The SIBYLS beamline Kohzu monochromator has often exhibited hysteresis when moving the theta 2 motor. This manifests itself in a decaying beam and suboptimal intensity. In order to fully understand the problem it helps to have a general understanding of the motors in the monochromator. The figure below illustrates the various motors and their relative motions.
The axis of the Theta motor lies on the surface of the first crystal and the angle of Theta determines the wavelength of X-rays that will be selected for a particular experiment. The second crystal is designed to take that monochromatic X-ray beam coming from the first crystal and reflect it down the beam pipe to the SAXS or MX endstation. The second crystal is ~30 meters from the crystal sample when doing a crystallography experiment and it is critical that the um-sized X-ray beam be aimed directly at the small crystalline samples. We use the Chi 2 motor on the second crystal to steer the beam from left to right (horizontal beam steering) and we use the M2 mirror to move the beam up and down (vertical beam steering). The axis of the Theta 2 motor lies on the surface of the second crystal, and it is used to align the surface of the second crystal so that it is perfectly parallel to the first crystal thus maximizing the flux of the X-rays that exits the monochromator. Typically we tuneup the beamline before a user starts their shift and during this tuneup procedure we optimize the angle of the theta 2 motor by monitoring an ion gauge positioned at the exit tube of the mono (Imono Out) and maximize the current. If we then monitor the Imono Out gauge for several hours after the tuneup procedure is complete we sometimes see that the intensity of X-rays falls precipitously.
We have always been concerned about this behavior, but we’ve never really understood what was causing it. Sometime in early 2006 we decided to take apart the mono to inspect the theta 2 motor. Ultimately we replaced the piezo dummy with the original piezoelectric motor that was designed for the monochromator. It is a Digital Piezo Translator (DPT) model manufactured by Queensgate Instruments in the UK. You can see the difference between the dummy and the actual piezo in the photograph below.
We then inspected the plate on the end of the dummy piezo and the ball bearing situated on the end of the stepper motor drive shaft. We noticed significant pitting on both the push plate and the ball bearing. You can see the pitting on the piezo push plate in the photographs below. We thought that this could certainly be the cause of our theta 2 problems. If the ball bearing was moving around in this divet as the theta 2 stepper motor was being adjusted then we thought this was a reasonable explanation for the theta 2 hysteresis.
We polished the push plate, installed the piezoelectric motor, and replaced the ball bearing with a silicon nitride bearing which is much harder than the original stainless steel bearing. Our hope was that the harder bearing and freshly polished push plate surface would alleviate our hysteresis problems. This didn’t really help and we continued to observe significant hysteresis when tuning up the rocking curve with the stepper motor.
We then decided to use the newly installed piezo to do our tuneups under the impression that for some reason the stepper motor was the cause of the hysteresis. The photo below better illustrates the two motors which control the theta 2 stage and their relationship to each other.
The two photos above show the theta 2 stage. The second monochromator crystal is attached to the left half of the stage. The right side of the theta 2 stage is stacked on top of the Z2 and Y2 stages which are in turn attached to the main theta stage. You can see that the two halves of the theta 2 stage are held together by a heavy duty spring. The theta 2 angle can be adjusted in two ways. The first is to drive the theta 2 stepper motor which screws a shaft, containing a ball bearing at its tip, into a brass block. The ball bearing pushes against a steel plate at the tip of the theta 2 piezo “motor”. This causes the two halves of the theta 2 stage to be pushed apart. The second method is to adjust the theta 2 angle is to activate the theta 2 piezo “motor” which then pushes against the theta 2 stepper motor. Both the stepper motor and the piezo motor achieve the same result which is to force the two halves of the stage apart thus adjusting the theta 2 angle. The primary difference between the two motors is the speed and range of motion that can be achieved. The stepper motor is relatively slow, but can move very large distances. Whereas the piezo motor is essentially instantaneous, but only has a total range of 70 um.
As previously mentioned, we found that when performing a tuneup using the stepper motor there was a significant settling time (typically several hours) when theta 2 would “fall off the rocking curve”. This was observed as a decay in the Imono Out signal and indicated to us that the second crystal was drifting farther and farther from being parallel to the first mono crystal. We believed this odd behavior to be some “feature” of our theta 2 stepper motor. Our solution for the past couple of years has been to do a rough tuneup with the stepper motor, let things settle, and then do a fine tuneup with the piezo. This has worked reasonably well, but never really solved our original problem. Even when only using the piezo motor we would still on occasion see theta 2 fall off its rocking curve.
For various experimental reasons we are now revisiting the theta 2 problem. So during this last 6 week autumn shutdown we opened the monochromator to replace the piezo push plate and the Silicon Nitride ball bearing. We again saw similar pitting on the push plate. However, this time the ball bearing remained un-pitted. I guess the Silicon Nitride was hard enough to withstand the forces that were wearing away the surface of the push plate.
Our current working hypothesis is that the spring which forces the two halves of the theta 2 stage together was severely over tensioned. If you examine the hook on the left side of the theta 2 spring in the photos below you will notice that is appears the spring is under so much tension that it is starting to bend open!
This has resulted in the Silicon Nitride ball bearing, that acts as the flexure for the two halves of the theta 2 stage, essentially drilling a pit in the surface of the push plate. So we loosened the spring by sliding the two spring support blocks closer together. The spring is still tensioned enough to hold the two halves of the stage together, but hopefully the tension will not be so overwhelming that it causes the theta 2 stepper motor to drill a hole in the push plate.
Results will be posted once we’ve thoroughly analyzed the situation.