- Email: [email protected]
A remote control system for the LELA experiment
Nuclear Instruments and Methods in Physics Research North-Holland, Amsterdam
A239 (1985) 235-242
235
A REMOTE CONTROL SYSTEM FOR THE LELA EXPERIMENT M. CASTELLANO, N. CAVALLO, F. CEVENINI and P. PATTERI INFN Sezione di Napoli and Istituto di Fisica Sperimentale dell'Università di Napoli, Napoli, Italy
Received
30 November 1984
A modular system for closed loop computer control of stepping motors has been realized and used for optical component movement of LELA experiment in the radiation risk area. A CAMAC module, controlling up to 15 stepping motors, a NIM dual motor driver and a special purpose circuit for computer interfacing are described .
1. Introduction Since 1977, when the feasibility of a free electron laser (FEL) using an helicoidal undulator magnet on the electron beam of the Stanford Superconducting Linac was demonstrated [1,2], one of the major subjects of interest has been how such a device can be operated on a recirculating beam machine, such as an electron storage ring [3-6]. The LELA (laser ad elettroni liberi in Adone) experiment has been designed to study problems involved in the interaction between radiation and the recirculating electron beam of a storage ring (Adone of LNF) [7] using a transverse undulator of 20 periods. Accurate measurements of the angular and spectral distribution of spontaneous radiation from the undulator were performed in the first stage of the experiment [8,9] . The results agreed well with calculated distributions, with no pattern distortion due to an imperfect magnetic field. In fig. 1 one of the measured radiation spectra is shown. The gain measurements and the construction of a suitable optical cavity for laser oscillation are the next step of the experimental program . We have already reported gain measurements which agree with theoretical predictions [8,10,11] . In this paper we shall describe the control system for mechanical movements developed for this experiment . The layout of the experimental set-up is shown in fig. 2. The PDP 11/40 computer used for control and data acquisition is installed in a remote control room from which no direct view of the apparatus is possible. This fact, as well as the necessity for remote control of mirrors, detectors and beam monitors, located in a high radiation risk area outside the experimental bunker (whose limits are indicated by dashed blocks in fig. 2), 0168-9002/85/$03 .30 © Elsevier Science Publishers B.V . (North-Holland Physics Publishing Division)
indicates the necessity for a closed loop automatic control, without, on the other hand, losing the possibility of a manual and direct control from inside of the bunker. For all those movements which do not require a sensitivity better than a few micrometers, it was decided to use commercial micrometric translators moved by stepping motors . One of the typical problems of the high resolution movements, which is of particular importance in our case, is the reproducibility of positions during the successive scans. We discarded, for financial reasons, the possibility of associating an absolute encoder of opportune sensitivity and range (- 1 pm on some centimeters) to each movement. Therefore we decided to use incremental optical encoders, which would guarantee the mechanical actuation of the step, together with reference points (start and end of run), to be totally reproducible within a single step of the motor. Because of the peculiar features of this system we decided to make out own main circuits to get the best results. The advantage of integrating the control system for the stepping motor motions with the data acquisition system has led us to prefer a standard CAMAC solution for the computer interface. On the other hand, the need for a physical separation of the different functions and the difficulty of establishing a priori the number and the type of movements necessary for the different phases of the experiment, have influenced our choice of a system with modular structure and separate functions . In sect. 2 we describe the general layout of the system by emphasizing the distribution of functions and their link in a closed loop configuration . Sect. 3 is dedicated to a description of the circuits built for this system and their specific performance. A brief account
236
5000
Fig. 1. Spontaneous radiation pattern measured in the radial dimension . Experimental points are interpolated by means of a cubic spline algorithm.
Fig. 2. Layout of the experimental hall. Concrete shields delimiting the accessible area are shown as dashed blocks.
M. Castellano et al. / Control system for the LELA experiment
of the way in which the system has performed in measuring the spontaneous radiation is presented in sect. 4. 2. General layout of the control system A schematic view of the control system is shown in fig . 3. The computer controls the stepping motor motions through the control module, a home-made CAMAC circuit with a drive capacity of 15 motors . For each motor two distinct pulses for forward and backward movement reach, through coaxial cables, the power driver ; this is a standard NIM module located inside the experimental bunker . This circuit can drive motors with different characteristics of voltage and speed and can make them execute steps of half or full length, as will be described in the next section. S. M. CONTROLLER MODULE (FOR IS S . M .)
The power driver supplies the current to the four motor phases under control of the step command, and it contains also a circuitry to control a shaft encoder which can be connected with the motor. A four digit counter memorizes the motor steps by counting the driving pulses. If the shaft encoder is connected with the motor the counter increment is conditioned by its consent. The power driver can also be operated manually for testing the movements or setting up the apparatus . The motor counters can be read by the computer. To minimize cables, a multiplexer unit is located inside the experimental bunker. This unit decodes motor addresses from the computer and multiplexes the addressed counter content to a CAMAC input register. The same unit also takes care of all possible end-ofstroke signals generated by the shaft encoder electronics. A stop pulse is sent to the computer, and a status
TTL INPUT TTL CRATE OUTPUT REGISTER CONTROLLER ' REGISTER
II I I I I II
EXPERIMENTAL AREA
(FOR 15 STEPPING
Fig . 3. Schematic view of the control system.
237
MOTORS)
238
M. Castellano et al. / Control system for the LELA experiment
SN74S288 PROMs, is controlled by an AM2909 sequencer . A motor movement program is obtained writing, with CAMAC subaddress A as motor address, a 16 bit word in the 16 x 16 RAM of the module, the most significative bit being the movement direction . The motor address is also inserted into a 16 x 4 FIFO memory . When all the involved movements have been programmed, the subaddress 0 has to be inserted into the FIFO. This subaddress does not correspond to a motor address and is used as a separation word in the FIFO itself. When a start command is issued by the computer, motor addresses are sequentially extracted from the FIFO, addressed RAM content is decremented by 1 and the relative output flip-flop is set. If the decremented RAM content is greater than 0 the number is rewritten into the RAM and the address is reinserted into the FIFO. This operation is repeated until the separation word is extracted from the FIFO. A pulse is then generated and sent to all outputs, which are gated by the output flip-flop. A step is thus simultaneously executed by all programmed motors. FIFO scanning is then repeated, and pulses generated, until no more addresses are contained in the FIFO.
register is strobed, enabling it to reconstruct the stop source. A closed loop control system has thus been implemented . All the movements, once programmed, are automatically executed allowing the computer to make a preliminary analysis of data and/or to interact with the operator for an on line control of the measurement . When the end of movements is reached or a fault signal is detected a CAMAC LOOK-AT-ME interrupt makes the computer resume control. 3 . Circuit description 3.1 . The stepping motors' control module
This circuit, housed in a double unit CAMAC module, controls movements up to 15 stepping motors. For each motor two LEMO connectors are provided for clockwise and anticlockwise movement commands. The number of steps and the movement direction of each motor can be programmed, and a pulse for each step is generated by the respective output . The pulse train frequency may be programmed on a wide range (0.1 Hz to 1 kHz). The circuit, whose logic diagram is shown in fig. 4, is microprogrammed . The sequence of microinstructions, memorized in three
a 3a A F Q
4 6 ) (111 S TEPS PIS
W LL )
MEMORY
n
Q
2R
a U
J W
F. I . F. O. MEMORY
O
W U.
u.
m
Fig . 4. Logic diagram of the stepping motor control module.
CLOCK (5 MHz)
~ä
W w
ADDRESS,
u. u. m
lxW
LL
I18x4)
D: W
EXTERNAL STOP IN
(16X24) MICROSTUCTONS R
W
a W
aw
0
o:W 0 0 Q
o o4
r
OUT
FORWARD PULSES
OUT
BACKWARD PULSES
M. Castellano et al. / Control system for the LELA experiment
The end-of-movement flip-flop is set in the status register and a CAMAC LAM to the computer is issued. The 200 ns microinstruction cycle makes the variable delay introduced between successive pulses completely negligible . In this way it is possible to execute peculiar movements of optical devices such as parallel translation of mirrors, displacement and/or rotation of the optical cavity axis, with great advantage for the results of the experiment. 3 .2 . The power driver module
The power driver module is a two unit NIM module (fig . 5). It drives two independent four phase stepping motors and monitors their position and status through a multiplexed interface to the remote control system. Each driver, whose logic diagram is shown in fig. 6, receives clockwise or anticlockwise motion commands, TTL levels low true, through two coaxial LEMO connectors. The driver can also be manually operated by means of 'two front panel pushbuttons which, if briefly pressed, cause the motor to execute a single step in either direction, and, if continuously pressed, give way to a 100 Hz oscillator which drives the motor to a fixed speed run. In this way it can be fully operated as a stand-alone module. The step generation logic is based on a 32 X 8 PROM.
239
Four PROM bits drive the motor phases currents . Three other bits are used as less significant bits of the 5 bits next address to the PROM itself . The most significant bit of the next address, controlled by a switch on the printed circuit, selects a half or full step movement (0.9° or 1 .8°, respectively). The second one is derived from step command and determines the clockwise or anticlockwise step direction. The possibility of changing the step size allows us to achieve the best compromise between movement speed and accuracy for each application . Current to the four motor phases is derived from an external source, so as to adapt it to specific voltage or current requirements . The driver itself can be set on current or voltage mode by a suitable choice of a resistor in the final power stage. The same circuit has been used in this way with different requirements of power and speed. A differential sqt"^re wave type shaft encoder can be associated to each v it. The two bit status of the encoder, before and after the step execution, is used as a 4 bit address to a look-up table PROM . A forward or backward consent to the step counter or a missing response signal to the alarm system is then generated. If the encoder eletronics is disabled by a front panel switch, the step counter is incremented or decremented directly by the step command. The step counter content is visualized on a four digit display on the front panel and can be read by the
Fig. 5. Frontal view of NIM dual motor driver module .
240
M. Castellano et al. / Control system for the LELA experiment
4ooS/ T
2o0S/T C K Q
IN C
CO
LL P B . pP Q
YC) Q
m
STEP COMMAND LOGIC
SINGLE OR CONTINOUS STEP LOGIC
N W
x a
1
O
LOOK-UP
I TABLE I PROM L .-
I
O 1_
IN C
E . O. S. LOGIC
W
O Q 2
1
W
u IN
°
Q
F F
MISSING RESPONS
FF
N
SHAFT ENCODER LOOK-UP
cLK
a J ó
Lt,
FF
7
I
TABLE PROM
FWD
E.O.S .
R.C. INTERFACE
REMOTE CONTROL
ENABLE CLR
SWD I
TIMING LOGIC
4DECADES COUNTER
FF
FWD/ BWD DELAY YES
NOT
ENC .
Fig. 6. Logic diagram of motor driver module. PHOTOEMITTING DIODE
PHOTOEMITTING
LE PHOTOTRANSISTOR
Fig. 7. Schematic view of a linear movement with LEDs and phototransistor for end-of-stroke logic.
M. Castellano et al / Control system for the LELA experiment
computer through a rear panel CANNON connector. Two end-of-strokes are envisaged for each movement . In order to have maximum flexibility, any logical function of two independent signals can be implemented by jumpers on printed circuits for each end-ofstroke, and used to generate an alarm to the multiplexer unit . The end-of-stroke status is also displayed by two front panel LEDs . For all linear movements used in spontaneous radiation measurements, a coincidence between two optical switches has been used to generate end-of-stroke signals. One of the switches is connected to the linear movement, the second one to the circular movement of the stepping motor axis to ensure an absolute position reference within one motor step (fig. 7) . For linear movements driven by micrometric screws with a 0.5 mm pitch/turn, a single step of a motor of 400 steps/turn corresponds to a displacement of 1 .25 ltm, well below the mechanical resolution of the screw itself . The resetting of the starting point has been tested over a period of many days of continuous running of the linear movements, thus guaranteeing its use as a reference point, within the resolution limits of the whole mechanical system . 3.3 . The multiplexer unit
In each power driver module the counters and the end-of-stroke status are displayed on the front panel to allow safe operation when used in manual mode. The multiplexer unit ensures interrupting and polling capability when operating in remote computer controlled mode . This unit is linked to a CAMAC 1/O register
241
from which it receives a four motor address and two control bits, and to which it sends 16 data bits . No communication handshake with the computer is foreseen . Each double power driver module is linked to the multiplexer unit via 16 data lines, multiplexed between the two counters, 6 fault signals and 4 control lines. All fault signals from the drivers are ORed together to promptly detect a wrong condition and to send a stop pulse to the motor control module. Their status is also strobed in a register so that, by means of a polling routine, the computer can reconstruct the fault source and undertake the correct recovery action . The step counters of the drivers can be independently read or reset, allowing a reproducible starting point to be created, moving forward or backward up to an end-of-stroke signal, and resetting the counter. 4. Conclusions The system described in this paper has been used for a series of preliminary measurements for the implementation of the optical cavity of the free electron laser built on the Adone storage ring . In particular, the measurement of the spontaneous radiation has been instrumental in testing the regularity of the undulator magnetic field and its complete compensation . The measurement was carried out using a monochromator undergoing horizontal and vertical translations in order to scan in steps of 1 mm the farfield pattern radiated by the undulator at a distance of 20 m. The scanning was obtained by using stepping-motors controlled by the system described in the present paper. Hundreds of
6500
6000
5500
5000
I -1 .50
I -1 .00
I -0 .50
I 0 .00
1
0 .56
Fig. 8. Spontaneous radiation peak wavelength X. [A] (vertical axis) as a function of angle 0 [mrad] (horizontal axis). The continuous line represents the theoretical curve.
242
M. Castellano et al. / Controlsystem for the LELA experiment
frequency and position scannings have been carried out for several days . In particular, some polarization mea-
surements have requested circular scannings around a common axis. Some of the collected data have been plotted in fig. 1 .
Owing to the limited lifetime of the stored electron
bunches, the beam has been injected several times before completing the data collection . Therefore, the perfect reproducibility of the detector position has proven
essential for the accurate measurement of the radiation pattern.
A clear evidence of the correct performance of the
system was provided by the return to the reference position after performing the exact number of steps every time the horizontal movement was activated. In contrast, the vertical movement exhibited some amount of nonlinearity which produced a 6 steps deviation from
the initial position after 10 vertical scannings. This
behaviour was due to the complex dependence of the vertical platform position on the angular position of the step-motor shaft. The six step deviation lies within the
limits of accuracy corresponding to a 10 gm resolution. For high precision scanning the end-of-stroke signal was
used to restore the exact position.
The agreement among the measurements performed during several days stands out clearly from fig. 8 where the peak wavelength 7t c is plotted versus the angle 0. X, is related to 0 through the quadratic relation A~= /Y(1+K Z 2 +y 20 2 n Ji o being the undulator pitch, K the effective value of the normalized field intensity, y is the beam energy and n the harmonic order. The continuous line represents
the theoretical curve, which fits the experimental points nicely.
As a final comment we want to remark that the
location of the different functions among several places has proved particularly useful in view of the remote location of the computer.
The system reliability has allowed us to perform
several totally automatized measurements .
References L.R . Elias, W.M . Fairbank, J.M .J. Madey, H.A . Schwettman and T.I . Smith, Phys. Rev. Lett . 36 (1977) 717. D.A .G. Deacon, L.R . Elias, J.M .J . Madey G.J . Ramian, H.A . Schewettman and T.I . Smith, Phys. Rev. Lett . 38 (1977) 892. C. Bazin et al., in : Physics of Quantum Eletronics, vol. 8, eds., S.F. Jacobs, H.A . Pilloff, M. Sargent, M.O. Scully and R. Spitzer (Addison-Wesley, New York, Reading, 1982) p. 89. [4] A. Luccio, in ref. [3], p. 153.
[5] G. Dattoli, A. Marino and A. Renieri, Storage Ring Operation of the Free Electron Laser, Report CNEN 80/22cc (1980).
[6] D.A . Deacon, Basic Theory of the Isochronous Storage Ring Laser, Phys . Rep. 76 (1981) 349. R. Barbini and G. Vignola, LELA - A Free Electron Laser Experiment in Adone, Frascati Report LNF-80/12 (1980). [8] R. Barbini et al ., in : Bendor Free Electron Laser Conf ., J. de Physique C1-44 (1983). M. Castellano, N. Cavallo, F. Cevenini, M.R . Masullo, P. Patteri, R. Rinzivillo, S. Solimeno and A. Cutolo, Nuovo Cim. 81B (1984) 67 . [101 M. Biagini et al., SPIE 453 (1984) 275. [11] M. Biagini et al ., subnutted to Nucl. Instr. and Meth .
A239 (1985) 235-242
235
A REMOTE CONTROL SYSTEM FOR THE LELA EXPERIMENT M. CASTELLANO, N. CAVALLO, F. CEVENINI and P. PATTERI INFN Sezione di Napoli and Istituto di Fisica Sperimentale dell'Università di Napoli, Napoli, Italy
Received
30 November 1984
A modular system for closed loop computer control of stepping motors has been realized and used for optical component movement of LELA experiment in the radiation risk area. A CAMAC module, controlling up to 15 stepping motors, a NIM dual motor driver and a special purpose circuit for computer interfacing are described .
1. Introduction Since 1977, when the feasibility of a free electron laser (FEL) using an helicoidal undulator magnet on the electron beam of the Stanford Superconducting Linac was demonstrated [1,2], one of the major subjects of interest has been how such a device can be operated on a recirculating beam machine, such as an electron storage ring [3-6]. The LELA (laser ad elettroni liberi in Adone) experiment has been designed to study problems involved in the interaction between radiation and the recirculating electron beam of a storage ring (Adone of LNF) [7] using a transverse undulator of 20 periods. Accurate measurements of the angular and spectral distribution of spontaneous radiation from the undulator were performed in the first stage of the experiment [8,9] . The results agreed well with calculated distributions, with no pattern distortion due to an imperfect magnetic field. In fig. 1 one of the measured radiation spectra is shown. The gain measurements and the construction of a suitable optical cavity for laser oscillation are the next step of the experimental program . We have already reported gain measurements which agree with theoretical predictions [8,10,11] . In this paper we shall describe the control system for mechanical movements developed for this experiment . The layout of the experimental set-up is shown in fig. 2. The PDP 11/40 computer used for control and data acquisition is installed in a remote control room from which no direct view of the apparatus is possible. This fact, as well as the necessity for remote control of mirrors, detectors and beam monitors, located in a high radiation risk area outside the experimental bunker (whose limits are indicated by dashed blocks in fig. 2), 0168-9002/85/$03 .30 © Elsevier Science Publishers B.V . (North-Holland Physics Publishing Division)
indicates the necessity for a closed loop automatic control, without, on the other hand, losing the possibility of a manual and direct control from inside of the bunker. For all those movements which do not require a sensitivity better than a few micrometers, it was decided to use commercial micrometric translators moved by stepping motors . One of the typical problems of the high resolution movements, which is of particular importance in our case, is the reproducibility of positions during the successive scans. We discarded, for financial reasons, the possibility of associating an absolute encoder of opportune sensitivity and range (- 1 pm on some centimeters) to each movement. Therefore we decided to use incremental optical encoders, which would guarantee the mechanical actuation of the step, together with reference points (start and end of run), to be totally reproducible within a single step of the motor. Because of the peculiar features of this system we decided to make out own main circuits to get the best results. The advantage of integrating the control system for the stepping motor motions with the data acquisition system has led us to prefer a standard CAMAC solution for the computer interface. On the other hand, the need for a physical separation of the different functions and the difficulty of establishing a priori the number and the type of movements necessary for the different phases of the experiment, have influenced our choice of a system with modular structure and separate functions . In sect. 2 we describe the general layout of the system by emphasizing the distribution of functions and their link in a closed loop configuration . Sect. 3 is dedicated to a description of the circuits built for this system and their specific performance. A brief account
236
5000
Fig. 1. Spontaneous radiation pattern measured in the radial dimension . Experimental points are interpolated by means of a cubic spline algorithm.
Fig. 2. Layout of the experimental hall. Concrete shields delimiting the accessible area are shown as dashed blocks.
M. Castellano et al. / Control system for the LELA experiment
of the way in which the system has performed in measuring the spontaneous radiation is presented in sect. 4. 2. General layout of the control system A schematic view of the control system is shown in fig . 3. The computer controls the stepping motor motions through the control module, a home-made CAMAC circuit with a drive capacity of 15 motors . For each motor two distinct pulses for forward and backward movement reach, through coaxial cables, the power driver ; this is a standard NIM module located inside the experimental bunker . This circuit can drive motors with different characteristics of voltage and speed and can make them execute steps of half or full length, as will be described in the next section. S. M. CONTROLLER MODULE (FOR IS S . M .)
The power driver supplies the current to the four motor phases under control of the step command, and it contains also a circuitry to control a shaft encoder which can be connected with the motor. A four digit counter memorizes the motor steps by counting the driving pulses. If the shaft encoder is connected with the motor the counter increment is conditioned by its consent. The power driver can also be operated manually for testing the movements or setting up the apparatus . The motor counters can be read by the computer. To minimize cables, a multiplexer unit is located inside the experimental bunker. This unit decodes motor addresses from the computer and multiplexes the addressed counter content to a CAMAC input register. The same unit also takes care of all possible end-ofstroke signals generated by the shaft encoder electronics. A stop pulse is sent to the computer, and a status
TTL INPUT TTL CRATE OUTPUT REGISTER CONTROLLER ' REGISTER
II I I I I II
EXPERIMENTAL AREA
(FOR 15 STEPPING
Fig . 3. Schematic view of the control system.
237
MOTORS)
238
M. Castellano et al. / Control system for the LELA experiment
SN74S288 PROMs, is controlled by an AM2909 sequencer . A motor movement program is obtained writing, with CAMAC subaddress A as motor address, a 16 bit word in the 16 x 16 RAM of the module, the most significative bit being the movement direction . The motor address is also inserted into a 16 x 4 FIFO memory . When all the involved movements have been programmed, the subaddress 0 has to be inserted into the FIFO. This subaddress does not correspond to a motor address and is used as a separation word in the FIFO itself. When a start command is issued by the computer, motor addresses are sequentially extracted from the FIFO, addressed RAM content is decremented by 1 and the relative output flip-flop is set. If the decremented RAM content is greater than 0 the number is rewritten into the RAM and the address is reinserted into the FIFO. This operation is repeated until the separation word is extracted from the FIFO. A pulse is then generated and sent to all outputs, which are gated by the output flip-flop. A step is thus simultaneously executed by all programmed motors. FIFO scanning is then repeated, and pulses generated, until no more addresses are contained in the FIFO.
register is strobed, enabling it to reconstruct the stop source. A closed loop control system has thus been implemented . All the movements, once programmed, are automatically executed allowing the computer to make a preliminary analysis of data and/or to interact with the operator for an on line control of the measurement . When the end of movements is reached or a fault signal is detected a CAMAC LOOK-AT-ME interrupt makes the computer resume control. 3 . Circuit description 3.1 . The stepping motors' control module
This circuit, housed in a double unit CAMAC module, controls movements up to 15 stepping motors. For each motor two LEMO connectors are provided for clockwise and anticlockwise movement commands. The number of steps and the movement direction of each motor can be programmed, and a pulse for each step is generated by the respective output . The pulse train frequency may be programmed on a wide range (0.1 Hz to 1 kHz). The circuit, whose logic diagram is shown in fig. 4, is microprogrammed . The sequence of microinstructions, memorized in three
a 3a A F Q
4 6 ) (111 S TEPS PIS
W LL )
MEMORY
n
Q
2R
a U
J W
F. I . F. O. MEMORY
O
W U.
u.
m
Fig . 4. Logic diagram of the stepping motor control module.
CLOCK (5 MHz)
~ä
W w
ADDRESS,
u. u. m
lxW
LL
I18x4)
D: W
EXTERNAL STOP IN
(16X24) MICROSTUCTONS R
W
a W
aw
0
o:W 0 0 Q
o o4
r
OUT
FORWARD PULSES
OUT
BACKWARD PULSES
M. Castellano et al. / Control system for the LELA experiment
The end-of-movement flip-flop is set in the status register and a CAMAC LAM to the computer is issued. The 200 ns microinstruction cycle makes the variable delay introduced between successive pulses completely negligible . In this way it is possible to execute peculiar movements of optical devices such as parallel translation of mirrors, displacement and/or rotation of the optical cavity axis, with great advantage for the results of the experiment. 3 .2 . The power driver module
The power driver module is a two unit NIM module (fig . 5). It drives two independent four phase stepping motors and monitors their position and status through a multiplexed interface to the remote control system. Each driver, whose logic diagram is shown in fig. 6, receives clockwise or anticlockwise motion commands, TTL levels low true, through two coaxial LEMO connectors. The driver can also be manually operated by means of 'two front panel pushbuttons which, if briefly pressed, cause the motor to execute a single step in either direction, and, if continuously pressed, give way to a 100 Hz oscillator which drives the motor to a fixed speed run. In this way it can be fully operated as a stand-alone module. The step generation logic is based on a 32 X 8 PROM.
239
Four PROM bits drive the motor phases currents . Three other bits are used as less significant bits of the 5 bits next address to the PROM itself . The most significant bit of the next address, controlled by a switch on the printed circuit, selects a half or full step movement (0.9° or 1 .8°, respectively). The second one is derived from step command and determines the clockwise or anticlockwise step direction. The possibility of changing the step size allows us to achieve the best compromise between movement speed and accuracy for each application . Current to the four motor phases is derived from an external source, so as to adapt it to specific voltage or current requirements . The driver itself can be set on current or voltage mode by a suitable choice of a resistor in the final power stage. The same circuit has been used in this way with different requirements of power and speed. A differential sqt"^re wave type shaft encoder can be associated to each v it. The two bit status of the encoder, before and after the step execution, is used as a 4 bit address to a look-up table PROM . A forward or backward consent to the step counter or a missing response signal to the alarm system is then generated. If the encoder eletronics is disabled by a front panel switch, the step counter is incremented or decremented directly by the step command. The step counter content is visualized on a four digit display on the front panel and can be read by the
Fig. 5. Frontal view of NIM dual motor driver module .
240
M. Castellano et al. / Control system for the LELA experiment
4ooS/ T
2o0S/T C K Q
IN C
CO
LL P B . pP Q
YC) Q
m
STEP COMMAND LOGIC
SINGLE OR CONTINOUS STEP LOGIC
N W
x a
1
O
LOOK-UP
I TABLE I PROM L .-
I
O 1_
IN C
E . O. S. LOGIC
W
O Q 2
1
W
u IN
°
Q
F F
MISSING RESPONS
FF
N
SHAFT ENCODER LOOK-UP
cLK
a J ó
Lt,
FF
7
I
TABLE PROM
FWD
E.O.S .
R.C. INTERFACE
REMOTE CONTROL
ENABLE CLR
SWD I
TIMING LOGIC
4DECADES COUNTER
FF
FWD/ BWD DELAY YES
NOT
ENC .
Fig. 6. Logic diagram of motor driver module. PHOTOEMITTING DIODE
PHOTOEMITTING
LE PHOTOTRANSISTOR
Fig. 7. Schematic view of a linear movement with LEDs and phototransistor for end-of-stroke logic.
M. Castellano et al / Control system for the LELA experiment
computer through a rear panel CANNON connector. Two end-of-strokes are envisaged for each movement . In order to have maximum flexibility, any logical function of two independent signals can be implemented by jumpers on printed circuits for each end-ofstroke, and used to generate an alarm to the multiplexer unit . The end-of-stroke status is also displayed by two front panel LEDs . For all linear movements used in spontaneous radiation measurements, a coincidence between two optical switches has been used to generate end-of-stroke signals. One of the switches is connected to the linear movement, the second one to the circular movement of the stepping motor axis to ensure an absolute position reference within one motor step (fig. 7) . For linear movements driven by micrometric screws with a 0.5 mm pitch/turn, a single step of a motor of 400 steps/turn corresponds to a displacement of 1 .25 ltm, well below the mechanical resolution of the screw itself . The resetting of the starting point has been tested over a period of many days of continuous running of the linear movements, thus guaranteeing its use as a reference point, within the resolution limits of the whole mechanical system . 3.3 . The multiplexer unit
In each power driver module the counters and the end-of-stroke status are displayed on the front panel to allow safe operation when used in manual mode. The multiplexer unit ensures interrupting and polling capability when operating in remote computer controlled mode . This unit is linked to a CAMAC 1/O register
241
from which it receives a four motor address and two control bits, and to which it sends 16 data bits . No communication handshake with the computer is foreseen . Each double power driver module is linked to the multiplexer unit via 16 data lines, multiplexed between the two counters, 6 fault signals and 4 control lines. All fault signals from the drivers are ORed together to promptly detect a wrong condition and to send a stop pulse to the motor control module. Their status is also strobed in a register so that, by means of a polling routine, the computer can reconstruct the fault source and undertake the correct recovery action . The step counters of the drivers can be independently read or reset, allowing a reproducible starting point to be created, moving forward or backward up to an end-of-stroke signal, and resetting the counter. 4. Conclusions The system described in this paper has been used for a series of preliminary measurements for the implementation of the optical cavity of the free electron laser built on the Adone storage ring . In particular, the measurement of the spontaneous radiation has been instrumental in testing the regularity of the undulator magnetic field and its complete compensation . The measurement was carried out using a monochromator undergoing horizontal and vertical translations in order to scan in steps of 1 mm the farfield pattern radiated by the undulator at a distance of 20 m. The scanning was obtained by using stepping-motors controlled by the system described in the present paper. Hundreds of
6500
6000
5500
5000
I -1 .50
I -1 .00
I -0 .50
I 0 .00
1
0 .56
Fig. 8. Spontaneous radiation peak wavelength X. [A] (vertical axis) as a function of angle 0 [mrad] (horizontal axis). The continuous line represents the theoretical curve.
242
M. Castellano et al. / Controlsystem for the LELA experiment
frequency and position scannings have been carried out for several days . In particular, some polarization mea-
surements have requested circular scannings around a common axis. Some of the collected data have been plotted in fig. 1 .
Owing to the limited lifetime of the stored electron
bunches, the beam has been injected several times before completing the data collection . Therefore, the perfect reproducibility of the detector position has proven
essential for the accurate measurement of the radiation pattern.
A clear evidence of the correct performance of the
system was provided by the return to the reference position after performing the exact number of steps every time the horizontal movement was activated. In contrast, the vertical movement exhibited some amount of nonlinearity which produced a 6 steps deviation from
the initial position after 10 vertical scannings. This
behaviour was due to the complex dependence of the vertical platform position on the angular position of the step-motor shaft. The six step deviation lies within the
limits of accuracy corresponding to a 10 gm resolution. For high precision scanning the end-of-stroke signal was
used to restore the exact position.
The agreement among the measurements performed during several days stands out clearly from fig. 8 where the peak wavelength 7t c is plotted versus the angle 0. X, is related to 0 through the quadratic relation A~= /Y(1+K Z 2 +y 20 2 n Ji o being the undulator pitch, K the effective value of the normalized field intensity, y is the beam energy and n the harmonic order. The continuous line represents
the theoretical curve, which fits the experimental points nicely.
As a final comment we want to remark that the
location of the different functions among several places has proved particularly useful in view of the remote location of the computer.
The system reliability has allowed us to perform
several totally automatized measurements .
References L.R . Elias, W.M . Fairbank, J.M .J. Madey, H.A . Schwettman and T.I . Smith, Phys. Rev. Lett . 36 (1977) 717. D.A .G. Deacon, L.R . Elias, J.M .J . Madey G.J . Ramian, H.A . Schewettman and T.I . Smith, Phys. Rev. Lett . 38 (1977) 892. C. Bazin et al., in : Physics of Quantum Eletronics, vol. 8, eds., S.F. Jacobs, H.A . Pilloff, M. Sargent, M.O. Scully and R. Spitzer (Addison-Wesley, New York, Reading, 1982) p. 89. [4] A. Luccio, in ref. [3], p. 153.
[5] G. Dattoli, A. Marino and A. Renieri, Storage Ring Operation of the Free Electron Laser, Report CNEN 80/22cc (1980).
[6] D.A . Deacon, Basic Theory of the Isochronous Storage Ring Laser, Phys . Rep. 76 (1981) 349. R. Barbini and G. Vignola, LELA - A Free Electron Laser Experiment in Adone, Frascati Report LNF-80/12 (1980). [8] R. Barbini et al ., in : Bendor Free Electron Laser Conf ., J. de Physique C1-44 (1983). M. Castellano, N. Cavallo, F. Cevenini, M.R . Masullo, P. Patteri, R. Rinzivillo, S. Solimeno and A. Cutolo, Nuovo Cim. 81B (1984) 67 . [101 M. Biagini et al., SPIE 453 (1984) 275. [11] M. Biagini et al ., subnutted to Nucl. Instr. and Meth .