The ISIS Proton Synchrotron: Beam Data Acquisition and Analysis Using NI PXI and LabVIEW

"With guidance from NI engineers, we rapidly created easy-to-use, well-designed LabVIEW user interfaces to control the systems and process and display data. These PXI-based systems have contributed towards a 20 percent increase in accelerator performance."

- Sarah Whitehead, ISIS – Science and Technology Facilities Council

The Challenge:

Improving the beam diagnostic system of the ISIS 50 Hz synchrotron to better monitor, control, and improve the accelerator’s performance.

The Solution:

Installing several new PXI-based data acquisition systems to increase the quantity, accuracy, and speed of the obtained diagnostic data.

Author(s):

Sarah Whitehead - ISIS – Science and Technology Facilities Council
Bryan Jones - ISIS - Science and Technology Facilities Council

 

ISIS is the world-leading pulsed neutron and muon source based at the Rutherford Appleton Laboratory in the UK. The facility’s 50 Hz synchrotron accelerates proton beams up to 800 MeV, delivering a mean beam power of 0.24 MW. A recent major expansion of ISIS requires higher beam currents from the synchrotron and has led to a program of detailed studies and upgrades.

 

Currently, the ISIS synchrotron accelerates a beam of 2.8 by 1013 protons to 85 percent of light speed, 50 times per second. Each acceleration cycle lasts 10 ms, during which the beam makes 12,000 revolutions of the 163 m circumference, with the proton energy gradually rising from 70 to 800 MeV.

 

Beam Loss Monitoring

The amount of beam the synchrotron can accelerate is limited by the proportion of the beam lost during acceleration, which is the beam that escapes the accelerating fields and hits the edge of the beam pipe. The lost beam can activate the synchrotron components, which makes hands-on maintenance problematic. Therefore, monitoring loss levels is crucial to successful operation of the facility. We distributed 39 beam loss monitors, which are 3 m long, argon-filled gas ionization tubes that indirectly detect protons striking the edge of the beam pipe, around the circumference of the synchrotron. Previously, the loss levels were displayed in the ISIS control room in a histogram showing total loss in each of the 10 synchrotron sections summed over an acceleration cycle.

 

We improved this display by installing a new PXI-based data acquisition system. The system consists of an NI PXI-1045 chassis housing a PXI-8105 controller, a PXI-6651 synchronization module, and five PXI-6133 digitizers. We chose PXI for its compactness and high-performance data acquisition and synchronization capabilities.

 

With the PXI system, we acquire data at 20 kS/s and stream it over the lab network using NI DataSocket software. We developed an NI LabVIEW VI to receive and display the data so multiple users can simultaneously access and analyze it in different ways.

 

The main feature of the VI is an ActiveX 3D graph control displaying the loss (Figure 1).This is similar to the existing histogram display, but with four times the spatial resolution and a time axis to show loss at 50 µs intervals. At a glance, beam loss can be identified on individual monitors at any time within the acceleration cycle. The GUI can also be used to print graphs, save reference data, and view individual monitors or the sum of all monitors. It recreates the old histogram display for comparison.

 

Multichannel Profile Monitoring

The two new multichannel profile monitors (MCPMs) use an array of 40 electron multipliers (channeltrons) to measure real-time transverse beam profiles within the ISIS synchrotron (Figure 2). The circulating proton beam interacts with the residual gas in the beam pipe to produce positive ions. A high-voltage drift field is applied across the beam, which directs the ions toward the monitor.

 

 

 

The data acquisition system consists of a PXI controller and chassis and five 8-channel NI PXI-6133 modules, which are sampled simultaneously using a PXI-6651 timing and synchronisation module. On every 50 Hz pulse of the ISIS synchrotron, the 40 channels are read at a user-selected sample rate for the acceleration cycle. From the front panel of the VI, we can view the profile of the beam at a selected time, graphs that display the 90 percent beam widths and centres over the acceleration cycle, and a 3D plot of all the profiles (Figure 3). Also, we can choose to subtract background data, save data, and load and superimpose the saved data over the currently selected profile.

 

Because the gain of the 40 channeltrons varies greatly, the MCPM needs to be calibrated using a motorized, single-channel profile monitor (SCPM), which is housed in the same assembly as the MCPM. The SCPM is the predecessor to the MCPM and has one channeltron driven by a DC motor across the beam. The SCPM moves across the beam using a step size that aligns it with the transverse positions of the 40 channeltrons in the MCPM. The data is saved to a database, and using the HTTP server from the LabVIEW Internet Toolkit, the MCPM VI can access the data and use it to create a calibration file that is applied to the raw data from the MCPM. The PXI system simplified the process of simultaneously collecting and analysing large amounts of data from the 40 channels.

 

We also developed a profile simulator to test all the electronics and computer hardware and software in the absence of a proton beam. The simulator is a microcontroller that can be switched into the circuit using the digital lines of one of the NI PXI-6133 modules. Different profiles are preloaded to the microcontroller and selected using an NI PXI-6533 digital I/O module. An NI PXI-5412 signal generator module provides a waveform equivalent to the intensity falloff that occurs during the acceleration cycle. This waveform is applied to potentiometers that adjust the output level of each channel (Figure 4).

 

 

 

Beam Position Monitoring

Accurate determination and correction of the transverse position of the proton beam is a major factor in reducing beam loss and thereby increasing acceleration efficiency. We installed 10 horizontal and 10 vertical beam position monitors in the synchrotron. Each monitor has two electrodes on which the beam induces a charge as it passes (Figure 5).

 

In the past, we acquired data from these monitors via five 4-channel digital oscilloscopes at 100 MS/s with 8-bit resolution and a 250 kb record length. Then a PC read out the data and a library of FORTRAN routines processed it. An IDL GUI provided measurement control and data visualisation.

 

The new PXI-based system consists of a PXI-1045 chassis housing a PXI-8105 controller, a PXI-6651 synchronisation card and 10 PXI-5124 digitisers. These provide 200 MS/s sampling at 12-bit resolution and a 512 Mb record length. Now we can acquire data over a whole acceleration cycle at double the previous sample rate and 16 times the previous resolution. The data is processed via a LabVIEW VI that synchronizes and configures the digitisers and acquires 10 ms of data from each monitor electrode. The data is processed and a file containing the beam position on each revolution of the synchrotron is written for each monitor. Then the IDL software picks up the file for further analysis and display.

 

We use LabVIEW modules to subtract the signal baseline, find signal peaks and troughs, and calculate the beam position. The data can also be converted and sent to our existing FORTRAN DLLs to benchmark the system. Data acquisition is triggered from any PC in the ISIS network using network-published shared variables. We tested the new system and can closely monitor and correct down the beam positions to ±1 to 2 mm.

 

Successful Development of a Beam Data Acquisition System

With guidance from NI engineers, we rapidly created easy-to-use, well-designed LabVIEW user interfaces to control the systems and process and display data. These PXI-based systems have contributed to a 20 percent increase in accelerator performance.

 

Author Information:

Sarah Whitehead
ISIS – Science and Technology Facilities Council

Figure 1.
Figure 2.
Figure 3.
Figure 4.