Using antennas to improve PCS in-vehicle performance
Degradation of PCS phone signal levels, when used inside a vehicle, is less than that found at cellular frequencies. Test range results confirm, however, that PCS portable use is enhanced by an externally mounted antenna.
The new PCS band brings with it the use of portable phones at microwave frequencies. The operation of these phones inside an automobile deteriorates, and performance suffers because of losses attributable to the shielding effects of the metal car body. The following discussion quantifies the performance of a hand-held PCS portable phone operating in an automobile and offers suggestions on how performance can be improved.
One objective that contributes to optimum performance is an omnidirectional signal radiated from the portable. The radiation patterns represent the relative field strength in the horizontal plane. The zero position in the figures is oriented to the front, or forward direction, of the test vehicle, a sedan. The average signal level was determined by averaging the measured signals in one-degree increments. The important point to recognize is the amount of pattern distortion or nulls caused by signal reflections inside the test vehicle. After attempting to measure the actual radiated signal from a portable phone held by the driver, we determined that the variations attributable to the position of the portable in relation to the driver were not repeatable. Hence, the driver and portable phone were replaced with a standard halfwave, vertically polarized, dipole antenna mounted off the headrest in a position simulating an actual portable antenna. This resulted in credible measurements in lieu of a person holding a portable, which might result in questionable portable phone efficiencies.
The method of measurement began with the construction of a halfwave reference antenna tuned to 1,930MHz. The center-fed dipole was fed through a coaxial quarterwave balun, which was part of a 50V air line, terminated in an N connector. The halfwave dipole reference was mounted on a five-foot fiber glass support pole and positioned on the center of a 24-foot-diameter ground level car rotator, as shown in Photo 1 on page 20. Probing the test range with the dipole showed a field-strength variation of less than 0.75dB over the vertical aperture between the bottom of the window opening and approximately 12 inches above the roof line of the vehicle. The test range used is a 200-foot, outdoor, ground-level range. A block diagram of the instrumentation for recording test results is shown in Figure 1 on page 22. A frequency-synthesized, phase-locked signal was used to obtain the data and accurately generate the patterns.
The reference signal level was measured for the free-space, halfwave dipole located at the test vehicle roof line. The reference dipole was replaced by the test vehicle, which was then rotated with the reference dipole in the same position that the portable antenna would be located if held by the driver. Figures 2, 3 and 4 show the pattern distortion attributable to operating inside a vehicle. Figure 2 on page 22 shows the pattern of the halfwave dipole inside the vehicle overlaid on the same dipole in free space with the vehicle removed. The nulls, as much as 20dB deep, are attributable to signal scattering caused by the vehicle. To the extent that an average signal has meaning, the signal level of the dipole inside the vehicle is about 3dB less than the free-space, halfwave dipole.
Figure 3 on page 22 shows the effect of a driver holding the dipole as though he were talking on a hand-held PCS phone. The nulls are much deeper, and in some directions, there is an additional signal loss of 5dB-10dB. The average signal level is 5dB less than the dipole without a person holding it. This is probably the best-case loss, because the efficiency of the portable phone and the actual positioning by a portable phone user will only deteriorate the signal further.
The free-space dipole is an excellent reference. However, in the real world, quarterwave and collinear antennas are used on vehicles. Therefore, it is necessary to tie the performance of the portable phone to a standard, externally mounted antenna. Figure 4 on page 24 compares a dipole inside the vehicle with a quarterwave whip on the roof of the vehicle. Notice that the roof-mounted quarterwave has an omnidirectional pattern and nicely fills in the nulls that result from using the portable phone inside the vehicle.
Figure 5 on page 24 shows that a quarterwave monopole has a vertical half-power beam-width of about 228 and a positive major lobe beamtilt of 208 above the horizon. The major lobe gain of the monopole on an 8l ground plane is 13dB relative to the halfwave dipole in free space. This same pattern shows that the performance of the quarterwave can be improved by lowering the major lobe through the use of a two-element collinear antenna design. Figure 5 also shows that the gain of the two-element collinear is 13dB greater than the quarterwave and is essentially equal to a halfwave dipole in free space when measured at the horizon.
Two-ground plane tests To further understand these signal levels, we undertook a series of vertical plane patterns. These patterns were made using two large ground planes which represented the car roof and the metal at the lower window level. Two metal disks, 48 inches in diameter (representing an 8l ground plane) were fabricated.
The two-ground plane test was conducted to determine the shape of the vertical (elevation) plane pattern. The test consisted of mounting the two disks 16 inches apart, then mounting the halfwave dipole 7 inches below what would represent the roof of the car and then rotating the system to produce a vertical plane radiation pattern. The 16-inch spacing was chosen to represent the average height of the window opening on a car. This is a 2.75l aperture at 1.9GHz, considerably larger in terms of wavelengths than at 450MHz or 900MHz. The four metallic supports represent the support pillars on the car. Figure 6 on page 26 shows a vertical plane pattern with the quarterwave on a single 48-inch ground plane, vs. the halfwave dipole between the two disks, vs. the halfwave dipole in free space as a reference. The halfwave dipole between the two disks shows a multilobe pattern, with one of the lobes near the horizon and within a couple of decibels of the halfwave dipole in free space. The data gathered in this plane can be used to explain why the signal levels measured in the horizontal plane on the car, at the horizon, show a 25dB average signal strength when compared to a halfwave dipole.
The use of a two-element collinear antenna design on the same 8l ground plane reduces the beamtilt from 1208 to 1118 above the horizon, as shown in Figure 6. The beamwidth and gain is essentially the same as the quarterwave in the major lobe.
Conclusions The average signal level from the standard antenna operating inside the vehicle, without any passengers, is about 5dB below the same antenna in free space. Further, the pattern distortion caused by a passenger operating the phone inside the vehicle increases the loss. The best solution to regain some of this signal loss is to use an external, vehicle-mounted antenna. The use of an external, collinear antenna, such as an Antenna Specialists’ ASPM1954T, would not only increase signal level but would also eliminate the effect of the nulls in addition to the benefit of an omnidirectional pattern.
The use of an externally mounted antenna will improve the performance of the PCS portable in two respects: It will provide gain over the antenna on the portable, and it will further enhance signal coverage.
The good news is that the vehicle’s degradation of portable phone performance at PCS frequencies is less than what has been demonstrated at cellular frequencies. However, the efficiency of the portable, together with the relatively small size of the antenna, results in in-vehicle system performance that is even more user-dependent than cellular. The best solution is an external antenna, or an active mobile repeater placed inside the car and hardwired to an external antenna.