infineon4engineers

Thread: Distance2Go Doppler and FMCW mode (range and speed)

1. Distance2Go Doppler and FMCW mode (range and speed)

Hey there,

I can't figure out how the Distance2Go processes I and Q data to get values for both range and speed within a single (or multiple?) measurement. What I know from theory so far is that you usually need a sequence of transmitted chirps (at least 2) to be able to resolve for both range R and speed v (or Doppler frequency fD).
Theory says this is done via a range-doppler-matrix and a two-dimensional FFT. But is that the case with the D2G board as well? I'm reading through the source code in DAVE but I did not figure that out yet. I stumbled across some 'mean' calculations, what to they do?
How can the board determine range and speed I if set the number of chirps to '1'?
In algo_api.h it is said: "Currently a single FMCW and a Doppler chirp is generated within a single frame." What does that mean?

Another question is the correlation between number of chirps, samples per chirp, data size and frame time.
As far as I understand it:
Number of chirps is the number of chirps that is generated within one frame.
Samples per chirp is equivalent to the number of samples per chirp, so if I'd have 2 chirps I need to store or evaluate 2 times the data.
Data size is the hardware buffer size (max 250)? So if I have 2 chirps and 250 samples per chirp, it evaluates the 250 samples from the first and after that the next 250 samples from the second chirp?
Frame time or minimum frame time is the time required to generate upchirp and downramp (limited by various hardware), e.g. ramp down time (PLL) and the time it takes to perform the calculations (algorithm process time etc). So I should get errors when setting the frame time lower that the minimum frame time.

I'd like to understand the system more, especially I'd like to learn about the resolution process of range and speed. Are there any sources I can get more information? I appreciate any help!

2. Hi,

1. Distance2Go version 1.2 does not have Range Doppler algorithm, so it is not possible to get range and speed for all targets. What it is offering in Radar_D2G_Doppler_FMCW project, is following.
It generates CW signal as first chirp and sample it at 20kHz sampling frequency (like a pure doppler signal), afterwards it generates FMCW chirp and sample this data in to second buffer with FMCW sampling frequency computed from chirptime and number of samples per chirp. Now, from first doppler data set, it computes speed of maximum doppler target while second FMCW chirp is used to compute ranges of all targets. Now ambiguity comes in case there are multiple targets, then one cannot assign that single speed computed from doppler data set to any target. This will not be case in near future as version 2.0 is planned to be released soon (hopefully before Christmas) and in this version multiple chirp support along with Range doppler algorithm with limited configurations will be provided.
2. In one frame of say 250msec, one can generate 8 chirps, with samples per chirp of e.g. 64. DATA_SIZE is samples per chirp if I remember correctly.
3. Minimum frame interval depends upon number of chirps, chirptime (up, down and wait) time, algorithm time, BGT and PLL power up time if duty cycle is enabled.

There is no document present at the moment for all these details. But for your info, a new board Position2Go is going to be released this year, which is BGTMTR12 and can give angle, range and speed as well. Stay tuned for this...

3. Hey,

1. During my measurements in the meantime I noticed that the calculated Doppler speed cannot be assigned to any of the possible 16 targets (target IDs) that come frome the range measurement. Looking forward to v2.0, I also heard about a new firmware release for November?
2. In the config.h of the DAVE project there are seperate variables for DATA_SIZE and samples per chirp so they are not identical but are somehow dependent on each other as far as I know.
3. I tried several setups in MATLAB (changing all the parameters to see their effects) and there is one fact that bothers me. I know there is a relation between minimum frame time, samples per chirp and chirps per frame (number of chirps). However, it matters which parameter I change first. For example setting samples per chirp first and chirps per frame afterwards has a different effect than changing the variables vice versa. Isn't it true that the order should be irrelevant?

Sorry for my late response, the last days were quite busy...

I got some more questions:
1) Is it always 1 doppler chirp that is sent per frame within the frame period? So the variable number of chirps only affects the number of FMCW chirps, but not the doppler chirps?
2) Is version 2.0 a software or hardware release? If it is a new hardware version then my guess is that the current hardware is not capable of doing range-doppler-algorithm?