SUPPLYING DIGITAL SATELLITE SIGNALS TO
MULTIPLE OUTLETS
Technical Memo by Maurits Roos, Teltrac Communications Ltd
A 1939 written RCA manual intended to assist TV receiver installers recommends a minimum signal level of 1 MV (one millivolt, the same as 1,000 microvolts - today on a signal level meter we would say the signal is +60 dBuV) to the TV receiver antenna terminals. They had it just about right, even 61 years ago.
Inside of the TV set there are circuits and (today) transistor of IC devices that amplify the relatively weak off-air terrestrial TV signal. Just as our satellite receiving systems have signal amplifiers (the LNB is the first one - typically 60 dB of gain there, followed by more gain circuits inside of the satellite receiver - typically an additional 60 dB or more), the TV set has to boost the signal before it can be converted into picture and sound. An amplifier stage has two relevant characteristics of interest here:
It is the latter which is important to us initially. Any amplifier device has a noise generator built into the system. Creating noise, while at the same time amplifying signal, is unavoidable. Our LNBs are a perfect example of this. A Ku band LNB with a "noise factor" of 0.6 dB is better (more sensitive) than a second LNB with a noise figure (1.0 dB), that amplifier created noise follows the signal through the system to the satellite receiver.
RCA's 1939 technology began with the "noise floor" of the receiver as a limiting factor in reception. In other words, how much noise did the receiver's amplifier stages generate while processing the TV signal? Reverting back to the dB scale - if the television receiver is recommended to have 60 dB (uV) of signal for a "noise free" picture, what was RCA saying about the receiver proper? First, that it had to have some significant difference between the noise the receiver created and the signal that it amplified. This "ratio" of the desired signal to the noise of the system is abbreviated by engineers as the S/NR or signal to noise ratio. The pivotal word here is ratio.
If 60 dB is an impairment free picture, at what point does the noise itself start? The answer is not 0 dB (uV) because 0 dBuV would be no noise at all.
Every amplifier creates noise - the question is only "how much noise?". Low noise amplifiers (as in LNB - "low noise block" [downconverter] have evolved over decades of tiny improvements.
In the 1939 TRT12 television set from RCA, the gain or amplifier stages inside the TV set created a "noise floor" of around 20 dBuV. Disconnect the TV aerial from the TV set, hook up test equipment and measure the amount of noise present in the system with no signal. The answer was 20 dBuV or around 10 microvolts. Thus, if the TV set creates 20 dBuV or noise, and it takes a ratio of 40 dB S/NR to produce a noise free picture, we need 40 dB more signal than we have noise.
The essence of the 1939 RCA text has not changed. Terrestrial TV reception still requires 40 dB more signal than noise, and 40 dB more desired signal than "interference" to be blemish free. The only change from 1939 is the actual (measured) "noise floor" of modern TV receivers. Today, the noise floor of a quality TV set is in the region of 15 dBuV (6 microvolts of noise). This means that if 15 dBuV is the receiver "noise floor" and you need a 40 dB signal to noise ration (S/NR) then the minimum signal required is 15 + 40 = 55 dBuV. You can prove these numbers to yourself by taking a signal level metre and TV set connected to a terrestrial TV antenna and attenuating the signal to the metre while simultaneously watching the picture quality on the TV set. With no external (antenna - masthead) amplifier, as the measured signal hovers around 55 dBuV you will begin to notice "graininess" (noise) on the screen. Reduce the signal level lower in level and the "grain" turns into on-screen snow.
What you have just read contains one immutable law of physics. It is not how much signal you have that determines the quality of the TV picture - it is the ratio between the signal and all forms of interference. For this discussion, we have to include "receiver generated noise" as a form of interference.
There is one more thing to consider. That every amplifier creates noise; every amplifier. But some create more noise than others. Which leads us to another immutable law of physics:
The noise factor of a receiving system is most affected by the noise created within the first amplifier stage of the system. In other words - if you have a very low noise amplifier as the first amplifier, the amount of noise in the entire system is reduced by the noise factor of that first amplifier.
What you have just read pertains to an analogue signal. Along came the world of digital.
The laws governing the signals do not change for digital but the signal level and the signal to noise ratio (S/NR) requirements are vastly different.
The satellite integrated receiver decoder (IRD) typically only requires 8 dB signal to noise ratio (S/NR) rather than the 40 dBuV a analogue terrestrial receiver requires, to produce the perfect picture with no imperfections, providing there are no impedance mismatches causing additional errors in the data stream.
This means is that if 20 dBuV is the receiver "noise floor" and you need 8 dBuV signal to noise ratio then the minimum signal required is 20 + 8 = 28 dBuV considerably lower than a analogue terrestrial signal required to produce a noise free picture.
It would not be advisable to run with a signal level at this threshold point because it would not allow for any signal loss due to rain fade, which is a common problem of Ku band systems in bad weather.
A typical signal from a 60cm dish of Optus B3 with a signal to noise ratio (S/NR) of 17 dB (9dBuV above Threshold). The signal level can be attenuated to a level of 37 dB before there is any noticeable effect on the Bit Error Rate (BER), as long as the cable and connectors are of good quality, such as Belden, Comscope or Times Fiber.
This leads to another immutable law of physics. You provide a better quality signal to the IRD with fewer errors, with a lower signal level with no amplification, rather than a higher signal level with a cheap line amplifier, providing less signal to noise ratio (SN/R) and more errors.
The input to a digital IRD is commonly specified as 75 ohms (impedance) across some stated bandwidth (such as 950 2,050 MHz with an input signal level ranging between two limits ( 44 to 79 dBuV [Pase DVR-500] as they specify)
The "F" connector on the rear of the digital IRD says it is for the IF input. And it specifies the input frequency range (for example, 950-1450). The integrity of this connection is critical to the performance of many digital IRDs.
The specified input impedance at the F connector is 75 ohms. You might expect that if you are using 75 ohm cable, and a 75 ohm F connector to plug into this jack, you will indeed have a 75 ohm connection. Guess again.
Experience to date suggests that any mismatch (variation) from the 75 ohm impedance of the interconnecting cable from the LNB can cause high bit error rates (BER) to appear. You can create similar "mismatch" conditions with improper use of signal splitters or unterminated splitter ports and/or improper use of line amplifiers.
Bit error Rate (BER) is measured in exponent notation; the Bit Error Rate (BER) quantifies the performance level of the digital link. A BER of 1x10-3 expresses the probability of one bit error occurring in a block of 1,000 bits. A BER of 5.0x10-5 is superior to a lower BER or 9.0x10-4 because there is a probability that less errors will occur. BER also may be expressed as 5E-4 or 3E-3, which is the equivalent to a BER of 5 x 10-4 or 3 x 10-3
Proper digital operation involves accuracy of timing by the transmission path components. If some part of the path caused a portion of the incoming signal to be delayed in time at the receiver, the information that is delayed finally arrives at the receiver simultaneous to later transmitted information. Now the poor receiver is confused: Which is the real timing? The signal that has been delayed by faulty components in the system or the signal that got through "on time?"
When the IRD is confused the bit error rate escalates. When the BER exceeds the software parameters built into the system the picture and audio at the output of the receiver stop because the receiver believes the BER has climbed to a dangerously high level.
A poor impedance match at the input to the receiver, a poor quality signal splitter can cause all of this to occur. There can be other causes:
When supplying a digital satellite signal to multiple outlets, if splitters are used all ports that do not have a receiver connected to them must be terminated. To avoid this problem a better option is to use directional couplers or line taps.
The ports of a directional coupler are isolated from each other and do not require to be terminated due to their design. Directional coupler ports do not pass power, so the receiver cannot power the LNB from a directional coupler port.
This is overcome by powering the LNB with a separate power supply using a power injector to power the LNB or if a satellite distribution amplifier is used it will supply power to the LNB via its Sat input.
There is a common myth that you cannot reticulate L band satellite signals on the same cable system because its very hard to keep the higher frequency L band signal at the same level as their terrestrial counterparts, this could not be further from the truth.
What you have just read demonstrates that it is true that the higher L band satellite frequencies have more system loss, but it is still the UHF/VHF analogue terrestrial signals that are the limiting factor in the design because they require a higher signal level to produce a blemish free picture.
Should you have any queries or require any further information, please do not hesitate to contact me on +64 (03) 348-8641 or (025) 246-0089 Email mroos@teltrac.co.nz
Email us |
Home |
Request a Pricelist: |
Please Sign our Guest Book |
Search our Mega Web Site |
How to place an Order |
WebMaster |
Telsat® Communications Ltd
Palmerston North, (Knowledge City) NEW ZEALAND. Established in 1988. Tel: +64-6-356-2749, Fax: +64-6-355-2141. |