Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!clyde.concordia.ca!nstn.ns.ca!news.cs.indiana.edu!att!linac!pacific.mps.ohio-state.edu!zaphod.mps.ohio-state.edu!unix.cis.pitt.edu!dsinc!netnews.upenn.edu!eniac.seas.upenn.edu!depolo From: depolo@eniac.seas.upenn.edu (Jeff DePolo) Newsgroups: sci.electronics Subject: Re: POLICE hand-held RADAR units Message-ID: <37578@netnews.upenn.edu> Date: 14 Feb 91 21:36:10 GMT References: <1991Feb12.035201.16098@nntp-server.caltech.edu> <2470014@hp-vcd.HP.COM> <1991Feb14.015812.14576@nntp-server.caltech.edu> Sender: news@netnews.upenn.edu Reply-To: depolo@eniac.seas.upenn.edu (Jeff DePolo) Organization: University of Pennsylvania Lines: 94 In article <1991Feb14.015812.14576@nntp-server.caltech.edu> fjs@nntp-server.caltech.edu (Fernando J. Selman) writes: >This is the power of the >doppler effect, that it allow you this type of accuracies, when >USED PROPERLY. But I doubt you can eliminate systematic errors to >this level using one of the toy guns produced by MPH Industries, Inc. >The cop in question had less than a second to make the measurement. I was >changing lanes from just passing another car. The uncertainty in the >cosine factor would be enough to reduce the speed measured from 71mph >to 65mph. That's quite a bit of error from the cosine effect. This means that you would have to be off-axis by almost 25 degrees. If the cop was 500 feet away, you would have to be over 200 feet laterally from his position. Considering how wide lanes are, this would be difficult, unless there was a very wide median. Even if being clocked while changing lanes away from the radar gun, your lane-change vector wouldn't come near the 25 degree offset. It's probably on the order of 5 degrees, at highway speeds for a normal lane-change. Say your lane change took 1 second. For a 15 foot-wide lane, this would yield a vector of only 8 degrees off-axis at 70 MPH. Considering a lane change usually takes more than 1 second, this is an overestimate. Both being off-axis and changing lanes away from the radar source would obviously yield a speed reading LESS THAN your true speed, so it doesn't buy you anything in court. >The humidity of the day was 82%, even though the manufacturer of >the toy gun claim 90% for operational limit, I highly doubt this number >as I have never find electronics instruments rated for work >above 80% noncondensing. All of the radar gun info I have gives manufacturer's ratings that are valid up to 90% humidity at 99 degrees. The MPH K-55, however, has a bad reputation of being unstable at high temperatures (per NBS Tests, among others). I've never seen any definate effects of high temperature in the radar guns I've played with, although this is more or less on an experiemental basis and not in an oven in a laboratory. > As for the main point in your posting, if the speed of light >were to be half that assumed by the logic of the instrument, no amount >of calibration would allow you to have a proper reading for all frequencies >not equal to the calibrating frequency. The fact that we measure a >frequency shift seem to confuse a lot of people into believing that >the speed of light is not the gauge in a dopple radar. Refresh your >physics! The doppler shift is directly proportional to the speed of light (radar), true. So, the error in the doppler shift (that is, your speed) is directly proportional to the error in the true speed of light (radar) relative to the design estimate. This error is negligible - we know accurately how fast radar travels to well within a small fraction of a percent. The error induced due to the variation in frequency from the design frequency of the radar gun vs. the true transmitter frequency is also negligible (maybe 10 MHz/10.25 GHz = > .1%). The point is, on a road free of other traffic, free of sources of interference, at a 0-degree angle, with normal ambient temp/humidity conditions, the radar gun should be accurate to well within 1 MPH. If any of the above conditions is changed, all bets are off. Most authorities consider +1 MPH/-2 MPH to be a good margin of error. I consider this wishful thinking. There are too many variables involved in doppler radar to make it foolprof, in particular, target acquisition, interference, proper calibration of the counter, frequency alignment of the transmitter, and proper operation by the operator. In summary, the following things aren't going to help you get off: error in the estimate of the velocity of a radar wave, error in the transmitter frequency of the radar gun, variatons in the speed of the radar wave as it passes through the atmosphere due to changes in density/temperature/ etc. These might work: it was too hot, it was too humid, there was external interference, there was another car around, the gun was not being operated properly, the gun was broken, the gun wasn't calibrated recently, the gun wasn't certified by the State, the gun wasn't checked with tuning forks before and after the stop, martians ate my speedometer (faulty-vehicle approach). The design of the radar gun makes it extremely accurate under perfect conditions - your only hope is to bring up of the less-than-perfect conditions and exploit it. Ignore the theory behind the transition of 10.525 GHz into 71 MPH - the judge just plain doesn't care (he's a lawyer, not an engineer :-) --- Jeff -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Jeff DePolo N3HBZ Twisted Pair: (215) 386-7199 depolo@eniac.seas.upenn.edu RF: 146.685- 442.70+ 144.455s (Philadelphia) University of Pennsylvania Carrier Pigeon: 420 S. 42nd St. Phila PA 19104