MEMS L3G4200D, by ST microelectronics, three axis gyro scopefor satellite dish alignment.
Are the micro electrical mechanical systems (MEMS) three axis, yaw, pitch and roll accuracy good enough to align a C or Ku bandsatellite dish? I will experiment with the L3G4200d gyro one I get the backordered circuit development board; it may be a month or more from now. Briefly,the L3G4200D gyro is ultra-stable three-axis digital output gyroscope with 16 bit-ratevalue data output.The L3G4200D gyro is the same deviceas found in the Apple iPad 2 and Samsung’s Android based Galaxy tablet. Gyro compasses, yaw pitch and roll display has already been incorporated into “APPS”for these devices. Answers have been found, but virtually no documentation exists on details. In the hopes of gaining insight to the ST gyro, I pose thequestion in a group where dish alignment is often discussed. What I desire fromthe L3G2400D gyro is:
1. North seeking capability. This may take hours to complete, but the results hould provide high accuracy to establish true North without using a magnetic compass or GPS receiver. Accuracy should be better that one degree. My main problem is establishing due south within a degree or better. As simple as this sounds, it is nearly impossible without fancy surveying equipment. A magnetic compass is useless. Its accuracy is thrown off by the dish’s metallic effect,even accounting for magnetic declination correction. A true north seeking gyro will provide a true reading, independent of metallic effect, and is easy toapply to the dish alignment result.
2. The L3G4200D must be capable of replacing traditional rotary encoders for motion control. It really simplifies mechanical design when asmall gyro can replace a rotary encoder. In this case the size of the L2G4200Dis 4x4 mm. Compare this size with a typical rotary encoder of over an inch in diameter or larger. Well may not be fair, the L3G4200D needs to be supported bya local CPU and support circuitry that also may exceed a one inch size, but is not dependent on the same constraints required for a rotary encoder’s mounting requirements. There is more freedom with a gyro than a rotary encoder mounting requirements.
The L3G4200D three axis gyro is a rate device. It outputs to the external world degrees/second for yaw, pitch and roll. I need to convert the rate output values to position or degrees. A simple integrator will provide the conversion, however the time between reads must be precisely known. I do not know the best way of configuration the L3G4200D or the best time clock to use. I know windows supports a high performance clock mode with microsecond precision (and have used it often), but it’s also affected by time to interrupt delay and other OS overhead. Ideally the L3G4200D gyro should provide a time stamp synchronized to the actual time of data capture.
For the L3G4200D, the basic concept is simple to understand; however, the details are difficult to unravel and understand. The most time is spent on the details, not the concept.
Ok back up! What the hell is this nut case trying to accomplish.
Ideally I am trying to replace the actuator pulse count with a rotary optical encoder mounted at the dish’s axis of rotation. This is same axis found in HH motors for USALS motion. Using an optical encoder at the dish rotational axis can provide USALS capability even though actuator driven. Take this one step further; the optical encoder can be replaced by a MEMS gyroscope. The result is a polar mount structure than can be improved by a simple MEMS gyroscope. Use the actuator for C band dishes, its strong, rugged and dependable. But, use the MEMS gyro for precise control. Forget about the actuator pulse count; use the output from MEMs gyro for precise PID motion control. Even forget about modified declination tilt, use a small second actuator to provide exact tilt declination per satellite position using the MEMS gyro output. Only one axis is needed, but a second axis will provide conformation.
Sorry for being long winded, however this application has caught my attention with a fury.
I dream of the day when I can get my 6 foot dish aligned and tracking. So far only dismal results have been achieved. OH! My mounting poleis vertical from the ground, I have checked it many, many times using a level and two different types of digital levels; it is correct as well as all otheralignments. However there must be a mistake and I feel only more exotic methods will finally shed some light where the miss alignment is. Part of the problem is the extreme sloppy alignment method for aligning the C/Ku band LNB. Just a set screw? He guys, there must be a better method. Maybe a MEMS Gyro mounted onthe C/Ku band LNB will enable alignment. Watcha you think? Sorry for sarcasm,but when you have spent hours for only dismal results you have got to be annoyed.
Suggestions, corrections, criticisms, even flames are allappreciated.
Rob
Are the micro electrical mechanical systems (MEMS) three axis, yaw, pitch and roll accuracy good enough to align a C or Ku bandsatellite dish? I will experiment with the L3G4200d gyro one I get the backordered circuit development board; it may be a month or more from now. Briefly,the L3G4200D gyro is ultra-stable three-axis digital output gyroscope with 16 bit-ratevalue data output.The L3G4200D gyro is the same deviceas found in the Apple iPad 2 and Samsung’s Android based Galaxy tablet. Gyro compasses, yaw pitch and roll display has already been incorporated into “APPS”for these devices. Answers have been found, but virtually no documentation exists on details. In the hopes of gaining insight to the ST gyro, I pose thequestion in a group where dish alignment is often discussed. What I desire fromthe L3G2400D gyro is:
1. North seeking capability. This may take hours to complete, but the results hould provide high accuracy to establish true North without using a magnetic compass or GPS receiver. Accuracy should be better that one degree. My main problem is establishing due south within a degree or better. As simple as this sounds, it is nearly impossible without fancy surveying equipment. A magnetic compass is useless. Its accuracy is thrown off by the dish’s metallic effect,even accounting for magnetic declination correction. A true north seeking gyro will provide a true reading, independent of metallic effect, and is easy toapply to the dish alignment result.
2. The L3G4200D must be capable of replacing traditional rotary encoders for motion control. It really simplifies mechanical design when asmall gyro can replace a rotary encoder. In this case the size of the L2G4200Dis 4x4 mm. Compare this size with a typical rotary encoder of over an inch in diameter or larger. Well may not be fair, the L3G4200D needs to be supported bya local CPU and support circuitry that also may exceed a one inch size, but is not dependent on the same constraints required for a rotary encoder’s mounting requirements. There is more freedom with a gyro than a rotary encoder mounting requirements.
The L3G4200D three axis gyro is a rate device. It outputs to the external world degrees/second for yaw, pitch and roll. I need to convert the rate output values to position or degrees. A simple integrator will provide the conversion, however the time between reads must be precisely known. I do not know the best way of configuration the L3G4200D or the best time clock to use. I know windows supports a high performance clock mode with microsecond precision (and have used it often), but it’s also affected by time to interrupt delay and other OS overhead. Ideally the L3G4200D gyro should provide a time stamp synchronized to the actual time of data capture.
For the L3G4200D, the basic concept is simple to understand; however, the details are difficult to unravel and understand. The most time is spent on the details, not the concept.
Ok back up! What the hell is this nut case trying to accomplish.
Ideally I am trying to replace the actuator pulse count with a rotary optical encoder mounted at the dish’s axis of rotation. This is same axis found in HH motors for USALS motion. Using an optical encoder at the dish rotational axis can provide USALS capability even though actuator driven. Take this one step further; the optical encoder can be replaced by a MEMS gyroscope. The result is a polar mount structure than can be improved by a simple MEMS gyroscope. Use the actuator for C band dishes, its strong, rugged and dependable. But, use the MEMS gyro for precise control. Forget about the actuator pulse count; use the output from MEMs gyro for precise PID motion control. Even forget about modified declination tilt, use a small second actuator to provide exact tilt declination per satellite position using the MEMS gyro output. Only one axis is needed, but a second axis will provide conformation.
Sorry for being long winded, however this application has caught my attention with a fury.
I dream of the day when I can get my 6 foot dish aligned and tracking. So far only dismal results have been achieved. OH! My mounting poleis vertical from the ground, I have checked it many, many times using a level and two different types of digital levels; it is correct as well as all otheralignments. However there must be a mistake and I feel only more exotic methods will finally shed some light where the miss alignment is. Part of the problem is the extreme sloppy alignment method for aligning the C/Ku band LNB. Just a set screw? He guys, there must be a better method. Maybe a MEMS Gyro mounted onthe C/Ku band LNB will enable alignment. Watcha you think? Sorry for sarcasm,but when you have spent hours for only dismal results you have got to be annoyed.
Suggestions, corrections, criticisms, even flames are allappreciated.
Rob
