A slow wireless bridge: finding the cause

First the numbers, then the guesses.

Complaint “bridge brakes” without measurements does not mean anything: everyone has his own idea of speed.

Three things are measured: signal level, signal-to-noise ratio, and real bandwidth between points.

The bandwidth is removed by an internal test of the device, not by downloading a file from the Internet.

Measurement is done in both directions: the channel is often skewed, and the bottleneck is visible only in one direction.

If you write down three numbers, you can go on; without them, any step will be at random.

An antenna with a narrow beam loses a noticeable proportion of the signal already at a deviation of one or two degrees.

Hover down the level indicator, slowly, first horizontally, then in slope.

After tightening the fastener, the level is checked again: the clamping often shifts the antenna when tightening.

Both sides take turns and repeat the cycle, otherwise the second point “catch up” to the first.

The level recorded after the tip is the reference point, which is compared after six months.

There is little direct visibility: around the line between the antennas there is an area that should also be free.

A branch, roof, or visor touching this area takes away the signal, even though the antennas can see each other.

Summer foliage is a classic cause of seasonal degradation, which passes by itself in autumn.

Raising the antenna a couple of meters is often cheaper and more reliable than changing the equipment to a more powerful one.

Reflection from a flat roof or water on the track is also harmful: the signal comes twice and interferes with itself.

In free bands, neighbors appear without warning and yesterday's clean channel is busy today.

Built-in review of the air shows who is sitting and where, and it should be done from the roof, not from the office.

The channel is chosen not by number, but by the picture of employment, and rechecked once a season.

A wide channel gives more speed, but collects more interference – sometimes a narrow channel works faster.

Automatic channel selection is convenient until it starts switching in the middle of the working day.

Street dots are most often powered by the same cable that the data goes through.

On a long cable, the voltage sags, and the device begins to reboot under load.

Cheap cable with a low-cost residential behaves worse than copper on power supply.

The connectors on the street are sealed, otherwise the water rises through the cable and sits directly on the board.

Thunderstorm protection is placed at the entrance to the building, and it is usually the first to show that the line passed impulse.

Heavy rain at frequencies of five gigahertz weakens the signal, but usually does not tear the channel completely.

Wet snow on the antenna harms more than rain, and the radio transparent casing here helps.

If the channel falls only in bad weather, then the stock level was initially too small.

A normal supply is a few dozen decibels above the threshold, not “works, and fine.”

When there is no stock, it is not the change of equipment that helps, but the lifting of antennas and honest alignment.