In Limiters

By Harley Berman

The What and Why of Limiters

Limiters are crucial building blocks in Tx/Rx systems and are used to protect sensitive circuity from high power signals.  Take for example a simple radar system, which consists of a high-power pulse transmitter sharing an antenna with the Rx signal chain shown below.  The pulse exiting the power amplifier is well above the damage threshold of the LNA and, depending on the isolation of your duplexer, this pulse has the ability to cause catastrophic damage to your LNA.  To solve this problem, system designers commonly place a limiter between the antenna and LNA, which protects the sensitive LNA and Rx signal chain.

The Self-Jamming Problem

Limiters are devices that have very low small signal insertion loss, however as the input power to the limiter increases past a critical input power level (referred to as the limiting threshold), the insertion loss begins to increase linearly.  This limits the output power to some maximum level (referred to as the flat leakage) regardless of input power.


Using the Pout vs Pin plot on the left for an ideal limiter with a flat leakage and limiting threshold of +10dBm, it is clear how this limiter would protect sensitive circuitry from high powered signals, assuming the sensitive circuitry has a damage threshold ≤+10dBm.  What is less clear from the left plot is the negative effect this has on your system.  To see this more clearly, we need to look at the corresponding insertion loss versus input power plot shown on the right.  This plot shows when you are jamming yourself with your Tx signal and engaging the limiter, you cannot detect the return pulse because it is being attenuated by the high insertion loss of the limiter.  This effectively blocks all communications to your system until the self-jamming signal turns off, leaving you in the dark.  If this scenario occurs in a radar system, it limits how close to the antenna something can be detected. When the pulse returns quickly, as it does for close objects, you will not detect anything because the limiter is in a high insertion loss state.  This naturally introduces a key metric in a limiter – the recovery time, or, how fast the limiter can relax to a low insertion loss state after being jammed by a high-power signal causing high insertion loss.  The figures below show time domain representations of what this may look like using limiters with different recovery times.


Limiter Technologies and Their Recovery Times

There are several different technologies used to implement a limiter, each of which influence the recovery time of the limiter in a different way.  The simplest method of creating a limiter is using a single PIN diode with a parallel DC path to ground through an inductor.  One of the main benefits of this limiter technology is its simplicity of implementation, however the recovery time is typically on the order of a couple of nanoseconds.  To lower the limiting threshold of this limiter, one could add a Schottky diode antiparallel to the PIN diode to create a Schottky-PIN limiter. This has the benefit of lowering the limiting threshold however the recovery time remains roughly unchanged.  For better recovery times, one could use a CMOS SOI process, which has typical recovery times on the order of 1ns, however, if having the fastest possible recovery time is a priority, the best limiter technology to date is the GaAs Schottky diode process used on Marki Microwave’s HLM and DLM series of limiters.  It has been experimentally shown that the recovery time of our HLM series limiters (and by similarity our DLM series of limiters) is nearly instantaneous and are the best limiters available on the market today to solve the self-jamming problem outlined above.

Limiter Technology Representative Unit Recovery Time
GaAs Schottky HLM-40U Nearly instantaneous
CMOS SOI pe45361 1 ns
Schottky-PIN SKY16602-632LF 5 ns
PIN Diode SKY16601-555LF 5 ns


HLM-40U’s Nearly Instantaneous Recovery Time

I just made a rather bold claim that our limiters have a nearly instantaneous recovery time so now it is time to back up this claim with some experimental evidence.  To measure the recovery time of a limiter, one must inject a small signal CW tone to the limiter, which represents the signal of interest, and simultaneously inject a high-power pulsed tone which, represents the jammer.  To measure the recovery time, we want to view the signal of interest in the time domain immediately after the high-power pulse turns off and measure the time it takes for the signal of interest to go back to full power. This is the definition of recovery time.


The block diagram above was implemented in our lab using a C13-0140 directional coupler to combine the large signal jammer and the small signal input together.  The output of this was fed into an HLM-40U, which is the device under test in blue, and an FB-2250 bandpass filter was used to select just the 22GHz signal of interest.  An ATN10-0067 10dB attenuator was placed between the DUT and filter to reduce the high-power reflections of the jammer signal off the filter.  The output was fed into an oscilloscope with the measured output on the right (jammer signal in blue and desired signal in pink).

Immediately, it is clear that the pink trace does not look like either of the responses shown above for recovery time, since the measured waveform has some peaks while the high-power jammer tone is still on.  A closer inspection of this plot shows that these peaks only appear when the large signal tone representing the jammer crosses 0V.  This means that the HLM-40U is responding to the jammer’s zero crossing as if there was no jammer on at all and relaxing to a low insertion loss state almost instantaneously!


This result highlights that the HLM-40U is an ideal candidate for limiter applications that call for a near instantaneous recovery time. For questions on how we obtained this result, our limiter portfolio, or technical inquiries, reach out to [email protected] and we would be happy to help you. To use the HLM-40U in your system, contact [email protected]

Recommended Posts

Leave a Comment