IR Demo Project Weekly Report
July 12-16, 1999
Management
Highlights of the week include lasing cw up to 1.72 kW, a full kilowatt above our previous power record. (Details below).
Thanks to all our collaborators and supporters for your assistance in reaching and exceeding the 1 kilowatt design goal for the IR Demo FEL
FEL Installation/Maintenance Activities
Only minor adjustments were made to certain hardware and software as
part of debugging after the six-week down.
FEL Commissioning Activities
On Monday we worked to set up stable, high-current, cw electron beam, achieving up to 3 mA. The cathode quantum efficiency was great, high enough in principle to pull off about 50 mA of cw beam. Trips of the IR waveguide interlocks persisted, ultimately leading us to lower the trip thresholds of the detectors. These interlocks are intended to prevent overheating the warm windows. We still do not understand the cause of the trips, but it appears from a practical viewpoint that they are nothing more than a nuisance. The matter remains under investigation.
On Tuesday we installed the new sapphire mirrors for 3 µm lasing. That evening we restored stable cw current and pushed it to 4 mA.
On Wednesday we restored lasing, ultimately getting close to 500 W upstairs
in the Optical Control Room (OCR). The mirrors behaved very
stably. We ran cw current up to 4.8 mA, at which point voltage
droop on the gun tripped off the machine; i.e., the maximum cw current
we can run under the present configuration is 4.8 mA. We could not,
however, lase appreciably at this current without first improving the momentum
management of the recirculator.
On Thursday we worked to improve the momentum acceptance of the recirculator.
By 1500 we had delivered 670 W cw upstairs in the OCR.
With the estimated 30% losses in the optical transport line, we figured
this corresponded to about a kilowatt power output. Rather than
optimize lasing, we decided to install the power head directly at the
outcoupling mirror of the optical cavity for unambiguous measurements.
At 1744, within an hour after turning the machine back on, we broke
through 1.0 kW. Soon thereafter, mostly by incrementally adjusting
the orientation of the optical-cavity mirrors, we had run the FEL at cw
powers up to 1.72 kW, a full kilowatt above the previous power record,
at a current of 4.4 mA. The limit to going higher was a persistent
trip in one of the beam loss monitors. The machine ran stably at
1.7 kW; for example it ran at 1.65 kW uninterrupted for 14 minutes.
Realizing that at these powers we could deliver in excess of a kilowatt
upstairs, we
decided enough was enough and turned to measuring power output versus
current, the behavior of which was found to be linear. We also
calibrated both the optical transport losses at 3 µm and a newly
installed Brewster window pickoff (at the outcoupling mirror of the FEL)
for use with a second power head installed on the downstairs optical table
outside of the optical beam path.
By this writing (1045 Friday, 16 Jul 99) we have removed the power head
from the optical beam path and reinstalled it in the OCR. The plan
for the day is to take difference-orbit data with the existing machine
settings, a brief stop for a party this afternoon, then lase this
evening to commission certain diagnostics (e.g., the autocorrelator)
in the OCR. All week during owl shifts, electron-beam measurements
were made. The new, finer multislit screen for Multislit #1 (at the
exit of the injector cryounit) works very well in that it significantly
improves the accuracy of transverse-emittance measurements at that location.
The measured emittance there is 6.2 mm-mrad, consistent with previous
injector performance. Time-of-flight measurements could not be
made due to an inability to calibrate the path-length cavities; this is
a new
problem that we will fix. There is also a new, unwanted feature
in the longitudinal phase space -- an energy tail that will, if not removed,
make it difficult to see changes in CSR-induced energy spread.
We plan to track down the cause of the tail (is it the new settings of
the gun
solenoids?). Dispersion in the back leg remains low. The
rms bunch length at the wiggler was measured to be 0.1 mm. Various
other
diagnostic checks were made in preparation for starting the systematicCSR
measurements next week.