# RF Distance Measurement Techniques (ToF+Interferometry)

1. Dec 1, 2011

### ¡MR.AWESOME!

I am trying to design a system to measure the distance between two RF transceivers with as small of a resolution as possible. Time-of-flight measurement is nice and easy to do, but it has a poor resolution with low frequency timers. I am thinking of using a low-cost mcu that will probably have a clock rate of ~30Mhz. So the best possible resolution I could get would be (1/3x10^7)*c=~10m/s. I would like to get much better resolution than that.

I did some searching around and found some info on RF interferometry. Basically, you create a wave and split it in two. The first wave goes out and does it's thing and then comes back. The second one stays where it is. You then compare the phases of the two. Depending on the wavelength, you can determine the distance that the first wave traveled by the amount the phase shifted compared to the second wave.

Obviously, this will only be accurate if the distance traveled is less than one wavelength. By combining time-of-flight measurements with phase difference measurements, could you not get a much higher resolution? It seems to me that distance measurement with interferometry is almost exclusively done with lasers or other forms of light, which is only good for very small distance measurements. Why is this? I must be missing something.

2. Dec 1, 2011

3. Dec 1, 2011

### yungman

Use a ramp generator for each 30 MHz period and digitize the ramp that is proportion to the time between the two clocks. This will give you a single stop TDC quite easily with much much better resolution. If you design it good, and even if you use an 8 bit ADC, you get 256 more resolution from the 30MHz. This is the old style TDC that is sold on the market in the 90s.

I came up with the idea and designed a multi-stop TDC( Time Digital Converter) in the mid 90s using two quad sine wave and I got resolution of 70pS successfully into our TOF ( time of flight) SIMS. It was the first ever multi stop TDC at the time with this kind of resolution in 1995.

I use two sine wave 90 deg apart none stop so at any given time, one will be at the most linear part of the slope. It is a big project with a lot of software to linearize the sine wave and to choose which one to use. But no matter how you cut it, it is going to be a big project for you if you need multi-stop.

We blew CAMECA and other competitors out of the water at the time.

Last edited: Dec 1, 2011
4. Dec 2, 2011

### ¡MR.AWESOME!

Ah. Thanks for the reply. You definitely set me off on the right track. I wasn't finding anything useful before.

I don't quite understand the dual sine, quad phase method you described. Is there any literature that you could point me towards to clarify?

One shot is good enough for me, but I would still like to understand your method.

Thanks

5. Dec 2, 2011

### yungman

I don't know of any books, I came up with that idea myself, I did not pursue patent nor wrote a paper at the time as I had something bigger cooking!!! The idea is to have two pure sine wave 90 degrees apart. so at any given point of time, you have one slope at the most linear region ( rise and fall of the sine wave is quite linear.) I use mapping EPROM to compensate the little bend on the slope of the sine wave. EPROM program also determine which of the 4 slope to use for digitizing..........

Well even thought this is 16 years old and I am not working for the company anymore, I don't feel comfortable getting into more detail even though it's my original idea. I think you have a good idea already. It is not an easy project. That was the first try only. It worked and we moved on. Hind sight I would do it differently.

If you are doing single stop, that would be very easy. Like I said before, use your 30MHz clock and start a ramp each time and when the in coming trigger come in, you digitize the value of the ramp. You get easily 256 steps within each clock period. So you just combine the digitized value and become the LSBite when combine with the course counter run by 30MHz. Make sure you start the ramp before the edge of the clock as it takes a little time for the ramp to settle to a linear slope. That's was why I use the continuous sine wave so there was not settling time.