On 2/11/99, I took two of the rotator encoder read-heads to the Heidenhain office in Schaumburg, IL, to verify that the ERA 880C read-heads were functioning properly and to understand why we were having such a difficult time aligning these heads to the rotator tape. I worked with Rick Glos (VP-Sales), Tom Wyatt (Area Sales Manager), and Dan Braasch (Service Supervisor) to check the strength of the reference pulse signals and to understand how head misalignment affects the performance of the encoder.
Dan put one of the readheads on a bench setup and took about 30 seconds to align the head and get good output signals from both the counting track and the reference pulse. However, upon closer inspection Dan noted that while the crossing point for the reference signal was within spec, it was higher than he thought it should have been under ideal bench conditions. Ideally, the crossing point should be at zero when looking at the reference pulse signal on an analog oscilloscope. If the reference pulse signal looks like a "greater than" sign (>), then the crossing point is the apex of the V. If the reference signal is looser and looks more like a figure eight (8), then the crossing point is where the lines cross in the middle. In either case, we were seeing this crossing point at about 0.15V on the scope. There is an internal pot on the readhead that adjusts the reference signal bias voltage, so Dan adjusted this pot until the reference signal crossing point was at zero. Dan commented that while 0.15V is within the calibration specification, starting out with a bias this high may eat up some of the misalignment tolerance of the head. Thus, if the mechanical alignment was a little out, we might see a signal on the scope that wouldn't be adequate enough to fire the reference indicator on the Heidenhain PWM-8 phase meter. This is exactly what we were seeing at APO on Feb 4-5.
We installed the second readhead on the bench and found that the reference pulse crossing point was also set at 0.15V, so Dan adjusted the bias voltage on this head as well. Since all five of the heads we purchased were from the same lot, Dan says it's likely they were all calibrated on the same bench and are all likely to behave the same way. Although Dan showed me how to adjust the reference pulse bias voltage, he agreed to recalibrate the remaining three read-heads for us, so I will return them to him after we install and test the two heads he just calibrated. Calibrating them in-situ on the telescope will be too difficult given the location of the readheads.
We also played with angular misalignment and found that if the face of the head is not parallel with the face of the tape, then the magnitude of the reference signal varies. If the head is tipped such that the top of the head is farther from the tape than the bottom, the magnitude of the reference pulse decreases. When Scott Smith of Heidenhain was with me at APO, he noted that the amplitude of the reference pulse was smaller than it should have been. This suggests that our tape may not be perpendicular to the read-head mounting surface and that we may have to shim the head to adjust the angle. We also need to check the radial run-out of the tape along the top and bottom edges to make sure we don't have any wobble in the tape. If the tape pitches in and out, then we may occassionally miss reference marks due to the angular misalignment with the head.
Finally, we measured the vertical position of the head with respect
to the tape when the head and tape were "ideally" aligned. The distance
from the top of the head to the top edge of the tape was 0.685".
We can use this dimension to determine the ideal thickness of the shim
beneath the head mounting brackets. We also played with the vertical
and found that we could move the head approximately +/-.030" without compromising the output signals.
Once the revised telescope engineering schedule is worked out, I will schedule a trip to APO to install and align the two heads that have been re-calibrated. If we're successful with the alignment, then I will return the remaining three heads to Heidenhain for recalibration.
I want to note that Rick Glos, Tom Wyatt, and Dan Braasch at Heidenhain were all extremely helpful and supportive in identifying the source of our problems and in providing a solution that I'm confident will work. I also want to acknowledge the help given to me by Scott Smith last week at APO.