RELATED APPLICATION 0001 This is a continuation-in-part application of International patent application PCT/DE 00/03730, filed Oct. 20, 2000 and claiming priority of German patent application 199 50 869.0, filed Oct. 22, 1999.

DESCRIPTION 0002 The invention relates to a method for measuring geometric characteristics of workpieces on coordinate measuring machines or digitalizing machines by means of measuring apparatus, which supply measurement information as to geometric characteristics of the workpiece relative to the measuring apparatus, which is, in short, often characterized as measuring sensors.

INITIAL COMMENT 0003 In the following, the abbreviation CMA is used for coordinate measuring machines as well as for digital machines.

STATE OF THE ART 0004 Today, mostly CMAs are used for detecting three-dimensional geometric variables. These CMAs are mostly equipped with switching probes which, when contacting the surface, generate an electric signal. With this signal, the positions of the individual linear or rotational axes of the CMAs are frozen, that is, the axial positions are copied into an output register. This data is then used in the known kinematic of the CMA, including the probe, to generate a measuring point in a stationary coordinate system. When generating a measuring point, partially more extensive error compensations are carried out, such as probe ball radius correction, geometric or thermal machine position error corrections. The basis for the fault correction and, in part, also for the determination of the kinematics, is mostly a calibration process. 0005 If several points are sequentially measured in the same fixed coordinate system, the relative position of the measuring points in the spatially fixed coordinate system to each other is a measure for the relative position of the contact points on the workpiece surface. In this way, geometric quantities can be determined on the workpiece. 0006 To an increasing extent, measuring sensors are preferably used in lieu of switching probes for detecting free-form surfaces. Contacting as well as contactless sensors are used. These sensors determine the position of the measuring point on the workpiece relative to the recording thereof, that is, in a sensor fixed coordinate system, hereinafter known as RST coordinate system. A sequence of measuring points is, as a rule, detected during the movement of the CMA and one speaks of a scanning operation. 0007 For computing a measurement point in a stationary coordinate system, the sensor information in the RST coordinate system must be coupled in time synchronism with the information for the position of the sensor in the stationary coordinate system. In methods known at the present time, this takes place in that the result of the measurement of the sensor in the RST coordinate system is detected in time as close as possible to the result of the sensor position determination by the CMA in the stationary coordinate system in a mostly digital signal processing unit and, thereafter, a measuring point in the stationary coordinate system is computed from these two individual informations.

DISADVANTAGE OF THE KNOWN METHOD AND ADVANTAGE OF THE METHOD ACCORDING TO THE INVENTION 0008 In the known method, so that the information from the sensor and the CMA originates indeed at the same time point, the times for obtaining the measurement value, the measurement value processing and the measurement value transmission, must have the same magnitude for the sensor and the CMA so that the time synchronous coupling of the results corresponds also to the same measurement time point. 0009 Otherwise, the information of the sensor would be coupled to a position of the sensor which it did not have at the time point of obtaining the information and this would lead to an incorrect computation of the measuring points with the exception of some special cases. 0010 There are systems known wherein, for unequal but respectively constant signal running times, a correction is preferably carried out via a delay path of the more rapid information utilizing a ring buffer; these systems, however, are subject to disturbances because the assumption of constant signal running times (especially for optical sensors having complex signal processing) cannot be guaranteed and the signal transmission time up to the point of the signal processing unit (at which both informations are detected simultaneously) is generally not constant. This applies primarily when the signal processing unit used is a complex multitasking operating system such as Windows NT. 0011 Furthermore, all systems have as a precondition that the measurement data rate of sensor and CMA is the same, that is, a result from the sensor is coupled to a result from the CMA. 0012 The method of the invention overcomes these disadvantages in that the time point of the measurement is determined by an electrical signal (trigger) for the CMA as well as for the sensor. 0013 All CMAs, which are designed for operation with a switching probe, permit this function in that the trigger signal is treated like a signal of a switching probe. The sensors usable in accordance with the invention likewise have to exhibit a possibility for synchronization. This can be a trigger input for the exact determination of the measurement time point as well as an output signal (SYNC), which indicates the exact measurement time point. In the first case, a measurement data detection system needs to contain a trigger source in accordance with the method of the invention; in the second case, the sensor itself would function as a trigger source. 0014 A signal processing unit in accordance with the method of the invention ensures that precisely one measurement result from the sensor and one measurement result from the CMA per trigger is further processed. 0015 In contrast to known methods, different signal running times no longer have an effect because, in the signal processing unit, the simultaneousness of the measurement is determined by the sequence of the results and no longer by the time point of the detection of the results. Stated otherwise, the n-th result after an initialization of the CMA and the n-th result after an initialization of the sensor originate reliably at the same time point because they were triggered by the same trigger. 0016 Since especially contactless sensors are essentially more dynamic than the CMA, different measurement data rates of sensor and CMA are purposeful and possible with the method according to the invention, preferably when the ratio of the measurement data rates is a whole number. In this case, for the computation of measurement points, the results of the CMA are interpolated in accordance with known methods so that, after the interpolation, the same amount of information is present from the CMA and from the sensor and can be processed further as with the same measurement data rate. 0017 In order to make the identification more reliable as to which result belongs to the n-th trigger (for example, to detect possibly occurring signal transmission defects), the sensors as well as the CMA should, in a preferred variation of the method of the invention, in addition to its other informations, transmit a TAN-NR (transaction identifier) which is increased with each trigger and is again reset when a maximum number is exceeded. In this way, for a coupling of the CMA data and the sensor data, synchronism can be checked in a simple manner and synchronization reached. 0018 In a preferred variation, at least the sensor and preferably also the CMA have a FIFO output memory which ensures that even under not real-time capable operating systems (such as Windows NT), a complete and clear transmission of each individual information per trigger is ensured. Especially with the high data rate, which is possible for contactless sensors, it is thereby prevented that the signal processing unit (because of loading with other tasks) cannot process the signal transmission within the time pregiven by the measurement data rate and therefore individual informations are not even processed further. 0019 The FIFO output memory, especially in combination with the TAN-NR, permits, on the one hand, a very disturbance-free operation compared to present-day solutions and, on the other hand, it permits the information from sensors or CMA to be transmitted in selectable block sizes for several triggers instead individually per trigger which significantly reduces the load of the signal processing unit because of the data transmission. 0020 The method according to the invention therefore permits a significant increase of the measurement data rate with a simultaneous increase with respect to disturbance reliability and a reduction of the load of the signal processing unit without a significant increase of complexity of present-day conventional CMAs.

SUPPLEMENTARY DESCRIPTION 0021 In the following, the present invention will be described in greater detail with respect to the embodiments shown in the figures. 0022 FIG. 1 shows a block diagram of a first embodiment of the invention; and, 0023 FIG. 2 shows a block diagram of a second embodiment of the invention.

0024 The measuring system 1 in FIG. 1 includes a coordinate measuring apparatus 2 (CMM), a measuring sensor 3 (measuring probe), a signal processing unit 11 (data processing unit) and an external trigger source 4. The coordinate measuring apparatus includes its own internal control 5 (CMM-control) and the measuring sensor 3 includes its own internal control 8 (MP-control). 0025 The measuring sensor 3 is accommodated on the coordinate measuring apparatus 2 in a manner known per se and is rotatable via this apparatus in at least three mutually perpendicular spatial directions as well as about additional rotational axes as required. 0026 The controls 5 of the coordinate measuring apparatus 2 control the movements of the coordinate measuring apparatus 2 and therewith the positioning and, if required, the orientation of the sensor 3 in space and simultaneously control the generation of measurement values of the coordinate measuring apparatus. In a corresponding manner, the control 8 of the sensor 3 controls possible internal movements or measuring force generators within the sensor 3 as well as the measurement value generation of the sensor. 0027 The controls (5, 8) of the coordinate measuring apparatus and of the sensor are autarch, that is, both controls have their own clock generator and work frequencies whereby also the repetition rates of the measured value generation are determined in each case. Since the sensors usually exhibit a higher dynamic than coordinate measuring apparatus, the clock frequencies of the coordinate measuring apparatus and the sensor 3 are different. They distinguish from each other typically by one order of magnitude. Typical clock frequencies for coordinate measuring apparatus are approximately 100 Hz and typical clock frequencies of sensors 3 are approximately 3 KHz. 0028 The two controls have respective trigger inputs (6, 7) via which the measurement value generation can be triggered; that is, when a specific trigger signal, for example a rising signal flank, is applied, the respective actual adjusting values and measurement values are read out and are read out via data outputs (9, 10) to the signal processing unit for further processing. With this further processing, the measurement values of the coordinate measuring apparatus 2 and the measured values of the sensor 3 are coupled to each other. The measured value recordation of the coordinate measuring apparatus as well as of the sensor is triggered by the same trigger signal. For this reason, it is ensured that the measurement value generation takes place simultaneously and, accordingly, measured values, which belong to each other in time, are generated with the coordinate measuring apparatus and with the sensor. 0029 In the embodiment of FIG. 1, the trigger output lies directly on the trigger input 6 of the coordinate measuring apparatus as well as on the trigger input 7 of the sensor 3. In this case, the repetition rate of the trigger signal, which is generated by the trigger source 4, may not be greater than the maximum clock frequency of the coordinate measuring apparatus. It is, however, also possible to switch a frequency multiplier or frequency divider into one of the two signal lines in order to utilize the higher dynamic of the sensor 3. 0030 In the embodiment of FIG. 2, the measuring sensor 3 has an internal trigger source 4 whose output signal 17, on the one hand, triggers the measurement value generation of the sensor 3 via the trigger input 7 of the sensor control 8 and, on the other hand, is applied to the trigger input of the coordinate measuring apparatus 2 via a frequency-multiplier/frequency-divider 12. The pulse repetition rate of the trigger 4 is adapted to the dynamic of the sensor 3 and the frequency-multiplier/frequency-divider 12 is so matched that the pulse repetition rate of the trigger signal is adapted to the lower repetition rate of the coordinate measuring apparatus. The pulse repetition rate of the trigger signal is, as a rule, too high by one order of magnitude. Additionally, the frequency-multiplier/frequency-divider 12 can generate a constant time delay in order to thereby compensate constant running time differences of the trigger signal. 0031 The sensor 3 as well as the coordinate measuring apparatus each have, in addition to their own controls (5, 8), an identifier generator (14, 19). Each of these two identifier generators can be relatively simply configured as high-frequency counters, for example, with a counting rate of 10 MHz. The high-frequency counter counts the incoming trigger signals in each case. To each measured value, which is generated by the coordinate measuring apparatus 2 or by the sensor 3, the corresponding current count of the counter is then added as identifier whereby later a clear assignment of the measurement values of the coordinate measuring apparatus and of the sensor can be guaranteed with the signal evaluation in the signal processing unit 11. These measured values correspond to each other and are accordingly triggered simultaneously. The counters can operate cyclically, that is, when reaching the maximum number, the counters again start counting anew from 1. 0032 The coordinate measuring apparatus 2 as well as the sensor 3 have data stores (13, 15) wherein the respective measured values can be intermediately stored supplemented by respective identifiers. The capacity of the data stores (13, 15) should be adapted to the corresponding identifier generators (14, 19) so that the identifier is clear within each intermediately stored set of measured values, that is, the same identifier is not assigned to two measured values triggered at different time points. 0033 The data stores (13, 15) are preferably configured as FIFO stores. In addition, the coordinate measuring apparatus 2 has a unit 16 with which several measured values of the coordinate measuring apparatus are grouped together to blocks in advance of the data transmission in order to relieve the signal processing unit 11 because the data transmission does not have to take place continuously. Here, it is especially practical to provide in the sensor for a corresponding unit for forming data blocks in advance of the data transmission even though this is not shown in FIG. 2. 0034 In addition to the already mentioned trigger inputs (6, 7), the trigger output 17 of the sensor 3 and the measurement data outputs (9, 10), the coordinate measuring apparatus 2 and the sensor 3 include two further outputs (18, 20) which, to some extent, trigger the signal transmission. Signals are present at these outputs (18, 20), respectively, when sensor measured values are present, that is, when data packages are present ready for the data transmission. The application of these signals indicates to the signal processing unit 11 that a data transmission should take place and this unit then triggers the data transmission. 0035 The coupling of each measured value with an identifier, which is assigned only to this measured value and which is, in turn, determined by the incoming trigger signals, permits erroneous time allocations to be excluded, which, for example, would otherwise be unavoidable in an overtriggering. If, for example, an additional trigger signal is generated, for example, because of a disturbance source (which trigger signal however does not trigger a measured value generation because it followed the previous trigger signal at too small a time distance), then the identifier, which is generated by this additional trigger signal, is not assigned to a measured value and is correspondingly not stored or transmitted. In a later evaluation in the signal processing unit, a conclusion can be drawn from the missing identifiers that the corresponding measured values are not present and, because of this information, the measured values of the sensor and of the coordinate measuring apparatus, which are triggered simultaneously, can be correctly assigned to each other. 0036 Of course, the frequency divider ratio of the frequency-multiplier/frequency-divider 12 must be considered in the assignment of the respective measured values based on the identifiers. If the frequency divider ratio is, for example, {fraction (1/10)}, that is, the repetition rate of the trigger pulses arriving at the coordinate measuring apparatus amounts to {fraction (1/10)}th of the repetition rate of the trigger pulses arriving at the sensor 3, then the measured values of the sensors having the identifiers 1, 11, 21, 31 . . . are to be assigned to the measured values of the coordinate measuring apparatus having the identifiers 1, 2, 3, 4 . . . when each trigger pulse has led to a measured value. If, in contrast, no measured value is transmitted with the identifier 11, then it follows from this that the corresponding trigger pulse has not led to a measured value and, accordingly, the measured values of the sensors having the identifiers 1, 12, 22, 32 . . . are to be assigned to the measured values of the coordinate measuring apparatus having the identifiers 1, 2, 3, 4 . . . . 0037 Of course, also other identifiers can be generated and added to measured values in lieu of the previously described counts of the counters. 0038 If, in lieu of an internal trigger of the sensor, an external trigger is used, then also this external trigger can have an additional counter which assigns a count of a counter to respective ones of the originally generated trigger pulses. With a transmission of these respective counter counts to the signal processing unit, the respective identifiers of the measured values and the external counts of the counter can be checked as to plausibility. 0039 In lieu of using an internal trigger of the sensor, it is, of course, also possible to provide a trigger output on the coordinate measuring apparatus whose signal output then triggers the measured value recordation of the sensor, if required, after a suitable frequency multiplication.