precautions signal cable length >30m cannot be recommended.
-For cable length >30m the PUR has to be switched off or a 50 Ohm termination resistor has to be placed in
parallel to the MCA signal input.
-For long cable length, a slight resolution degradation has to be taken into account.
GBS -Elektronik GmbH, Bautzner Landstr. 22,
01454 Großerkmannsdorf / b. Dresden
Tel.: (0351) 217007-0 Fax: (0351) 217007-21
homepage: http://www.gbs-elektronik.de/
email: contact@gbs-elektronik.de