Electron Cloud in the LHC:
Multipacting tests in a superconducting magnet

With a wide-band power amplifier (1-1000 MHz, 100 W) we have excited TEM modes in a 1 m long coaxial structure accomodated in the cold bore of an 8.4 T magnet. The secondary electron yield and its possible reduction by a strong B-field could be inferred from the multipacting level reached under suitable resonance conditions (either for pulsed or CW excitation). Measurements of photoelectron yield were also possible using UV light from a Xe flash lamp mounted near or inside the coaxial tube.

Latest news

30/09/1997 F. Caspers, J.-M. Laurent, M. Morvillo, and F. Ruggiero, Multipacting tests with a resonant coaxial setup, CERN LHC Project Note 110 (1997).

26/08/1997 Received three capillary Xe flash lamps, with 0.16 joule per pulse, and a second large bulb Xe flash lamp, with 5 joule per pulse.

7/08/1997 Cold multipacting tests with the resonant coaxial setup indicate no significant influence of the magnetic field up to about 7.5 T (10000 Amps). On the left in the picture, from top to bottom: function generator, scope, signal generator and power amplifier. On the right: cryostat containing a two-in-one, 1 m long superconducting dipole.

28/07/1997 Started mounting the resonant coaxial setup in a short LHC magnet prototype. Since we haven't yet received any flash lamps, the first multipacting tests in an 8.4 T field (scheduled for next week) will not include photoelectric effect.

26/07/1997 Warm multipacting tests with "resistor probe". (Scope pictures compressed Postscript file, 251 kB)
A thin wire is radially connected to the middle of the inner conductor through a 1 kOhm inner resistor and comes out of the outer tube via a small hole with an external 100 kOhm resistor in series: the corresponding signal can be seen on a scope with high impedance. The generator signal with a frequency around 480 MHz is amplitude modulated (100% depth) with frequency fAM=2.5 Hz (first three pictures) or fAM=9 kHz (last three pictures) and amplified at maximum gain. In the `rt'-pictures the upper trace is the reflected signal attenuated by 40 dB and the lower trace the transmitted signal, both measured with a 50 Ohm load. The `R'-pictures show the `resistor probe signal' measured with a 1 MOhm (or higher) load.

Art) W=-36.7 dBm	BrR) W=-20.0 dBm	Crt) W=-19.3 dBm
Drt) 1 microsec/div	D'rt) 10 microsec/div	ER) 10 microsec/div

18/07/1997 More multipacting tests with the warm resonant coaxial setup, modified so that we can apply a dc-voltage to the inner conductor. (Scope pictures compressed Postscript file, 474 kB)
The generator signal with a frequency around 480 MHz is amplitude modulated (100% depth) with frequency fAM=3 Hz. In the `rt'-pictures the upper trace is the reflected signal attenuated by 40 dB and the lower trace the transmitted signal, both measured with a 50 Ohm load. The `f'-pictures show the forward amplifier signal attenuated by 40 dB and the `T'-pictures show the transmitted signal measured with a 1 MOhm load.

Art)	AT)	Bt) 0 V
Bf) 0 V	Crt) 0 V	C'rt) 0 V
Drt)	Ert)	ErT)
Frt) 0 V	Grt) 100 V	GrT) 100 V
Hrt) -50 V	Irt) -250 V	JrT) -440 V

17/07/1997 New multipacting tests with the resonant coaxial setup, modified so that we can apply a dc-voltage to the inner conductor. A thin wire is radially connected to the middle of the inner conductor through a 50 kOhm resistor and comes out of the outer tube via a small hole. When connecting inner and outer conductors (without dc-bias), the multipacting stops and a higher generator power is required to let it start again. Once this happens, however, we can reduce the generator power to the original level without stopping multipacting. This seems to indicate that the inner conductor acquires a positive charge during multipacting: after removing this charge, it is more difficult to start multipacting again. (Scope pictures compressed Postscript file, 323 kB)
The generator signal with a frequency around 480 MHz is amplitude modulated (100% depth) with frequency fAM=30 Hz. In the `rt'-pictures the upper trace is the reflected signal attenuated by 40 dB and the lower trace the transmitted signal, both measured with a 50 Ohm load.

Art)	Aft)	Brt)
Crt) 0 V	Drt) 100 V	Ert) 23 V
Frt) 200 V	Grt) 150 V	Hrt) 0 V

14/07/1997 First multipacting tests with the resonant coaxial setup show rather reproducible results. After solving some initial vacuum problem, we observe clear multipacting thresholds: with AM modulation, the reflected signal suddenly increases and the transmitted signal remains constant over the multipacting range (see for example figure Crt, showing two multipacting levels). Sometimes we can also detect a direct electron current collected by the weakly coupled port and superimposed on the transmitted signal: the positive current in figure Aft may be due to a discharge. (Scope pictures compressed Postscript file, 396 kB)
In the `rt'-pictures the upper trace is the reflected signal attenuated by 40 dB and the lower trace the transmitted signal, both measured with a 50 Ohm load. In the `ft'-pictures the upper trace is the forward amplifier signal attenuated by 40 dB and the lower trace the transmitted signal measured with a 1 MOhm load.

Aft)	Brt)	Bft)
Crt)	Cft)	Drt)
Dft)	Ert)	Eft)
Et)	Frt)	Fft)

7/07/1997 Tests with a dc bias in the multi-wire warm setup confirm that what we observe is really multipacting. Using a button-like probe, we can even measure a direct electron current during multipacting.

The oscilloscope pictures show discontinuity in the reflected signal and collected electron current during multipacting in the multi-wire warm setup, with AM modulation (at 5 Hz) of CW signal for resonant excitation at 47 MHz. 1) Upper trace: reflected signal with peak voltage of ~200 mV with a 40 dB attenuation, 2) Lower trace: signal from a button-like electron probe with ~400 mV peak value, measured with a 1 MOhm load (M. Morvillo).

4/07/1997 The onset of multipacting in the existing multi-wire coaxial setup, resonantly excited at 140 MHz, has been successfully detected electronically: a sudden jump in the reflected signal can be clearly observed when the (amplitude modulated) input power exceeds a threshold value. Contrary to preliminary results obtained so far with pulsed excitation, this effect is very reproducible and the pressure level does not seem to increase. Further tests will continue with a dc bias to cross-check that what we observe is really multipacting.

2/07/1997 Ordered two large bulb Xe flash lamps, with 5 joule per pulse (Oriel N. 6427): they should arrive at CERN by 11/07/1997. These flash lamps have a peak power of several hundred kW and could be mounted at one end of the coaxial tube, but we may face cooling problems for operation under vacuum.

27/06/1997 Cold multipacting tests with an 8.4 T magnetic field (short LHC magnet prototype in Bloc 4) are scheduled for the week 4-8 August 1997.

26/06/1997 Ordered two UV detectors of the semiconductor type: they should arrive at CERN in about 10 days and will be useful to check the proper functioning of the Xe flash lamps.

23/06/1997 Received wide-band power amplifier (AR 100W1000, 1-1000 MHz, 100 W, 50 dB gain, visible on the left of the picture). Multipacting tests at room temperature can start on an existing multi-wire coaxial setup by J.M. Laurent (vertical tube shown in the picture). Mechanical pieces required for cold multipacting test with B-field (inner tubes, teflon supports and coupling ports) ordered at the CERN workshop: they should be ready in about one week.

19/06/1997 Ordered three capillary Xe flash lamps, with 0.16 joule per pulse (Oriel N. 6425): they should arrive at CERN by 21/07/1997. These flash lamps deliver microsecond pulses of photons with energies up to 6-7 eV and could be mounted inside the coaxial tube. We will have to learn how to trigger them.

6/06/1997 Ordered wide-band power amplifier (Amplifier Research Model 100W1000, 1-1000 MHz, 100 W, 50 dB gain): should arrive at CERN by 27/06/1997.