Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Schottky or PIN diodes for TVS / ESD protection of LNA input?

Status
Not open for further replies.
G

gbugh

Member level 5
Joined
Mar 15, 2013
Messages
80
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Location
Fort Worth, Texas
Activity points
2,236
I see a lot of Schottky diodes designed and advertised specifically for TVS and I don't see PIN diodes advertised for TVS yet I see PIN diodes that are cheaper, have lower capacitance, lower ON resistance and just as low or lower leakage current.

The PIN diodes have a higher forward turn-on voltage also, which means I can receive high voltages from the magloop's high Q voltage peaks combined with strong local signals and it won't clip the input signal as quickly as the Schottky diodes would.

It seems like I have to use 2 Schottky diodes in series in each direction to do what just 1 PIN diode in each direction could do.

So can I just as well use PIN diodes to protect a high input impedance JFET input LNA used to sense a voltage maximum across the tuning capacitance of a magloop antenna?

Thanks, George
 

PIN diodes aren't suited for TVS applications. You should get a zener device actually specified for TVS operation. Schottky diodes may work, probably can't handle the same peak power levels as a true TVS.
 

NXP has released recently some TVS diodes for Wireless devices.
You'd look at..
 

BigBoss said:
NXP has released recently some TVS diodes for Wireless devices.
You'd look at..
Click to expand...

There are two separate problems here: TVS means voltage-spike protection which is done by very fast Schottky diodes and TVS limiters are designed specifically for coaxial or stripline applications.

PIN limiters are typically designed as POWER limiters. PIN diodes may be slower, so they would not limit voltage spikes.

You have not defined the frequency of your LNA to be protected. At high RF frequencies it is a good idea to use a high-pass filter before the LNA, then there will be no voltage spikes calling for TVS. A waveguide section as a high-pass filter would not allow any "voltage" to pass at all.

PIN limiters are mostly designed for use in 50-Ohm lines. One class of PIN limiters uses a Schottky detector to generate the control current for the clamping PIN diode. Such limiters are fast, at in nanoseconds. Such power limiters do allow some power to pass through (leakage), and the LNA must be designed to withstand such leakage.

I would recommend to check Herotek and Mini Circuits for details of their limiter products.
 

jiripolivka said:
You have not defined the frequency of your LNA to be protected.
Click to expand...

The LNA would mount on a 1 or 2 meter diameter magnetic single loop antenna for covering 100KHZ to 30MHz. The LNA would sense the high Q voltage across paralleled varactors that would be used to vary the frequency, along with switching in various fixed capacitance values. I'm trying to keep the magloop Q close to 400. I was thinking I need a TVS device for lighting strikes near the antenna but most TVS devices have too much input capacitance that would swap the varactors' lower capacitance limit and so stop me from tuning as high in frequency. The LNA's input JFET, has a negative voltage limit that could be exceeded before a zener kicks in. I had settled on using Schottky diodes until I started studying PIN diodes more. I wanted to experiment with difference input JFETS and some of them have only a 3.5 volt max negative gate voltage before it damages the JFET. Unlike most LNAs, this design needs a very high input impedance, low leakage, low capacitance, rather than a 50ohm input impedance.

So it seems from what you all are saying, that fast, ultra low capacitance Shottky diodes are the only thing I can use but they could still get blown by a nearby lightning strike so I may have to live with that. Is that correct?

George
 

I like the low capacitance of that gas discharge tube.

I'll see if I can find one even lower than 60 volts, then after small coupling capacitors I can have Shottky diodes at the gate of each input JFET.

- - - Updated - - -

I don't see any lower than 60V. I'll try your recommendation, thanks!
 

gbugh said:
I like the low capacitance of that gas discharge tube.

I'll see if I can find one even lower than 60 volts, then after small coupling capacitors I can have Shottky diodes at the gate of each input JFET.

- - - Updated - - -

I don't see any lower than 60V. I'll try your recommendation, thanks!
Click to expand...

If I may, I would add some points to your design:

1. Using a LNA at frequencies below 30 MHz makes no sense as the galactic noise is very strong, so the background noise in the receiver is not defined by the thermal noise but by that external noise.

2. PIN diodes are "slow" diodes at above ~20 MHz, so their attenuating effect will be not much suitable below that frequency. Schottky diodes are faster and frequency independent, therefore I would recommend to use them in a limiter in front of LNA.

3. The above idea to add a small neon tube limiter is good. Due to its very low capacitance, you can add it into the antenna "feed" if there is some high-impedance point.

4. For the frequencies below 30 MHz, the JFETs as well as bipolar transistors are all suitable for receiver front-ends. To prevent damage from lightning spikes, the best is to set :LNA current as high as possible, and have a spare plug-in LNA ready, just for case. I have seen many designs of lightning detectors, with long wire antennas, and nobody reported any damage by lightning itself. Only if it hits nearby; then no protector would survive.
 

jiripolivka said:
If I may, I would add some points to your design:

1. Using a LNA at frequencies below 30 MHz makes no sense as the galactic noise is very strong, so the background noise in the receiver is not defined by the thermal noise but by that external noise.
QUOTE]

Thanks jiripolivka, I understand. That had been mentioned in an earlier thread also. I will have a later need for a similar circuit inside a faraday cage a stimulated ring of magnetic material that will be stimulated at its ferromagnetic resonance frequency, adjusted with static magnetic field bias. I am planning experiments with recieving signals in what is normally the null direction of the magloop by a similar magloop in another faraday cage. I'm expecting I may need the extra sensitivity.

My hypothesis is that magnetic field energy is radiated in the null direction and can still push and pull on precessing electron spins even though the flux direction is in the propagation direction so does not cut across conductors (faraday cage walls) so does not induce current flow. Still with a tuned sample of precessing electrons, it can pick up the signal even within another faraday cage.
Click to expand...
 

gbugh said:
jiripolivka said:
If I may, I would add some points to your design:

1. Using a LNA at frequencies below 30 MHz makes no sense as the galactic noise is very strong, so the background noise in the receiver is not defined by the thermal noise but by that external noise.
QUOTE]

Thanks jiripolivka, I understand. That had been mentioned in an earlier thread also. I will have a later need for a similar circuit inside a faraday cage a stimulated ring of magnetic material that will be stimulated at its ferromagnetic resonance frequency, adjusted with static magnetic field bias. I am planning experiments with recieving signals in what is normally the null direction of the magloop by a similar magloop in another faraday cage. I'm expecting I may need the extra sensitivity.

My hypothesis is that magnetic field energy is radiated in the null direction and can still push and pull on precessing electron spins even though the flux direction is in the propagation direction so does not cut across conductors (faraday cage walls) so does not induce current flow. Still with a tuned sample of precessing electrons, it can pick up the signal even within another faraday cage.
Click to expand...

I can see that you plan a challenging experiment! For such special case, please consider my points above but b ready to experiment. Then according to first results you will have to introduce modifications till you achieve a success. Please try to google "magnetic resonance hardware", there are low noise amplifiers designed to run in strong RF interference in MRI systems. This can help you.
Click to expand...
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top