Schottky for surge & polarity protection?

[kevj]

until we tried a pair of anti-parralel shottkys across the input. The earth braid to ground had to be beefed up to reduce noise.

Thanks Brian. I hear what you're saying, but just can't picture how exactly such a thing would be implemented.

I've got an antenna - a piece of wire... with a signal that goes to an input of a "radio stage", it sits inside a grounded SMA connector with no electrical connection to the signal line obviously.

If I were to tie a single shottky from the signal line (antenna) to ground, as it has a low forward voltage drop, but it still does have "some" voltage drop - the very low voltage signal itself shouldn't be effected, but if a spike were to happen, it would cause the schottky to "open" and that voltage would be quickly dumped to ground. Am I picturing this right? Sounds too simple.

Also, as EMP is a wave, it's actually kind of an A/C signal induced in the antenna - so should I put a pair of schottky's - one pointed either direction?

Thanks everyone. Don't know why I'm having such a hard time picturing this.

[BrianT]

A single diode won't do it unless you know the pulse is always the same polarity which is highly unlikely in RF systems. The shottky diode does not 'open' as you mention. It will be reverse biased or forward biased depending on the polarity of the incoming signal. The diode will withstand some reverse bias and breakdown at some voltage from 20 to 400 depending on the diode specs. That 400 volt signal is being poked straight into your radio front end so you must eliminate this reverse breakdown by using two diodes where one protects the other and the voltage across the pair is never more than the forward voltage drop of the diode - about 250 mV

Your 'single wire' antenna has both an earth part (the outer of the SMA) and the active line (the centre pin on the SMA). Two shottky diodes in "anti-parallel" means one has the anode to the antenna with cathode to earth and the other is the reverse of that. The forward voltage drop of the shottky is about 250 mVolts so no matter what the polarity of the offending spike is, the diodes clamp the voltage to no more than 250 mV. RF, which is typically below 10 mV rides through unaffected.

HTH
BrianT

[kevj]

Outstanding. Thank you very much Brian - exactly what I was looking for.

[Ioannis]

Have a look at the file I attached. I think it is exactly what you want.

Ioannis

[languer]

Transorbs (like SMBJ) and other TVS (transient volt suppressors) are great for transients. In fact, for lightning protection and such TVS devices are used widely (these are what you find in your surge protection power strips). Schottky diodes would never whidstand lightning transients (and live to see another day).

But schottly diodes have very little capacitance, transorbs generally have a great deal of capacitance. So if you hang a transorb directly across the antenna input, you may as well not worry about reception / transmission (since you will have none).

In your case, a schottky diode in an anti-parallel configuration (as was suggested) maybe all you need. Should you require to add the TVS (there are some very good bidrectional ones - look up AVX's TVS) to the antenna line, you will want to have an RF choke (RF inductor) in series with the TVS. The choke connects between the antenna line and TVS, and the TVS betweeb the choke and power rail (GND most likely). Lookup the following appnotes, they have very good info:
What is a Silicon TVS and How Does it Work?
Selecting the Optimum Transient Voltage Suppressor