Support
home news pca support products contact us about Dielectric employment

 
Technical Papers
Contact Info
Who to call when you need technical assistance.
 
Waveguide or Coax for 1 MW DTV

By: J. Stenberg

Waveguide transmission lines have been widely used in high power UHF NTSC installations for several decades. The high transmitter power necessary for 5 Megawatts of ERP often required the use of waveguide. Their high efficiency provided an economical alternative to coax for long transmission line runs. Deciding to use waveguide for NTSC was primarily based on a trade-off between increased power handling, higher efficiency and increased windload. Other performance issues were considered essentially equivalent.

DTV implementation raises the question of whether these tradeoff assumptions remain valid. Lower DTV ERP requires lower transmitter average power than NTSC. The disadvantage of higher windload often outweighs the efficiency advantages especially on existing installations. New sizes of coax transmission line with increased power handling can provide more alternatives. Variation of group delay across the 6 MHz band due to waveguide dispersion means that 8-VSB pre-correction may be required at the transmitter to reduce EVM. This paper will discuss the tradeoffs to consider when choosing between waveguide and coax for DTV. It will provide theoretical and measured data on both.

This paper will talk about the criteria that determine which type of line to use for DTV and the tradeoffs to consider.

First, we will look at efficiency, what effects it and how coax line moding and average power rating are related.

Second, we will look at windload and compare waveguide and coax.

Lastly we will discuss DTV performance.

Efficiency, commonly measured in dB/100 ft, is the value all broadcasters are interested in since it defines the electricity costs. Obviously higher efficiency is a desirable goal.

Efficiency is determined by:

  • Whether the line is waveguide or coax
  • The cross sectional dimensions, the larger the cross section, the higher the efficiency.
  • The shape, circular cross sections provide the highest efficiency since the surface area is minimized.
  • The materials, high conductivity copper or aluminum and a minimum amount of dielectric.

Click Graphic Above for Detailed View

This chart compares various types and sizes of line. The values are based on 1000 ft of line at channel 17.

As you can see waveguide can provide a higher efficiency than coax. In this case a 12% increase is possible by using WR1800 instead of 7-3/16" coax.

To maximize efficiency in a coaxial system the largest diameter possible should be used.

This is however limited by the frequency at which a waveguide TE11 mode can propagate in the coax. Coax line should not be operated above this frequency since its field orientations are not the same as the normal TEM coax mode. Energy excited in this mode may become trapped, cause overheating and high VSWR.

Higher order modes must be excited by a discontinuity in order to propagate. This can include diameter or direction changes and the use of dielectric insulators.

Higher impedance lines (since a+b is smaller) have higher cut-off frequencies than low impedance lines.

The cutoff frequency is determined by this formula where a and b are the radii of the inner and outer conductors.

Click Graphic Above for Detailed View

This chart shows the theoretical and practical values for the cutoff frequency of standard transmission lines.

The practical values are the highest frequencies recommended for normal use.

  • 8-3/16" is usable to channel 52
  • Our new 7-3/16’ is usable to channel 69
  • 6-1/8" lines are usable to well above channel 69

Note that modes may still be a problem at frequencies below these practical values if the geometry is changed or sufficient distance between components is not maintained.

Since the peak power capacities of coax are considerably higher than those required for DTV it is not considered a limiting factor. Average power on the other hand is a limiting factor and must be considered.

Average power ratings are defined by setting a maximum allowable temperature on the inner conductor. This temperature corresponds to the temperature above which oxidation of the copper conductor occurs rapidly.

There are several methods available to increase the power ratings of coax line:

  • Increase the efficiency of the line by using higher conductivity materials or larger diameters.
  • Increase the pressure inside the line thereby providing better thermal conductivity.
  • Design a method to remove more heat from the inner and transfer it to the outer and then to the ambient air.

Click Graphic Above for Detailed View

Dielectric’s new EHT/Line utilizes a unique enhanced heat transfer process which provides a significant decrease in the inner conductor temperature for a given power level.

As shown a standard 6-1/8" 75 ohm line at channel 38 is rated at 61 kW. Our new EHT/Line is rated at 85 kW for the same inner temperature.

Bullet wear is not effected by this increase in power since the inner remains at or below the same temperature.

EHT/Line is available in all sizes and configurations.

Click Graphic Above for Detailed View

Windload is the greatest disadvantage of waveguide and is most often the controlling factor. Existing tower installations more than likely will not have the loading capacity necessary for waveguide.

As this chart shows the loading is as much as 5.8 times greater than 7-3/16" coax.

DTW or circular waveguide windloads, while lower than rectangular, are still 1.8 to 2.4 times greater than 7-3/16" line.

Degradations to the 8VSB signal from waveguide and coax are not the same.

The change in phase relative to a change in the input frequency is always linear for a coax line. It does not have a lower cut-off frequency and is not dispersive.

Waveguide on the other hand has a phase versus frequency behavior which is non-linear and introduces group delay to the signal.

The amount of group delay depends on three critical factors:

  • First, the broad dimension of the guide which sets the lower cut-off frequency.
  • Second, the overall length of the run. The longer the length, the higher the total group delay variation.
  • Third, the channel of operation. The closer to the cutoff frequency the guide is operated, the greater the group delay.

Click Graphic Above for Detailed View

This chart illustrates how these three factors affect the total group delay variation. It is based on DTW1500. Other sizes would have similar values but on different channels.

When the guide is operated near cutoff, channel 30 in this case, the total group delay variation is large. For 1500 ft 29 nS of delay results at the output of the line over the 6 MHz channel.

Channel 38 is somewhat better at 18 nS for 1500 ft.

When the guide is operated well above the cutoff frequency the effect is minimized. At channel 54 approximately 9 nS results for 1500 ft.

Click Graphic Above for Detailed View

To quantify the degradation in 8VSB signal to noise ratio, tests were performed at Comark utilizing an HP 89441 vector signal analyzer and several test sections of waveguide. The frequencies were adjusted close to the cutoff so that large group delay variations could be produced in a short run.

This chart shows the results of these tests. Note that above approximately 22 nS the SNR drops below 27 dB. For virtually any length there will be some reduction in SNR.

For the reasons on the prior chart, correction must be made at the transmitter for any length of waveguide if maximum DTV performance is desired.

This can be done utilizing fixed compensation either at RF or digitally. It is not clear at this point what the performance cost of this correction is although a higher peak to average ratio is likely to result.

Dynamic adaptive pre-correction which requires coupling and feedback from the top of the run would provide the highest level of compensation. This, however, does not appear practical at this point.

The choice thus comes down to:

Is the increased efficiency of waveguide worth the additional windload and non-optimum DTV performance?



Dielectric Communications
22 Tower Road
P.O. Box 949
Raymond, ME 04071
(207) 655-8100
(800) 341-9678
Fax: (207) 655-8177
dcsales@dielectric.spx.com


Home | Broadcast | Products | Support | PCA | News | Employment | About Dielectric | Contact Us
All pages and images © 2005 Dielectric Communications, A unit of SPX Corporation
Created by Image Works