METHOD OF MANUFACTURING A TRANSVERSE ELECTRIC MAGNETIC (TEM) MODE TRANSMISSION LINE AND SUCH TRANSMISSION LINE

Abstract
Method of manufacturing a transmission line including the steps: forming an element with at least one longitudinal groove on a surface of said element, said longitudinal groove being defined by two opposite wall portions in the element and having a longitudinal opening adjacent to said surface, andlocating a conductor line in said at least one longitudinal groove. The method is distinguished by the steps:forming the conductor line from a metal strip upon punching the same from a sheet of metal,attaching the metal strip to at least one holding device, andmounting said at least one holding device, with the attached metal strip, on said element, so as to locate said metal strip in said longitudinal groove at a distance from said opposite wall portions of said element. The invention also concerns a transmission line manufactured in accordance with said method.
Description
FIELD OF THE INVENTION

The present invention concerns a method for the manufacture of a transmission line according to the preamble of claim 1. Further, it concerns such a transmission line, according to the preamble of claim 2.


PRIOR ART

The use of microstrip lines with air as the dielectric has been used in base station antennas since at least the mid-90s. The advantage is low loss and ease of manufacture in the sense that a flat ground plane can be used for the lines. The disadvantage of such lines is the necessary width, which for e.g. a 3 mm separation is 13 mm for a 50 ohms line. The absence of a dielectric, other than air, also means that the mutual coupling is rather high.


As another prior art in respect of the present invention U.S. Pat. No. 5,652,557 could be mentioned. This patent discloses a transmission line consisting of a longitudinal groove having longitudinal metal walls and a longitudinal opening. In the groove, a microstrip conductor line is formed.


However, the microstrip conductor of U.S. Pat. No. 5,652,557 is manufactured by disposing a conductor film on a dielectric. Such a production method can be complex and expensive.


THE OBJECT OF THE INVENTION AND ITS MOST IMPORTANT CHARACTERISTICS

It is an object of the present invention to propose a solution for or a reduction of the problems of prior art. A main object is consequently to devise a method for manufacturing a transverse electric magnetic (TEM) mode transmission line, that provides a transmission line having beneficial electric properties and at the same time is a method that is both rational and easy to employ.


According to the invention this is accomplished by a method having the features of claim 1. According to this method, a transmission line can be produced by forming a metal strip through punching a metal sheet and then fastening the metal strip in a longitudinal groove by means of a holding device or spacer. This is a rational method: it is easy to punch a metal sheet to form the metal strip, and the provision of a holding device enables an easy mounting of the metal strip in production.


According to another aspect of the invention, one or more of the above objectives is reached with a transmission line having the features of claim 2. This is a straightforward solution for a transmission line that lends itself to a rational production of the same.


We present an alternative transmission line geometry well suited to a base station antenna structure. As shown in FIG. 1a, the microstrip lines are rotated 90 degrees and placed in a longitudinal groove, or “canyon”, on a surface of an element. Said longitudinal groove being defined by two opposite wall portions in the element and having a longitudinal opening adjacent to said surface. At least a surface layer of said opposite wall are formed by an electrically conductive material, such as a metal. This makes the line resemble a coaxial line which decreases losses and the mutual coupling, as will be demonstrated in the following sections. The proposed line is well suited for base station antennas since these have long transmission lines extending primarily in the array direction. Furthermore, the grooves are readily accommodated if the antenna reflector is made from an aluminum extrusion. The proposed line is similar to the trough or channel line of [1-2] (references, in brackets, are listed at the end of the description), but differs in that the centre conductor is aligned with the top edge of the channel. The proposed “canyon stripline” also has an advantage in the use of a metal strip as opposed to a wire, since a strip allows us good control of the characteristic impedance.


The characteristic impedance of the line is controlled by changing the height of the line. FIG. 1a also shows an inter-connect between the two lines. In the manufacture, the complete line, or metal strip, could be punched out of a single piece of sheet metal with little loss of material and then bent 90° along two lines. The U-shaped line pair could then be attached to at least one holding device or spacer, inserted into the grooves of the element and fastened at a distance from the wall portions forming the grooves using said at least one holding device or spacer. The at least one holding device is mounted on said element, e.g. in the vicinity of the opening of a longitudinal groove. The holding device 101c could for instance be made of a plastic material. One design of such a transmission line is shown in FIGS. 1a, 1b and 1c, but there are numerous alternatives how this could be done. For instance, a single line, or metal strip, could be punched out and used in a single groove. Moreover, the holding device or spacer could be made of any material with sufficient isolation properties.


We compare the proposed transmission line with a standard air-microstrip line of similar dimensions. Our simulations, using a software, HFSS, for 3D electromagnetic-field simulation, confirm that the proposed line has lower loss and mutual coupling than air microstrip.





SHORT DESCRIPTION OF THE DRAWINGS

Embodiments exemplifying the invention will now be described, by means of the appended drawings, on which



FIG. 1
a illustrates two parallel transmission lines, according to the invention placed in two longitudinal grooves or “canyons”,



FIG. 1
b illustrates the two parallel transmission lines in an embodiment such that the complete assembly could be punched out of a single piece of sheet metal,



FIG. 1
c illustrates a possible solution to attach the line to the reflector,



FIG. 2
a illustrates the geometry of a transmission line according to the invention,



FIG. 2
b illustrates the field distribution in a transmission line according to FIG. 2a,



FIG. 3 illustrates the relationship between impedance and strip height for a transmission line according to the invention,



FIG. 4
a illustrates a cross-section of a line with the simulated field distribution,



FIG. 4
b illustrates the simulated loss of the line in FIG. 4a,



FIG. 4
c illustrates a cross-section of a line with the simulated field distribution,



FIG. 4
d illustrates the simulated loss of the line in FIG. 4c,



FIG. 5
a illustrates a cross-section of two lines with simulated field distribution, in order to consider mutual coupling,



FIG. 5
b illustrates the simulated loss of the lines in FIG. 5a,



FIG. 5
c illustrates a cross-section of two lines with simulated field distribution, in order to consider mutual coupling,



FIG. 5
d illustrates the simulated loss of the lines in FIG. 5c,



FIG. 6
a illustrates a simulation of two parallel transmission lines with a cross-over in between,



FIG. 6
b illustrates a Smith chart in connection to the simulation of 6a, and



FIG. 6
c illustrates the simulated loss of the lines in FIG. 6a.





DESCRIPTION OF EXEMPLARY EMBODIMENTS


FIG. 1
a illustrates two parallel lines of the proposed “canyon line” design. The two lines (dark) are connected by a cross-over part at the centre and the complete assembly could be punched out of a single piece of sheet metal as shown in FIG. 1b. Here, 101b denotes an output end of 50 Ohms, 102b denotes a 70 Ohms transformer, 103b denotes a 50 Ohms output end, 104b denotes a 50 Ohms input end and, finally, 105b denotes folded cross-over sections. FIG. 1c shows a possible solution for a plastic spacer 101c to attach the line 102c to the reflector 103c. The reflector could for instance by of extruded aluminium.


The first step in the design is to determine the dimensions of our proposed line for 50Ω operation. FIGS. 2a and 2b illustrates the geometry and field distribution respectively in the proposed TEM transmission line. A vertical strip is placed in a groove in e.g. an aluminium extrusion. As seen in FIG. 2b, the field distribution is quite similar to a stripline. The characteristic impedance is thus roughly inversely proportional to the strip height h over a rather large region, since the fundamental effect is that we increase the capacitance per unit length. The calculated impedance using HFSS is shown in FIG. 3. FIG. 3 illustrates line impedance vs. strip height h for the proposed “canyon line”. Other dimensions: width w=10 mm, depth d=15 mm, thickness t=2 mm, compare FIG. 2a. In FIG. 2a height h is referenced by 201a, width w by 204a, depth d by 202a and thickness t by 203a. For narrow strips the behaviour deviates since the relative amount of fringing fields will be larger, just as for a regular microstrip and stripline. We also note that the linear behavior is broken as the height h approaches the groove depth d, which in this case is 15 mm. Thus, for a robust design, we should choose the groove width w and depth d so that we can change Z0 over a desired range without having to let the height h be close to the depth d. Using HFSS, we have compared two 300 mm long transmission lines, one being a traditional air microstrip and the second our proposed line. The cross-sections of the lines with the simulated field distribution are shown in FIGS. 4a-4d along with the simulated loss. FIGS. 4a-4d illustrate an HFSS simulation of a single transmission line. The line length is 300 mm or 1λ at 1 GHz. In FIGS. 4a and 4b: a microstrip line of width 13 mm, thickness=2 mm, distance to ground=3 mm. In FIGS. 4c and 4d: the proposed “canyon line” with depth d=15 mm, width w=10 mm, height h=10.2 mm, thickness t=2 mm. We note that the losses are smaller in the proposed line, although the material is the same and the dimensions quite similar. The reason is the more efficient use of the conducting strip area when the field is distributed on both side of the strip.


The resulting E-fields, in Volts/meter [V/m] of the simulations can be summarised for FIG. 4a as:















Reference numeral
















401a
402a
403a
404a
405a
406a
407a
408a



















E-field
125
375
625
875
1125
1375
1625
1875


[V/m]









The E-fields for FIG. 4c as:















Reference numeral

















402c
403c
404c
405c
406c
407c
408c



401c
415c
414c
413c
412c
411c
410c
409c



















E-field
125
375
625
875
1125
1375
1625
1875


[V/m]









Note: The loss in the air microstrip, 0.22 dB/lambda at 1 GHz, seems much too high. It is possible that HFSS overestimates the loss due to radiation, but simulations using a larger air volume and smaller port areas could not confirm this. We are still confident however, that the proposed line has lower losses.


HFSS Simulations of Two Parallel Lines

In order to compare the proposed transmission line to air microstrip, we should also consider the mutual coupling. For this purpose, we have made simulations of the two comparable geometries shown in FIG. 5a and FIG. 5c. FIGS. 5a-5d illustrates an HFSS simulation of two parallel lines. Length=300 mm. FIG. 5a: microstrip lines as in FIG. 4a with an edge-to-edge separation of 9 mm. Total width=35 mm. FIG. 5c: proposed “canyon lines” with a separation of 5 mm. Total width=25 mm. Both pair of lines are 300 mm long and the case of FIG. 5c (the proposed “canyon line”) is in fact the same as in FIG. 1a, except for the cross-over section omitted in FIG. 5c. It turns out that the proposed line may be used with as little as 15 mm center-to-center separation, but this is not possible for the microstrip due to the width of the strip (13 mm). Instead, we chose to increase the separation so that the microstrip pair of lines occupy a width of 35 mm compared to the 25 mm for the proposed “canyon stripline” pairs. Since the “canyon line” uses space below the ground plane, it is perhaps still a fair comparison. Note that a comparison between two transmission line geometries is always arbitrary to some degree. It is therefore possible that another choice of e.g. height-to-width ratio would give a slightly different result for the microstrip line. However, the microstrip design used here should be rather beneficial since the distance to ground of 3 mm reduces mutual coupling. The simulation results in FIGS. 5a-5d clearly show that the fields are more confined in the proposed line. The field strength at the second line due to a 1 W signal input on the first line is 10 times weaker compared to the microstrip line. This corresponds to a 20 dB lower coupling as seen in the S21 parameters. Note the periodic behavior in the coupling caused by the finite line length of 300 mm, or 1λ at 1 GHz.


The resulting E-fields, in Volts/meter [V/m] of the simulations can be summarised for FIG. 5a as:















Reference numeral
















501a
502a
503a
504a
505a
506a
507a
508a



















E-field
45.255
72.882
117.38
189.03
304.44
490.29
789.61
1271.7


[V/m]









The E-fields for FIG. 5c as:















Reference numeral





















505c
506c
507c
508c
509c



501c
502c
503c
504c
514c
513c
512c
511c
510c




















E-field
28.100
45.255
72.882
117.38
189.03
304.44
490.29
789.61
1271.7


[V/m]









Finally, we have simulated a design where a cross-over section is placed between two parallel lines of the proposed “canyon stripline”. FIGS. 6a-6c illustrates an HFSS simulation of two parallel “canyon lines” with a cross-over between them. The Smith chart in FIG. 6b shows S11 with the reference plane at the cross-over point. The cross-over section is 10 mm wide and is placed 2 mm above the reflector. The total line length is 300 mm and the insertion loss is 0.05 dB at 1 GHz. As seen in FIG. 6a, two 50 ohms lines are connected at a point halfway along the ground plane. The total transmission line length is thus as before 300 mm. It turns out that it is quite easy to achieve a low reflection at low frequencies, but at higher frequencies the inductance becomes a problem. At 1 GHz, see FIG. 6c, the reflection is quite acceptable, −40 dB. The simulated loss at 1 GHz has increased from 0.03 dB for the straight line (FIG. 2a) to 0.05 dB, which indicates that radiation is a small concern at this frequency.


The resulting E-fields, in Volts/meter [V/m] of the simulation can be summarised for FIG. 6a as:















Reference numeral

















601a
602a
603a
606a
606a
606a
607a
608a
609a




















E-field
10.834
17.448
28.100
45.255
72.882
117.38
189.03
490.29
1271.7


[V/m]









CONCLUSIONS

We have presented numerical HFSS simulations of a “canyon stripline” and compared it to a traditional air microstrip. The design is well-suited for integration in base station antennas using extruded reflectors. The simulation results show that the proposed line has lower losses and lower mutual coupling.


There are many variations of the invention. For instance, the wall portions forming the groove(s) in the element can be made of a thin sheet of conducting material or paint placed on a non-conducting material. The holding device could be snap-fitted to the metal strip by the use of e.g. arms fitted into one or more holes of the metal strip. The holding device could be attached to the element by the use of e.g. rivets that could be an integral part of the holding device. This would enable the metal strip with the spacer(s) to be mounted in a single step.


REFERENCES



  • 1. Brian C. Wadell, “Transmission Line Design Handbook”, Artech House, 1991.

  • 2. H. A. Wheeler, “Transmission-Line Properties of a Round Wire in a Polygon Shield”. IEEE Transactions on Microwave Theory and Techniques, Vol. MTTT-27, No. 8, pp. 717-721, August 1979.


Claims
  • 1. Method of manufacturing a transverse electric magnetic (TEM) mode transmission line comprising: forming an element with at least one longitudinal groove on a surface of said element, said longitudinal groove being defined by two opposite wall portions in the element and having a longitudinal opening adjacent to said surface, at least a surface layer of said opposite wall portions being formed by an electrically conductive material, andlocating a conductor line in said at least one longitudinal groove, the method further comprising:forming the conductor line from a metal strip upon punching the same from a sheet of metal,attaching the metal strip to at least one holding device, andmounting said at least one holding device, with the attached metal strip, on said element, so as to locate said metal strip in said longitudinal groove at a distance from said opposite wall portions of said element.
  • 2. A transverse electric magnetic (TEM) mode transmission line, comprising: an element formed with at least one longitudinal groove on a surface of said element, said longitudinal groove being defined by two opposite wall portions in the element and having a longitudinal opening adjacent to said surface, at least a surface layer of said opposite wall portions being formed by an electrically conductive material, anda conductor line located in said at least one longitudinal groove,
Priority Claims (1)
Number Date Country Kind
SE0601971-5 Sep 2006 SE national
PCT Information
Filing Document Filing Date Country Kind 371c Date
PCT/SE07/00834 9/24/2007 WO 00 7/6/2009