Jamming        Enforcement      Doppler radar  

This is something I've always been curious about ,  I guess more for the technical challenge than the  practical use of such a device ( I drive a diesel ok !) . Now I make this plain as day ,to use one of these devices is totally illegal and would land you in a heap of crap when you get caught. It would be a double whammy cos two government departments would come down on you like a ( metric) tonne   of bricks , the law enforcement division and the radio regulatory  people as well . So don't say you have not been warned!

Just about all enforcement speed radars work on the Doppler principle  when a eg radio wave ( or sound )  at a frequency of "n" Hertz strikes a reflecting moving surface some of its signal ( or all what ever doesn't get absorbed ) changes frequency, the amount of frequency change is directly proportional to the relative speed of the reflecting surface.\ and the incident wavefront . This is described by the formula below 

F doppler = 2R.Fo              R is speed in meters/sec.  Fo  frequency Hz   c  is speed of light m/s-1

Law enforcement Doppler radars work at three common frequency bands,  known as X band  typically 10.525 Ghz,  K band 24.150 GHz   and Ka band  33.5 Ghz and  a few frequency variants    up to 36 GHz ( 3 GHz wide so there's a lot of room to hide in up here!). They are all typically Gunn oscillator based devices  and operate as direct conversion receivers with a "Zero"  Intermediate frequency .The fact that they are Gunn oscillators is the first problem as they are not a particularly stable RF source (but that is of no consequence for their successful operation ) The IF is in fact in the audio range from 50 Hz up to maybe 10 Khz  depending on the frequency of the Gunn oscillator and the maximum speed the device is designed to display .  So there in lies another  hurdle to overcome, the sensitive D.C receiver is essentially narrow band  ( there is a constraint to this statement and I will explain it later  as you can also argue that it is in fact a wideband receiver as well !)  the detection range of all radars is described with the familiar Radar Range Equation.  I guess we are all familiar with the inverse square law for radio signals  ie, where to double the transmission distance you have to  increase the radiated RF power by a factor of four ?,  well in the radar equation, the detection range of a system is related to the fourth power ! that's why commercial radars eg ships, aviation , land based , have such powerful transmitters  (Kilowatts) to achieve good range!      
2           2 
r   =      Pt . G .  l .  s            Radar Range Eqn
            2             4     
p) .  R
Think of it this way,  there is the inverse square law operating when the radar transmitter sends a signal out to a remote reflecting object OK ? Now the signals bounce off that distant  object . Think of that reflecting surface again as a second  transmitter( it is sending out a signal is it not??   the fact that it receives its transmitting power from the  initial illuminating source is only of interest)  so its sends out its signal ,( even if it is only a reflection)  It still obeys the inverse square law but it is radiating its signal back to the initial signal source ( which at this stage is receiving signals )  so there is two inverse square laws involved here  add them together and you get the Fourth power in the equation.  That's heaps of potential signal LOSS,  The take home message is:   to get good range, a radar transmitter needs lots of  transmitting grunt! and because of the fourth power constraint in  the radar range equation , also means  you can pick them up way before they can pick you up ! 

Enforcement Doppler radars don't have a lot of grunt . The Gunn oscillator output is probably at maximum of about 100 mW . However at these frequencies you can have a lot of antenna gain in a small volume. ( typically 15-25 dB) of directional antenna gain associated with that 100 mW and end up with an EIRP of up to 10Watts down the boresight of the antenna .
                                 (+20 dBm and 20dB gain gives erp of 40dBm (10 W) !
The antennas typically have a 3dB  beamwidth of about ~ 20 degrees .  These radars are used in a directional manner ,so the high gain antennas are a plus ! you can pick off your target  , the detection beam is highly directional ! (that is a "reciprocal" statement for receiving the signal too)
The outer reaches of range of the typical stationary Doppler radar is probably approaching  a mile ( 2 Kms) on a good sized  large flat nosed vehicle i.e  large truck/tractor unit.  ( lots of plane reflecting surface,) . Given all the constant variables  ie one type of Doppler radar, the range is entirely dependant on the size and shape of the reflecting vehicle.  The range against a small motorcycle ( small frontal reflecting area) is poor, maybe 300  hundred meters, a low sporty "aerodynamically shaped "sharp" cars give less Doppler reflection that large "squarer" cars . ( go to google.com ,search  and read all about how stealth aircraft achieve low radar signatures)  The detection  range is somewhat less in a moving situation due to the large amount of signal clutter reflected off  the road directly ahead of the speed detection vehicle. this backscatter, swamping weak target  high frequency Doppler shift. However the advent of  Fast Fourier Transformations (FFT)  algorithms in DSP processing chips processing the incoming audio, do go a long way to minimise this  . There in lies another problem . The newer  Doppler radars are getting smart,  they can do spectral analysis on their targets and pick out the fast ( higher frequency Doppler shifts) from the slower speeds and also with the microprocessors do mathematical manipulation so a moving Doppler detection unit can show the actual "ground speed" of an approaching or receding vehicle,  another potential problem!

See what I mean by "Know thine enemy"  if you are going to  make a Doppler jammer  then you must take into account every conceivable variable  you may come across.  You must account for them or your final result my not work ,or end up with a number of limiting constraints or you may end up doing exactly the opposite to what you wish to achieve! ( end up with an enforcement notice for twice the speed you "were doing"  ). it is not an easy project (it was when enforcement radars were simple devices) but not now !
There are several ways, some are ideas I have not tried and others have suggested. There is the Brute force method .This I have tried.  You synthesise a Doppler shift and transmit it to the enforcement radar on the same or close frequency  . This is easy to do, simple electronically  and can be done using a ne555 timer  running as a  variable square wave audio oscillator generating the required "Doppler" shift frequency.  You run the 555 off a 6 volt regulated  supply and  supply the gunn oscillator from pin 3 of the 555 , (it will source up to 200 mA max), this is enough current for the average 10mW gunn diode.  Simply point the gunn osc with its 20 dB horn antenna, at the enforcement radar and bingo! you can make it read any speed you want it to read, by just turning the variable resistor that controls the square wave frequency
                Hawk being "suckered"                                         Stalker being "Suckered"

on the 555 oscillator .
However, you must use the correct frequency verses Doppler shift  i.e. at 10.525 Ghz  33Hz of doppler shift gives 1 mph  eg 3300 Hz square wave into the gunn osc  will show 100 mph on the enforcement radar. There are other scaled "Doppler constants" for 24 GHz and 33 Ghz  .  easy huh ! and yes it works, Ive tried it (against a  unit operating at 24 GHz) when you point it at the  radar! BUT in practice its about as useless as "titts on a bull" ( sorry to pop your bubble) I have seen variations of  these sort of circuits for sale on the 'net. Some are quite sophisticated with digital readouts for setting the "jamming speed" and are triggered into action by the receipt of a signal by a conventional radar detector. 
I give them full marks for the Concept of "electronic countermeasures" but I wouldn't trust one of those devices to protect me against enforcement radar here!  All the jamming circuits / units I've  seen are low power devices, maybe 50mW output ? into a 20 dB horn . I made up a simple device as described above with the NE555 and a 5mW 24 Ghz gunn osc and a 25db horn .              

    Output spectrum of the 555 modulated gunn osc                 24 Ghz  gunn osc and horn antenna 
The maximum "jamming" range I could get was about 250m  (you're excited about this aren't you ! I can tell ) that was the maximum range at which I could make the 24 GHz radar read the speed I wanted . HOWEVER when we tried it in a moving car coming towards the Doppler radar it was a dismal failure, the received "true Doppler" signal swamped the effect of the synthesised Doppler  jamming signal completely ,not enough grunt!. In order for it to work, we would have to increase the power of my jamming oscillator to a level such that the Doppler radar could pick up my jamming signal before it picked up its  true reflected signal. 
 This brings the variables into the equation like frontal area of the jamming vehicle verses detection range etc. Note that the above simple experiment shows that the closer you are to the Doppler radar the harder it is to jam!   Lets do some simple maths and try to work out how much jamming power I need to hit the radar before it gets me.  
        Assume my 250m jammer puts out eirp of +7 dBm and a 25 dB horn gives  eirp of +32 dBm ( just over 1 Watt )  To double the jamming range, I increase my gunn osc by 6dB ( 4X ) to eirp of +38 dBm, we now jam at 500m . Increase  another 6 dB , this takes it out to 1000m  ( eirp of 44dBm ) and to be safe we increase another 6 dB and take it up to an eirp of +50 dBm ! .( almost 100 Watts EIRP.)  That will enable us to make the Doppler radar read at 2 Kms away line of sight ( if we point straight at it AND it is looking straight down the  antenna boresight at us) ( this is in an optimum configuration hardly what you might find in real life or in a worst case scenario)  But what about the  antenna gain.? Well  we cant really increase it  much and there is a limit to the amount of gain we want to go to, as the higher the gain, the larger ( physically) and more directional it is , the less is the "spread" or coverage  of the "protective" signal ) 
 We have had to increase the output of our signal source ( gunn osc) by +18 dB!  we would have to use a +25 dBm  output oscillator , we are now into the relm of high power impatt diode oscillators and this   translates  to mucho $$$$$$ and high current high voltage sources 30-70 volts (depending on the manufacturer and the device etc.)  its looking like its much cheaper to drive at a legal speed isnt it !. 

There will be those out there who will be saying  "bollocks" I can talk on my 5 mW gunn oscillator transceiver to someone else  over a line of sight distance of greater than 10 miles ! and that's true, yes you can . I mentioned before that the Doppler radar receiver can be treated as a narrow band or a wide band receiver and I will clarify this, When we send our synthesised pulsed Doppler signal to the radar, we are "punching" into the receiver in the wideband mode . I'll explain .When we flood the enforcement radar with our signal, the diode in the receiver is operating in "video detector" mode ie converting pulsing rf into pulsing  dc at our generated audio rate ,  It is NOT MIXING  signals to produce an audio IF, so our radar receiver is just a crystal set. Much like those early passive radar detectors we used to make years ago that had a chopper diode ahead of the video detector diode in the tuned cavity . This video detected chopped DC signal was fed into a high gain amplifier and then a tone at the chopping rate was heard on a loudspeaker . In essence you "heard"  the enforcing radar.  They were not very sensitive compared to the superheterodyne receivers we have now (which incidentally still don't have an RF amp before the diode mixer, so they can be multi frequency) 
A straight video detector diode operated in an optimal mode with  correct bias matching etc has a TSS (Tangential sensitivity) minimum detectable signal  of about -50 to -60  dBm ! that's still a lot of signal . Remember that is a diode with good matching and optimum bias current , NOT a heavily biased mixer diode sitting in a cavity swamped with a couple of milliwatts of RF from  the gunn oscillator cavity next door!  In video detector mode and the gunn osc going , the diode in "TSS mode"  as deaf as a post!   (There are some enforcement radars that use this principle to see if they are being jammed , if there is still audio signal coming out of the detector/mixer diode when the gunn oscillator is turned off then this must be an interfering signal and not a true Doppler generated one !) and the read JAM on the digital display  !   

                                                Clever Hawk detecting Jamming !!!!!!!!
This hopeless lack of sensitivity  in broadband mode, is  why we have to hit the Doppler receiver with so much RF to make it read our synthesised speed .  
This is why I wouldn't have much faith in those Internet  passive radar jammer plans.....  just not enough returned signal !  
The other problem is that the Doppler receiver diode is hugely more sensitive in MIXING MODE  and we are talking microvolts of signal (-100 dBm ) so its at least ~50 dB more sensitive in "narrowband"  mode ( crystal set verses zero IF "direct conversion" )  it is much more sensitive to its own synchronised reflected Doppler signal , that's why I couldn't jam  or over ride the radar with my gunn oscillator. 

The only way we can have any real cost effective effect on the Doppler radar is to get into its receiver in "Narrowband mode" . This is very tricky and if we could do it  in "narrowband " mode we could knock it off with the described  5mW gunn osc  and antenna from a couple of miles line of sight. ! it is not easy and I haven't tried this idea yet. You need to phase lock your jamming oscillator in CW mode to within a couple of Kilohertz, of the doppler radar signal , this will enable a mixing process to occur within the radar receiver and produce a very strong interfering signal , There in lies the difficulty, Varactor controlled Gun oscillators with  wide  tuning range, particularly at 33 GHz would present a problem . also the doppler radar  gunn oscillator is not particularly stable either. Now we have two  independent variables to account for ! to lock the jamming oscillator, one could produce a simple Direct conversion receiver at the frequency of interest with a very wide bandwidth IF  , this IF could be fed into a frequency/ phase detector  PLL ? operating at ~2 Khz ,the  amplified voltage out put of the PLL phase detector drives the frequency controlling varactor of the jamming oscillator? . Thus one attempts to lock the jamming gunn oscillator to the Doppler radar . This system would have to be modulated by a ramp voltage on the tuning, to sweep the  jamming receiver across a range of frequencies of interest. The ramp voltage would be inhibited when the wanted signal appeared within the bandwidth of the receiver allowing "lock" to be achieved .  In the interest of rapid acquisition of lock, one may have to start with a very wide band IF of many MHz then switch to narrower bands to "zoom in" on the wanted signal .This would have to be done incrementally as his is similar to having the sweep speed on the spectrum analyser too fast when in too narrow a bandwidth IF filter, at too great a dispersion , The display response is so distorted and at a much reduced level ,this is what would happen with a narrow audio IF and a fast wide sweep by the search ramp voltage.   starting to get complicated isn't it ?  ( its easier just to drive slowly !!)  but not nearly as much an electronic  challenge!

The Third idea suggested by a friend Peter Williams  is in the same vein as passive radar jammers ( that work in theory but DON'T work in practice  for the ratio of  reflected to modulated  signal strength reasons I've mentioned above) . This technique gets around the problem of synchronising the return signal to the transmit  signal so we can get into the radar receiver in "narrowband" mode . We use the received signal as our source of  our transmit signal . like a transponder principle?  Pick it up on a horn antenna amplify it by 50-70 dB (by what ever maximum level cheque book and rf isolation allows,) then retransmit it back to the Doppler radar BUT  chop the final amplifier off and on at an appropriate audio rate  to OOK (off on keying) modulate the returned signal , This will put sidebands on it! I'm not sure how this will go as the return signal already has a Doppler shift on it and we are adding another double sideband signal on top of that as well ! god only knows how the Doppler radar will interpret it  (we could make matters worse for ourselves )

The Fourth idea I had is a variation on the one above.  Use a grunty super regenerative detector quenching at around 3 Khz or so.(I dont know if you have seen on a spectrum analyser how wide the   noise is that  comes out of one of those things!)  to lock to the received signal and re radiate a dirty  strong signal back to the  doppler radar . Super regens at this frequency may not be all that sensitive and   from the research I have done on them,  maybe only in the order of -60 to --70 dBm for lock to occur . However they are a very  simple receiver . I have had a 10 Ghz gunn oscillator self super regenerating but the sensitivity was very poor! though it did  quiet when it received a strong signal !


Seen on the dashboard of the average patrol  car !

This is the 33 Ghz  casing                                                     close up shot of PC board  with low noise op amp (LT1115)      and  AGC     chip  LP2951



Business end  Tx cavity and Rx cavity  face each other  note small phase shift  probe to convert to circular polarisation    33.5 Ghz gunn assly  5.5 v   perhaps 100 mw ?? 

 The  unit operates  off 12 volt , so i wired up 12v gell cell to the doppler  module, also I connected an Audio amp with speaker  to the audio output . The  unit is very sensitive ,  you can "hear  " people  walking  down the  road 100m away and it reminded me of a heart beat through a doppler heart beat monitor , the  swish swish of the legs moving .I guess the shorter wavelength gives better resolution ?   Anyway I drove to a good long clear  length of  road to assess the maximum range of the doppler  unit . 
The first thing I notice is the frequency  of the doppler shift is very high, about 100 Hz per mile an hour  at 33 Ghz , so vehicles at 50-60 Kph have a very high doppler  pitch ,,too fast and they are  hard to hear ,( but no problem to the DSP electronics )  the slow 30 Kph sounds much more audible . The max  range for average size motor vehicles ,  I can hear with my own ear, is around 1000 m and you must have the unit aimed  down the boresight at this  range. Going by the "directivity " of the unit Id say the horn has quite a high gain  , so to get good detection range,  the officer using the device has to aim it well !  I may  try and work out what the beamwidth is ??   pity it did not have a varactor diode in the gunn cavity  as it would be cool to hake a CWFM radar out of it ! 

So there are a few ideas on the electronic countermeasures , as you can see to actually make any impact takes a huge amount of clever circuitry and $$$$ .  it is much easier and cheaper to drive responsibly !!    however  one could take another approach as used by stealth aircraft and absorb the  reflected microwave signal !! , There are available carbon impregnated foam polyethylene and conductive rubber that can be layered over the vehicle to reduce its radar signature...............  its not cheap $$$$.....

Well the next test is to take a large  piece of conductive  foam  1000 mm  by 1200  mm    as used to insert IC  into for  static protection .and wrap it around the front of my motorcycle  !   we will do a simple range test deriving reflected signal strength  using the AGC  voltage derived from the  agc chip in the  direction conversion Rx  board  with the motorcycle  naked ( no foam ) and then do another run with the  foam wrapped around the  front area . 

 Unfortunately  the foam I have is of relatively high resistance , several K ohms / cm  so may not work as well as foam with an impedance/resistance  of 377 ohms per centimeter . The  problem is although the carbon impregnated conductive  foam will absorb microvaves well , it is a rapid  impedance transformation between free space and the  material . When you have mismatches ,  you get reflections !    these are what we are trying to minimize so i am not sure how well this experimet will work . Real microwave absorbant foam  is of the correct resistance and usually has pyrimids  shapes cut into it so reflected signals bounce around in the vallys of the pyramids and eventually get absorbed with very little signal being  reflected .   This could be impractical on a vehicle  however would work but would be a bastard in the rain , it would act like a sponge  your motor  bike would end up very front heavy.    to be continued ...................

.......  but  that is another story  !!  drive safely !!