Unlike some communications signals the E1/T1 physical form is very specific and can be sensitive to incorrect termination, transmission or cable lengths. This 2.048Mbps signal stream is in fact bipolar and follows a precise waveshape defined in ITU recommendation G.703. For maximum robustness against electrical interference and attenuation, the signal envelope must fall within a pre-defined mask, for both amplitude and timing. While most equipment is capable of operating beyond the defined limits, there is no requirement to do so and good electrical and physical design is required to properly accommodate these signals.
When it comes to splitting these signals, our experience is that complete termination using correct transformer coupling, line interface units configured for line length and impednace, and re-generation of the E1 signal chain is the only waty to ensure error-free performance.
It amy be tempting to use video switches for these applications. At first sight, these appear to have plenty of bandwidth, bnc connections, perhaps transformer coupled and a wide variety of switching options.
If you fall for this temptation, you may be lucky and find that it doesn’t work at all.
if you are unlucky you’ll find that it appears to work, if you’re very unlucky it will appear to work well.
To explain, it is necessary to know the difference between composite video signals and E1 signals. The biggest difference is amplitude. Composite video signals are nominally 1V peak, compared to the 2.4V peak of E1 signals. This appears to be the right way round, feeding a larger signal into a smaller signal input might be ok (?) and look, the impedance is 50 ohm instead of 75 ohms, so it gets attenuated to the right sort of level anyway! Result!.
Well not quite. Unfortunately we need to consider what happens when that 2.4 V signal overdrives the video input. Let’s start with the coupling transformer ( if there is one). Transformers for video signals are very carefully designed to be suitable for the frequencies and amplitudes of video signals, especially when optimised for physical size as well. The overdriven transformer may go into saturation, causing it to stop transforming – in other words it cuts off the incoming signal amplitude at some ( practically indeterminate) level. Which still sounds ok , until we consider the effect on the pulse width. The ITU mask includes a very precise definiton of how wide the E1 pulse can be. A saturated transformer causes a variation in pulse width dependent on the signal content . That is, the variation in pulse width will depend on the signal content. Another term for vriation in pulse widt is jitter and this can have a profound effect on the signal timing. The effect on data can be quite subtle – occasional bit errors, perhaps a lost frame or even bit slip. And it will depend on a combination of data content, signal amplitude, existing signal jitter and electrical noise on the line.
In fact the situation is just as bad with non-transformer coupled video switches. Capacitive coupling causes asymmetry in the pulse width ( more jitter) and diode clamps on the video switch will also cause early signal cut-off, again leading to jitter.
And of course when considering the output signal from the video switch, you are dependent on the following E1 circuitry working beyond specification to capture the signal, which at best will be only half the amplitude required.
The errors created by using the wrong type of switch/splitter/converter for E1 signals can be quite difficult to track down and often blame is wrongly attributed to cable connections, lengths, E1 circuitry and so on. Worse still the system may appear to work for some time, before becoming unreliable.
So in conclusion, use the right tool for the job! Our range of E1 splitters are constructed to telecomms standards with telecomms level of testing. To view these, follow this link.