MTi 10-series and MTi 100-series Document MT0503P, Revision 0 (DRAFT), 11 Feb 2013 Xsens Technologies B.V. Pantheon 6a P.O. Box 559 7500 AN Enschede The Netherlands phone +31 (0)88 973 67 00 fax +31 (0)88 973 67 01 e-mail info@xsens.com internet www.xsens.com Xsens North America, Inc. 10557 Jefferson Blvd, Suite C CA-90232 Culver City USA phone 310-481-1800 fax 310-416-9044 e-mail info@xsens.com internet www.xsens.com
Revisions Revision Date By Changes A 11 Feb 2013 MHA Initial release 2005-2014, Xsens Technologies B.V. All rights reserved. Information in this document is subject to change without notice. Xsens, MVN, MotionGrid, MTi, MTi-G, MTx, MTw, Awinda and KiC are registered trademarks or trademarks of Xsens Technologies B.V. and/or its parent, subsidiaries and/or affiliates in The Netherlands, the USA and/or other countries. All other trademarks are the property of their respective owners. Xsens Technologies B.V. ii
Table of Contents 1 INTRODUCTION... 1 2 ORIENTATION AND POSITION ACCURACY... 2 2.1 ORIENTATION ACCURACY MTI 10-SERIES... 2 2.2 ORIENTATION ACCURACY MTI 100-SERIES INDOOR TEST... 4 2.3 ORIENTATION AND POSITION ACCURACY MTI-G-700 AUTOMOTIVE TEST... 6 3 GYROSCOPES... 10 3.1 WORKING PRINCIPLE... 10 3.2 ALLAN VARIANCE... 10 4 ENVIRONMENTAL TESTING... 11 4.1 INGRESS PROTECTION (IP)... 11 4.2 VIBRATION IMMUNITY... 12 5 NON-STANDARD CONFIGURATIONS... 13 5.1 AVAILABLE CONFIGURATIONS... 13 5.2 SPECIFICATIONS FOR NON-STANDARD CONFIGURATIONS... 13 Xsens Technologies B.V. iii
1 Introduction This document is the technical datasheet for MTi 10-series and MTi 100-series range of products. The document serves as a support for the performance values specified in Product Data Sheets. In the technical datasheet, validation tests can be found on claimed accuracies. Apart from the performance, this document also highlights important features of the product by substantiating it with technical charts and graphs e.g. Allan Variances and vibration behavior. The following names are used for different products: Product name Alternative naming Year of introduction Legacy MTi MTi 3 rd generation 2005 /MTi Legacy MTi-G MTi-G 3 rd generation 2007 /MTi-G MTi 10-series MTi-10; MTi-20; MTi-30 Direct link on www.xsens.com /MTi-10-series MTi 100-series MTi 4 th generation; MTi-100; MTi-200; MTi-300 2012 /MTi-100-series MTi-G-700 MTi-G 4 th generation; /MTi-G-100-series GPS/INS MTi-G-700 Bookmark Document name Document ID Availability [PL_10s] MTi 10-series product leaflet MT0501P www.xsens.com [PL_100s] MTi 100-series product leaflet MT0502P www.xsens.com [MTI_UM] MTi User Manual MT0605P www.xsens.com Note that this Technical Datasheet is a draft to support the performance claims in the product leaflets of the 4 th generation MTi and to provide more in-depth information about the MTi. The performance tests, although accurately planned and conducted, are not comprehensive and do not cover the wide range of conditions and environments where Xsens products are subject to in actual applications. New performance tests are planned in order to augment this document. Xsens Technologies B.V. 1
2 Orientation and position accuracy Prior to the release of products, Xsens thoroughly investigates the performance of it by subjecting the product to a whole range of tests which includes hardware, software, sensor fusion algorithms, userinterface and overall system testing. The plots and results in this chapter are the result of performance tests and validate the orientation and position accuracy as described in the leaflets of the MTi 10-series (MT0501P) and MTi 100-series (MT0502P). 2.1 Orientation accuracy MTi 10-series An indoor test was done, where an MTi 10-series and a legacy MTi were attached to an imar FSAS tactical grade IMU. The test included running, stationary periods, periods of centripetal accelerations with magnetic distortions. The equipment used is listed below: Brand Product series Product Rate of turn full scale Acceleration full scale Xsens MTi 10-series (4 th generation) MTi-30 AHRS 450 deg/s 50 m/s 2 Xsens Legacy MTi (3 rd generation) MTi AHRS 1200 deg/s 50 m/s 2 imar IMU-FSAS (0.75º/h ref) 450 deg/s 50 m/s 2 Xsens Technologies B.V. 2
The MTi s orientation is compared to the dead-reckoned orientation imar FSAS gyroscope data. This section details the results: there are various filter profiles, the results differ. Especially in yaw the differences are very easy to see. Some filter profiles heavily rely on the magnetic field (e.g. the high_mag_dep filter profile), where e.g. the VRU filter profile doesn't use magnetic field at all. These filter setting profiles were designed keeping in mind certain applications, in which the MTi is used Based on the test above, the following errors are determined: Type of MTi Setting profile Roll/Pitch Yaw (disturbed environment) MTi 3 rd generation MTi 4 th generation Typical error (RMS) Maximum error Typical error (RMS) Maximum error Machine 0.66 deg 5.63 deg 10.27 deg 33.26 deg General 0.45 deg 2.05 deg 10.46 deg 30.73 deg High_mag_dep 0.52 deg 2.20 deg 10.48 deg 26.66 deg Low_mag_dep 0.50 deg 2.19 deg 13.56 deg 26.62 deg Dynamic 0.42 deg 1.72 deg 9.34 deg 27.22 deg VRU 0.47 deg 1.92 deg 32.22 deg (no ref) 47.80 deg (no ref) Notes on roll/pitch: - The trial was very dynamic, which means that orientation is slightly less accurate because of synchronization errors (only for the maximum error). When there is a movement of 400 deg/s and there is a timing difference of only half a sample time of the reference IMU (running at 200 Hz), the error accumulates to 1 deg. The MTi and the reference were hardware-synchronized; so the errors that can be attributed to timing are approximately 0.5 deg. - For specifications [PL_10s], the most applicable filter profile (in this case dynamic ) was used. Notes on yaw: - In the trial, the magnetic field was almost never homogenous. Therefore the yaw accuracy is lower than specified in the product leaflet [PL_10s], which is valid only for yaw performance in a homogenous magnetic field. - The VRU filter profile doesn t use the magnetic field at all to estimate heading. The yaw accuracy in the VRU profile represents gyroscope integration drift. General notes: - As can be seen, there is a significant improvement in performance between the orientation accuracy of the 3 rd generation MTi and the 4 th generation MTi. Xsens Technologies B.V. 3
2.2 Orientation accuracy MTi 100-series indoor test The MTi 100-series stands out in because of its better gyroscopes (see also section 3.2). This results in a better orientation, especially during challenging conditions, such as vibrations and long-lasting accelerations. In an indoor test, the following indoor test was done, where an MTi 10-series and an MTi 100-series were attached to an imar FSAS tactical grade IMU. The test included multiple rotations resulting in centripetal accelerations; magnetic distortions were present continuously. Equipment used: Brand Product series Product Rate of turn full scale Acceleration full scale Xsens MTi 10-series MTi-30 AHRS 450 deg/s 50 m/s 2 Xsens MTi 100-series MTi-300 AHRS 450 deg/s 50 m/s 2 imar IMU-FSAS (0.75º/h reference) 450 deg/s 50 m/s 2 Xsens Technologies B.V. 4
The orientation of the MTi-30 and MTi-300 were compared against the imar FSAS, the orientation error is plotted. Notes on roll/pitch: - Although, the plotted data seems to be comparable for MTi-30 and MTi-300, a closer investigation tells otherwise. The MTi-10 series has an RMS error of 0.19 deg for roll and pitch, the RMS error of the MTi 100-series is 0.12 deg (>50% better). Especially in the maximum error, the MTi-300 shows that it is more robust. The maximum error of the MTi-300 is 60% lower than the maximum error of the MTi-30. - Roll and pitch are coupled and therefore cannot be regarded separately. Pitch inaccuracy between 300 and 375 seconds in the MTi-300 should therefore be combined with the roll accuracy at that point. Notes on yaw: - In the calibrated data signal, notice the long periods where the magnetic field has another direction. The 4 th generation MTi features a heading redefinition that lets the MTi converge to its new magnetic field after e.g. 30 or 90 seconds, depending on the filter state and the filter profile. The tactical grade IMU doesn't have magnetometers at all, so yaw is based on gyroscope (and gyroscope integration drift) only. Heading redefinition allows for a more predictable yaw in a range of applications that experience magnetic distortions. With the heading redefinition, yaw can change considerably and this should be taken into account when integrating. MTi configuration Roll/Pitch Yaw (disturbed environment) (setting profile) Typical error (RMS) Maximum error Typical error (RMS) Maximum error MTi-30 (dynamic) 0.19 deg 0.74 deg 9.16 deg 22.77 deg MTi-300 (dynamic) 0.12 deg 0.46 deg 5.95 deg 15.48 deg Xsens Technologies B.V. 5
2.3 Orientation and position accuracy MTi-G-700 automotive test In the MTi 100-series, the MTi-G-700 is excellent for automotive and other high-velocity applications. The MTi-G-700 compensates for accelerations and can cope better with magnetic distortions with the help of GPS. See for more details the MTi User Manual [MTI_UM]. The trajectory consisted of a drive from the city of Dordrecht to the city of Rotterdam in The Netherlands. The drive includes two tunnels where there was no GPS availability. The first GPS outage lasted for 20 seconds; the second GPS outage lasted for 80 seconds. The velocity was up to 100 km/h; long-lasting accelerations in the horizontal frame (centripetal accelerations, accelerating after a traffic light etc) were up to 0.3 g. Without sensor fusion filtering, the roll and pitch errors would be around 15 deg. Below the trajectory of the test is presented: the gaps in the trajectory represent GPS outages. Figure 1: The trajectory of a test with the MTi-G. The two GPS outages, indicated by yellow circles, are caused by driving in a tunnel. Xsens Technologies B.V. 6
The following equipment was used: Brand Product series Product Rate of turn full scale Xsens Xsens imar Navigation; NovAtel Inc. MTi 100-series (4 th generation) Legacy MTi-G (3 rd generation) Acceleration full scale MTi-G-700 GPS/INS 450 deg/s 50 m/s 2 MTi-G GPS-aided AHRS 300 deg/s 50 m/s 2 IMU-FSAS (0.75º/h reference); OEMV-2 L1/L2/L2C GNSS: 2 cm 450 deg/s 50 m/s 2 A comparison was done of the 3 rd generation MTi-G and the 4 th generation MTi-G-700 against the IMU-FSAS and OEMV2 integrated solution. Notes on roll/pitch: - The two GPS outages that had a detrimental effect in the performance of 3 rd generation MTi- G, both during the GPS outage and upon GPS reacquisition. In the MTi-G-700, these GPS outages have hardly any effect on the performance. - The stability of roll and pitch when GPS is available has increased significantly as can be seen in the beginning of the trial. Notes on yaw: - The filter profile used is the automotive filter profile. This filter profile makes use of GPS course-over-ground to determine yaw. When there is a GPS outage, yaw inaccuracies are likely to be a result of gyroscope integration drift. General notes - The initialization phase for the MTi-G-700 is significantly shorter, noticeable in both roll/pitch and yaw. Xsens Technologies B.V. 7
Next to the orientation, the velocity and position error of the 3 rd gen MTi-G and the MTi-G-700 are shown below. As can be seen in the in the graph when there is GPS availability, position error of the MTi-G-700 as well as the 3 rd generation MTi-G stays within approximately 20 meter. However, when GPS outages occur, the difference in accuracy in both generation of MTi-G s can clearly be noticed. To ensure that the differences were visible, the y-axes are limited to +/- 40 meter only. The maximum errors during the GPS outages are as following: Outage @ 600s: duration 20 seconds @ 1700s: duration 80 seconds Position error 3 rd gen. MTi-G Position error MTi-G- 700 Horizontal Vertical Horizontal Vertical 250 m 25 m 10 m 5 m 850 m 18 m 80 m 18 m Xsens Technologies B.V. 8
Next to position, the MTi-G-700 also outputs 3D velocity, results of field tests are shown below: The vertical estimates of a GPS-L1 receiver are ~1.5 to 2 times worse than its horizontal estimates. The use of accelerometer and barometers make the vertical estimates smoother and more accurate. When GPS outages occur, both MTi-Gs keep estimating the velocity (dead-reckoning). To ensure that the differences were visible, the y-axes are limited to +/- 5 m/s only for horizontal velocity and +/- 1 m/s only for vertical velocity. The maximum errors during the GPS outages are as follows: Outage Position error 3 rd gen. MTi-G Position error MTi-G-700 Horizontal Vertical Horizontal Vertical @ 600s: duration 20 seconds 23 m/s 2 m/s 3 m/s 0.2 m/s @ 1700s: duration 80 seconds 15 m/s 0.3 m/s 3.5 m/s 0.3 m/s Xsens Technologies B.V. 9
3 Gyroscopes 3.1 Working principle The working principle of a MEMS gyroscope is explained on this webpage: http://www.analog.com/library/analogdialogue/archives/37-03/gyro.html 3.2 Allan Variance The Allan Variance is a common graph to visualize the characteristics of sensors, especially gyroscopes 1. Depicted below is the Allan Variance of the gyroscopes of the MTi 10-series and the MTi Figure 2: The Allan Variance of the gyroscopes in the MTi 10-series and the MTi 100-series 100-series. The noise density of the MTi 100-series is approximately 3 times lower than the noise density of the MTi 10-series. Noise of both gyroscopes is Gaussian (white) noise, indicated by the -1/2 slope till the cluster time of 10 seconds. Also note that the bias of the MTi-100 is stable for a longer time, making dead-reckoning of the gyroscopes much more accurate: this especially of importance in e.g. the MTi- 20 or MTi-200 or in filter profiles of the MTi-30 and MTi-300 that don t use magnetometers for heading reference. 1 N. El-Sheimy, H. Hou, and X. Niu, Analysis and Modeling of Inertial Sensors using Allan Variance IEEE Transactions on Instrumentation and Measurement, vol. 57, no. 1, pp. 140-149, January 2008 Xsens Technologies B.V. 10
4 Environmental testing The MTi s aluminum casing and the PCB design, including connectors and connections is specifically designed for harsh environments, such as humidity, vibrations and shocks. 4.1 Ingress protection (IP) All products of the MTi are designed to IP67. Note that the main connector of the MTi and the GPS antenna connection (SMA) of the MTi-G-700 are IP67 rated only when mated. The receptacle parts of the main connection and the SMA connection are specially designed to IP67. The top casing of the MTi 100-series is perforated with three holes in one of the sides. These holes are designed to allow the atmospheric pressure to be measured by the internal barometer. To make sure that the MTi 100-series has sufficient ingress protection, an IP67-rated vent is applied on the inside. Figure 3: the IP67 rated vent allows the internal pressure of the MTi 100-series to adopt the atmospheric pressure Xsens Technologies B.V. 11
4.2 Vibration immunity Next to the signal processing pipeline of the MTi and the vibration rejecting gyroscopes of the MTi 100-series, special attention has gone into the hardware design of the MTi. This section highlights these features. Xsens has tested the 4 th generation MTi against vibrations according to MIL-STD 202-204D (A); the results are shown below. There was no data loss during the measurements, which indicated that the MTi is immune to vibrations according to the MIL-202-204D A standard. During the test, the MTi was mounted in three different orientations (Z up, X up, Y up) 3x 15 minutes a sweep of 10-500-10Hz with amplitude of 10g peak to peak Figure 4: Accelerometer signals as recorded on the vibration table. In the top chart the output of the accelerometers is shown in bits (in this MTi 15000 and 60000 bits correspond to approximately -5g and 5g respectively). In the bottom chart, the power spectrum is shown (two overlaid lines, as the sweep went from 20-500 Hz and back). No samples were missed at all during the tests, so the MTi passed the test. Xsens Technologies B.V. 12
5 Non-standard configurations Each MTi has 4 configurable specifications: gyroscope full range, accelerometer full range, the interface protocol and the casing. 5.1 Available configurations The standard product code is MTi-###-2A5G4, where ### is the product type (e.g. 30 for an AHRS, 200 for a VRU or G-700 for the GPS/INS). Below is a list of possible options and configurations per specifications. MTi- ###- * A G~ -% Product Interface * Accelerometers Gyroscopes ~ Casing % 10: IMU 2 = RS232 A5 = 5g G4 = 450 deg/s No suffix = encased 20: VRU 6 = RS422 A8 = 15g G0 = 1000 deg/s -O = OEM board 30: AHRS 100: IMU 200: VRU 300: AHRS -G-700: GPS/INS Although all combinations are theoretically possible, not all products may be available from stock and surcharges may apply. Consult with Xsens sales department at sales@xsens.com to learn more. 5.2 Specifications for non-standard configurations Most specifications as in the data sheet are valid for all configurations, except for the two parameters listed here: - Full range of the accelerometers and gyroscopes differ according to the table in 5.1. - The noise density of the 15g accelerometer is approximately 30% lower than the noise density of the 5g accelerometer. Xsens Technologies B.V. 13