Abstract
A simple low-profile waveguide load pull tuner uses a horizontally only moving reflective probe and an adjustable attenuator placed between the DUT and the tuning probe. Both controls are using stepper attenuators mounted on the waveguide. This simplifies the design and minimizes the cost of manufacturing. The tuner has wideband 50 Ohm tuning capability allowing minimum risk of transient spurious oscillations. A de-embedding adapter removal calibration method allows generating thousands of calibrated points in a small fraction of the time needed for calibration of the full permutations.
Claims
1. A load pull tuner system comprising: A waveguide load pull tuner comprising: a waveguide transmission line having two broad walls and two narrow walls, a test and an idle port, at least one slot centered along a broad wall, and a remotely controlled mobile carriage travelling along the waveguide and holding a reflective tuning probe which is inserted into the at least one slot at a fixed penetration, and a RF energy-absorbing sliver, remotely insertable at adjustable penetration into the at least one slot; wherein the RF energy-absorbing sliver is placed at a fixed position along the waveguide between the test port and the reflective tuning probe; and a waveguide load pull tuner calibration method comprising: connecting the tuner to a pre-calibrated vector network analyzer, measuring s-parameters of the tuner at a multitude of penetrations of the RF energy-absorbing sliver and positions of the reflective tuning probe along the waveguide and saving in a calibration file for later use.
2. The load pull tuner system of claim 1, wherein the tuning probe of the waveguide load pull tuner is inserted into a first slot, which is at least one half of a wavelength long at a lowest frequency of operation of the waveguide load pull tuner, and the RF energy-absorbing sliver is inserted into a second slot, which is placed between the test port and the first slot.
3. The load pull tuner system of claim 1, wherein the RF energy-absorbing sliver and the tuning probe of the waveguide load pull tuner are cascaded and inserted into the same slot, and wherein the RF energy-absorbing sliver is inserted between the test port and the reflective tuning probe.
4. The load pull tuner system of claim 1, wherein the RF energy-absorbing sliver of the waveguide load pull tuner is an oval disc rotating eccentrically around an axis perpendicular to the slot and inserted gradually into the slot between full extraction and maximum penetration.
5. The load pull tuner system of claim 4, wherein the axis controlling the RF energy-absorbing sliver of the waveguide load pull tuner is remotely controlled by a second stepper motor and gear.
6. The load pull tuner system of claim 1, wherein the mobile carriage of the waveguide load pull tuner is remotely controlled by a first stepper motor and gear.
7. A calibration method for load pull tuner system as in claim 1, wherein the waveguide load pull tuner is connected with a pre-calibrated vector network analyzer, comprising the following steps: a) the RF energy-absorbing sliver is withdrawn out of the waveguide, defining a rotation angle ?o=0; b) the reflective tuning probe is moved close to the test port, defining a horizontal position Xo=0; c) s-parameters of the waveguide load pull tuner are measured and saved in an init matrix [S0]; d) s-parameters of the waveguide load pull tuner are measured for a multitude of rotation angles ?m of the RF energy-absorbing sliver between a minimum value ?min, corresponding to withdrawal, and a maximum value ?max corresponding to maximum insertion and saved; e) the RF energy absorbing sliver is withdrawn back to angle ?o=0; f) s-parameters are measured for a multitude of positions Xn of the reflective tuning probe and saved; g) the invers init matrix [S0].sup.?1 is cascaded with the s-parameters of step f) and saved; h) all permutations of s-parameters of steps d) and g) are cascaded and saved in a tuner calibration file comprising tuner s-parameters Sij (?m, Xn) for {i,j}={1,2}.
Description
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
(1) The invention and its mode of operation will be more clearly understood from the following detailed description when read with the appended drawings in which:
(2) FIG. 1 depicts prior art: a typical automated load pull test system.
(3) FIG. 2 depicts prior art: a perspective view of a capacitively coupled, vertically adjustable RF parallelepiped tuning probe (slug) and the relevant dimensions and parameters of the operation.
(4) FIG. 3 depicts prior art: front view of a single carriage/single-probe slide screw tuner.
(5) FIG. 4 depicts cross section of remotely controlled adjustable attenuator using a absorbing sliver insertable into the waveguide.
(6) FIG. 5 depicts cross section of a fixed penetration reflective probe and remote control thereof.
(7) FIGS. 6A through 6B depict the transmission and reflection of the adjustable attenuation of the gradually insertable absorbing sliver over a WR-12 (60-90 GHz) frequency band; FIG. 6A depicts the transmission and FIG. 6B depicts the reflection.
(8) FIG. 7 depicts a calibration setup of the tuner.
(9) FIG. 8 depicts a flowchart of the de-embedding calibration algorithm of the fixed penetration tuning probe tuner with adjustable attenuator.
(10) FIGS. 9A through 9C depict a cross section of the tuner; FIG. 9A depicts the overall tuner, FIG. 9B depicts the attenuator section and FIG. 9C depicts the fixed penetration tuning probe section.
(11) FIG. 10 depicts the reflection factor control and trajectories of impedance synthesis on the Smith chart.
DETAILED DESCRIPTION OF THE INVENTION
(12) This invention discloses a radio frequency (RF, microwave), computer-controlled impedance tuning system, suitable for load pull measurements, comprising at least one electro-mechanical impedance tuner and the calibration method thereof. The tuner (FIGS. 4, 5 and 9) uses a low loss waveguide transmission airline 904, which includes at least one slot 93, 96 placed centered 47 on a broad wall of the waveguide and allowing the absorbing attenuation sliver 44, 92 and the reflective tuning probe 95, 55 to either penetrate gradually or travel along the waveguide to control amplitude and phase of the reflection factor. The overall tuner can be seen in FIG. 10.
(13) The horizontal control of the fixed penetration probe 55 in the airline 54 is best accomplished using linear electric actuators (see ref. 8) or a remotely controlled fixed motor 53 and an ACME screw 52 moving the mobile carriage 50 along the waveguide 54 traversing the slot 51. This movement control of the tuning probe 55 is much simpler than a cumbersome vertical axis and offers significantly lower overall tuner profile. The cross section of this segment is shown in FIG. 9C and the horizontal movement towards 902 and away 903 from the test port.
(14) The attenuation control (FIG. 4) uses eccentric rotation control 43 of the absorbing sliver 44 and gradual insertion 46 into the waveguide cavity 45, by a second remotely controlled stepper motor 41. This is done either by directly attaching the sliver disc 44 to the motor axis 42 or, if needed, using some reduction gear. The motor 41 is also fixed at a position of the waveguide between the test port 91 and the tuning probe 95.
(15) The operation of the load pull tuner is visualized in FIG. 9A: the RF energy is generated by the DUT and is injected into the waveguide 904 at the test port 91. Some 97 of this energy is reflected at the absorbing sliver 92, as seen in FIG. 6B, and returned back to the DUT; this is a spurious effect, that must, though, be accounted for. Most of the arriving energy is absorbed by the sliver 92, as is obvious when calculating the inverse available loss |S21|.sup.2/(1?|S11|.sup.2) from FIGS. 6A and 6B, and the remaining part 98 traverses towards the reflective probe 95, which is held dielectrically 94 by a vertical control mechanism (axis) and is reflected back 901; the absorption by the sliver 92 and the reflection at the probe 95 determine the overall reflection factor 90 at the test port 91. Because the RF energy bounces back and forth between sliver 92 and probe 95, an exact calculation is complex. Therefore, the tuner is calibrated at every frequency for a combined multitude of horizontal tuning probe positions and sliver penetrations and the data are saved in a calibration file, retrieved and interpolated before use.
(16) The tuner calibration flow chart is shown in FIG. 8: in a first step, after the tuner is connected to a pre-calibrated vector network analyzer (FIG. 7, see ref. 11), the absorbing sliver 92 is fully withdrawn from the waveguide allowing full energy transfer and the tuning probe 95 is placed at an initial position, that can be defined using a limit switch; this is called the INIT position and defines ???.sub.0?0 and X=0; after that the scattering (s?) parameters of the tuner are measured and saved in an init matrix [S0]; after that the calibration enters two move-measure-save loops. In the first loop the tuning probe 95 stays at X=0 and the sliver 92 is rotated gradually into the waveguide slot 93; The registering can be in degrees and mm, but it is more convenient to stick with motor steps: depending on motor step resolution and possible reduction gear a 0 to 90 or 0 to 180 degree rotation corresponds to at least 50 or 100 motor steps or 90/50 or 1.8 degree per motor step, which translates to a relative penetration of h=H*sin ?=sin(?/4*Y) and dH/H=cos(?/4*Y) and X=0 to XMAX=(?/2)/resolution, where H is the internal height of the waveguide and resolution is the displacement per motor step, all expressed in motor steps X or Y. Once the sliver rotation, s-parameter measurement and saving in a file named PHI is finished, the sliver is rotated back to the initial state ?.sub.0 and a second move-measure-save cycle is started with the probe moving for X between 0 and 22; all the measured s-parameter data as a function of position X are cascaded (see ref. 10) with the invers matrix [S0].sup.?1 and saved in a file named X. Subsequently all s-parameter permutations of the data saved in files PHI and X are created in computer memory and saved in a tuner calibration file CAL(F) at the calibration frequency F; this file contains s-parameters of the tuner for a multitude of stepper motor combinations creating reflection factors covering the Smith chart. Using these anchor points proper interpolation functions allow high resolution tuning at combinations of horizontal probe positions and sliver penetrations between calibrated points. As an example: 20 penetration levels and 100 horizontal positions lead to 2000 calibration points; instead, the fine penetration resolution can be set to up to 200 levels and the horizontal movement up to 2000 steps, a total addressable space of 400,000 distinct reflection factors. The proposed calibration, instead requires a total of 100+20=120 points to be measured, lasting ca 2-4 minutes instead of a full calibration lasting 110 to 220 hours.
(17) The tuning mechanism is best explained using the diagram of FIG. 10: with the absorbing sliver withdrawn and the tuning probe at any arbitrary position X, the assumed reflection factor would be at point 100. The target being point 102, the control will follow either path 101-102 (i.e., increase attenuation and move the tuning probe) or path 103-102, i.e., move the probe first and then increase the attenuation. In both cases there is a single possible attenuation and horizontal robe setting, the choice is only the path to follow. Under circumstances, when the DUT is prone to spurious self-oscillations, the attenuation is set to maximum i.e. the starting point is always at 104 and the recommended move would follow a path moving X around the center 104 (which is invisible, since the vector has zero amplitude) and then reduce the attenuation to get to 102; this also reduces the risk of transient spurious oscillations during the movement of the tuning probe (see ref. 7).
(18) Obvious alternatives to the disclosed concept of a load pull system using a tuner with one horizontally only moving probe and a pre-positioned adjustable attenuator and the associated calibration method shall not impede on to the validity of the present invention.
