# Measure phase difference with less than a full wave?

I am interested in aircraft altitude measurement during approach. Off the shelf sonar doesn't provide enough distance, and inexpensive radio sensors don't provide a short enough distance. I have seen other radio wave projects attempt distance measure with "time of flight" and believe this is the wrong approach - a difference of phase angle would allow slower electronics and less critical rf circuits.

A search of this forum shows a number of discussion how impractical "time of flight" calculations are for simple electronics. I suspect the signs in many cities showing car speeds are using phase angle and not "time of flight", but they seem to quit measuring at 30-50ft.

I am trying to figure out what the lowest frequency will be needed for measuring 5ft to 200-300ft. Of course measuring down to 5ft is the critical frequency limitation and generally several cycles are needed for a the transmitted wave to measure accuratly. The table here is partially reproduced below.

  Freq      length     1/4 wave
................................
50  MHZ     6m         1.5m
100 MHz     3m         28cm
300 mhz     1m         25cm
500 mhz    .6m         15cm


Due to a less than perfect world, it looks like measuring down to 5ft (10ft round trip) will need at least 300mhz (1m full wave), but the 1/4 wave column got me wondering if some how less than a full wave can be measured accurately? I understand the 1/4 wave data is supplied for antenna purposes.

My project would be much simpler if I could use a 100mhz transmitter instead of 300mhz. I have found an off the shelf sensor (AD8302 board) for less than $10 with 1 degree phase resolution. Does anyone know if there is a simple theory or whether it is possible to measure phase difference with less than a full wave? Thanks • Proper low power pulsed radar is much better than what you are talking about. – Andy aka Jan 21 '18 at 23:32 • Police radar uses 30 GHZ pulses and trig equations to calculate speed, but the close the vehicle gets, the faster the CPU must calculate the average speed. Using digital filters you need to get at least 4 samples for a crude average, hence the close range limit for Police radar. At 5 feet ultra-sound would work much better, but not for a fast moving vehicle. Hence the conundrum. All types of velocity/distance measurement systems have a finite limit. – Sparky256 Jan 21 '18 at 23:49 • There is a trick called under-sampling, but it needs many samples (64 or more) to be accurate. I think that Police radar could do what you ask but it would need more software and a mini super computer to process the data in time. – Sparky256 Jan 21 '18 at 23:52 • Thanks I have been mulling over this project for several years. I investigated the RADAR approach a couple years ago. The closest I found needed an antenna made out of a large coffee can and had a very large circuit board. There is even a university group that has a public project on the Internet somewhere that I looked at. All the Internet projects I found measured speed, not distance. I need a solution more the size of an apple or orange. I think I can get the AD8302 board to work but I wanted some expert criticism before I spent time and money. – jwzumwalt Jan 22 '18 at 0:39 • According to your chart, accuracy at just 5 feet demands a 2 GHZ pulse burst for just a crude average. Your correct in that no one can build microwave boards as a hobby. Maybe look for "Buy radar" as a search term. Maybe something antique but usable out there. – Sparky256 Jan 22 '18 at 0:55 ## 2 Answers Use a chirp! Setup a low power transmitter, CW in say the 24GHz band (smaller antennas) arrange for it to sweep in a linear ramp upwards in frequency by maybe 100MHz or so over a few tens of ms. In the middle of said sweep you use a mixer with the IF fed from the transmitters output to mix with the reflected RF. At the mixer output you get an audio tone frequency proportional to range to target, process as appropriate. If you have much vertical movement you may want to average the result from an upward and downward sweep to reduce the impact of the Doppler shift. A completely alternative approach: Why not take a leaf from Bomber Commands book, specifically the dambusters mission, where they measured altitude by mounting two searchlights on the airframe spaced some distance apart and mounted at different angles, looking at where the two spots of light fell gave them a way to judge altitude. The modern approach is probably a quadrant photodiode with a collimator and a laser diode mounted on a servo, pulse the laser and use sync demod to detect the return then use the quadrant photodiode to adjust the servo to centre the reflected spot in the diodes field of view, you now know the angles, so altitude is just a case of trig. • +1 My expertise is in digital not rf electronics, but I understand most of your suggestion. I had not thought about antenna length except maybe I could use an antenna 1/10th or 1/100th of the wave length. I have no way to make a circuit board or test ghz wave lengths. my O-scope only goes to 250mhz. Circuits over 100mhz get critical very fast. But I think you are describing the beat frequency (method?) used in some of the old radio receivers that had very low frequency IF stages - once again something I had not thought of. This is theory to me, never actually worked on rf. – jwzumwalt Jan 22 '18 at 0:30 • I'd certainly go with the quad photodiode, but I'd probably apply an astigmatic lens with its optical X-Y axes 45 degrees relative to the quad detector. Use auto adjust by sum-difference to acquire circle of least confusion for height. Dynamic range will be the problem, though. Needs thought. – jonk Jan 22 '18 at 0:30 • The dynamic range is why you are doing synchronous detection and why you have a narrow band optical filter over the photodiode to remove as much that does not match the laser wavelength as possible. You may find that log amps are still your friends. – Dan Mills Jan 22 '18 at 9:59 Radar altimeters and low-cost speed-sensing radars typically use FM-CW techniques. The actual signal processing occurs at audio rates. You might want to read up on the existing approaches before trying to invent your own. • I don't see anything in the "Radar altimeters" link that is pertinent to my question, did I miss something? – jwzumwalt Jan 22 '18 at 0:23 • @jwzumwalt, you stated that you are "interested in aircraft altitude measurement during approach" using RF technique. This is called "Radar Altimeter". The Wiki article (provided by Dave) describes several approaches in this area, and states: "Radar altimeters are frequently used by commercial aircraft for approach and landing". So it appears that you want to invent your own altimeter. Your comment implies that are trying to dismiss efforts of hundreds of professional engineers who worked in this area for many decades. – Ale..chenski Jan 22 '18 at 1:16 • Gee, silly me! You described a radar(radio) altimeter, so naturally, I assumed that that was what you wanted advice on. If that isn't what you're after, you're going to have to clarify your question considerably. – Dave Tweed Jan 22 '18 at 2:45 • Certified aircraft components such as this will require$50-100million dollars of testing and approval and be sold for $200,000 dollars. I am looking for a solution for under$100. I think it is safe to say the design goals for a quadracopter are different than a 747. – jwzumwalt Jan 22 '18 at 18:17
• I'm not talking about testing and certification, just the basic technology, which is simple, well-understood and reliable. These attributes make it easier to pass certification, but they are also good attributes for what you need. Why would you reject this information out-of-hand? – Dave Tweed Jan 22 '18 at 22:10