matthuszagh / fmcw-radar Goto Github PK

For every no lead component on the board see if there is a leaded equivalent to exchange it with. The performance benefits (which are minuscule) are not worth the increased difficulty with hand-soldering.

Time to get real, now that the board is working.
How far did you get on the multiplexing? I saw your Kicad file, did you:

• Make it?
• Write the code FPGA code to support it?
• Python/pc code for it?
  I'm interested in pushing a little forward that idea, so let me know!

I omitted the ground cutouts described in this post. I'll probably get a pretty large reflection because of it.

Consider using ECP5 FPGA instead of Artix 7

I remember you previously said you wanted to make your own antenna.
I found this article: https://antennatestlab.com/3dprinting
and made my own horn antennas for the fmcw3. They cost nothing and are fun to build.

Exchange the current switch (EVP-AWBD2A) with the one used in the icebreaker board (PTS810 SJM 250 SMTR LFS) which is a much better size and is generally nicer.

Should use a switch on the FPGA backside to allow the mode pins to be set without soldering/desoldering resistors. This is useful, for instance, if you don't want the FPGA to be automatically configured from SPI flash at startup.

All mode pins have internal pullup resistors, so the switch should toggle between an open and closed circuit to ground. See this DIP switch for a potential part.

Ensure that the mode pins (M2, M1, M0) are labelled on the silkscreen. It's probably also a good idea to label the interfaces that can be used (e.g. Master SPI is 001 and JTAG is 101).

The power amplifier is enabled with a low voltage. Perhaps this should
be inverted.

Change 93LC46BT-I/SN to 93LC46BT-I/OT and add 10k pull-up resistors to EECLK and EECS (pull up to 3.3D).

I'd like to use a USB type C connector in place of the current micro B USB 2.0 connection. The plugs/receptacles are symmetrical, support higher speeds (if the interface/FPGA will allow it) and are the future, so it would be nice to gain experience with them. The interface controller will need to be changed, since the FT2232H does not support USB standards past 2.0.

Ideally, I could use this to power the board. However, that would require at least 1 boost converter (an amplifier in the transmitter requires a 10V power supply) and might be a bit tight on current. USB C should allow up to 3A of current at 5V, which would be theoretically sufficient, but it doesn't leave a lot of leeway.

I think this is coming from the power supply (the upstream switching
converter switches at 500KHz).

The first step is to measure the power supply noise and determine if
that is indeed the source of the noise. If so, the contents below
apply. If not, I don't yet have an alternate theory.

The tone comes from the TPS5420D switching converter. The problem
is some combination of the following:

1. TPS5420D produces too much ripple. Could use the old L7980 instead.
2. TPS7A7001 linear regulator (downstream) has inadequate PSRR. In
   this case, it should be replaced with the old TPS7A91 (we'll need
   2, 1 for the mixer and another for the power amp, in order to meet
   max current draw).

Possible solutions (some combination of these):

1. Revert to L7980 and TPS7A91. Have mixer and power amp use separate
   TPS7A91 to accommodate current draw.
2. Use a capacitance multiplier (see the art of electronics section
   8.15.1) between the converter and regulator.

Hello,

Is this patch antenna the one that Henrik used or your own design based on his post?