Some tests on the electrical
properties of tubes and foils for use in grounding and shielding of the Atlas
SCT
Tony Smith Feb ‘01
SCT barrel and
forward will have foil screens
Some Questions
Should barrel tubes be joined to the screen at both ends?
Should barrel tubes be isolated at one end or in the middle?
Should forward wheels have foil on the face of the wheel shorting the cooling tubes?
Some tests
Measurement of the inductance, resistance and pickup from a loop of tube
Magnetic field pickup from tube measures current flow in tube
Five configurations of foil and tube:
Tube only
Tube and foil with foil o/c at both ends
Tube and foil with foil short to tube at one end
Tube and foil with foil shorted to tube at both ends
Tube and foil with tube shorted to foil at both ends but tube o/c at one end
Some results
Tube has significant inductance
Resistance increases with frequency due to skin effect
Foil has much lower inductance
Foil shunts most of the current away from the tube if connected both ends
Current in tube is very small.
Very little change in current if tube o/c
Some conclusions
Barrel:
Incoming tubes shorted to foil at both ends and tube continuous through barrel
>>>Current flow in tubes giving rise to magnetic coupling would be reduced.
>>>Current flow in tubes giving rise to voltage drops along the tubes and hence
between modules would be reduced
Incoming tubes shorted to foil at both ends but tube o/c :
>>> Current flow in tubes giving rise to magnetic coupling would be near zero
>>> Voltage on modules close to the break may be significant relative to foil screen or other module on other side of the break
Forward:
Shorting cooling points with foils on the wheel face reduces the risk of common mode currents generating problem voltages between modules. – factor 40?
Measurements on plain tube loop
Figure 1 Loop of tube
|
Frequency |
Voltage (mV) |
Current (mA) |
Rs (ohms) |
Ls (uH) |
Vscope (mV) |
|
10 MHz |
1024 |
5.587 |
13.78 |
2.87 |
26.6 ( 105) |
|
5 |
884 |
9.175 |
0.335 |
3.03 |
24.4 (58.5) |
|
1 |
426 |
22 |
0.084 |
3.04 |
8.6 |
|
0,5 |
214 |
22 |
0.047 |
3.05 |
7.0 |
|
0.1 |
42 |
22 |
0.006 |
2.96 |
2.5 |
*
note Numbers in brackets are corrected for 22 mA drive current
From
the results above it can be seen that the inductance is constant as expected
over the 0.1 - 5.MHz band and there is a mall measurement error at 10 MHz.
The
skin effect explains the measured increase in the resistance with frequency as
the current travels in a smaller depth of the metal.
The
pickup voltage on the scope broadly follows the expected linear dependence on
frequency.
A second set of measurements were performed with a setup as in figure 2
Figure 2 Loop of tube plus foil
Here
a sheet of 20 um Aluminium foil was placed on a 36 cm diameter former made from
wood and card. The tube of the lower part of the loop was placed 3 cm inside
the former. Connections at either end of the foil at points A and B could be
made between the foil and the tube. These connections were made with 3 cm wide
foil
strips
wrapped around the tube and secured with clips.
Measurements
were made with no connections at A or B, just at A, and both A and B.
. Foil on former but no connection at A or B
|
Frequency |
Voltage (mV) |
Current (mA) |
Rs (ohms) |
Ls (uH) |
Vscope (mV p-p) |
|
10 MHz |
1020 |
5.836 |
13.17 |
2.74 |
27.36 ( 103.1) |
|
5 |
874.6 |
9.522 |
0.320 |
2.885 |
21.00 (48.51) |
|
1 |
407 |
22 |
0.035 |
2.901 |
11.00 |
|
0,5 |
203.9 |
22 |
0.047 |
2.906 |
6.70 |
|
0.1 |
40.96 |
22 |
0.018 |
2.916 |
1.57 |
Foil connected only at end A
|
Frequency |
Voltage (mV) |
Current (mA) |
Rs (ohms) |
Ls (uH) |
Vscope (mV p-p) |
|
10 MHz |
1.019 |
6.083 |
10.53 |
2.624 |
26.86 (97.14) |
|
5 |
NR |
NR |
-0.64 |
2.747 |
18.16 ( NC) |
|
1 |
404.7 |
22 |
0.094 |
2.885 |
10.72 |
|
0,5 |
203 |
22 |
0.055 |
2.894 |
6.56 |
|
0.1 |
40.87 |
22 |
0.022 |
2.910 |
1.76 |
Foil connected both ends
|
Frequency |
Voltage (mV) |
Current (mA) |
Rs (ohms) |
Ls (uH) |
Vscope (mV p-p) |
|
10 MHz |
1.015 |
7.322 |
10.635 |
2.160 |
1.78 (5.34) |
|
5 |
624.5 |
11.65 |
0.301 |
2.214 |
0.71 (1.34) |
|
1 |
317.9 |
22 |
0.221 |
2.259 |
0.79 |
|
0,5 |
160 |
22 |
0.135 |
2.269 |
0.77 |
|
0.1 |
32,6 |
22 |
0.072 |
2.310 |
0.47 |
Foil connected at both ends and tube open circuit at A
|
Frequency |
Voltage (mV) |
Current (mA) |
Rs (ohms) |
Ls (uH) |
Vscope (mV p-p) |
|
10 MHz |
1.014 |
7.299 |
10.92 |
2.165 |
1.80 (5.43) |
|
5 |
825.0 |
11.56 |
0.390 |
2.233 |
0.84 (1.56) |
|
1 |
320.6 |
22 |
0.210 |
2.276 |
0.77 |
|
0,5 |
161.0 |
22 |
0.155 |
2.285 |
0.78 |
|
0.1 |
32.89 |
22 |
0.095 |
2.320 |
0.78 |
|
|
|
|
|
|
|
Effective schematic with and without foil and electric and magnetic pickup mechanisms in detector input loop
If the screen is shorted at A and B then the voltage drop per unit length between A and B is the same in the screen as it is in the tube so the voltage on the screen relative to the module tied to the tube is near zero.
If the tube is o/c at one end, then the tube will be equipotential with one end of the screen, but the other end of the screen will have a voltage signal on it relative to the module.
If the tube is o/c in the centre the two halves of the tube will be equipotential with the respective ends of the screen but a module at the centre will see half the screen voltage drop relative to the screen and the full drop relative to its neighbour on the other side of the break.
