Mirrors/NawfalMotii79-PLFM_RADAR
1
0
Fork 0
mirror of https://github.com/NawfalMotii79/PLFM_RADAR.git synced 2026-06-08 14:44:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(mcu): PR-AB.a — move vector_0 out of inner beam_pos loop

Browse Source 
runRadarPulseSequence used to fire vector_0 (broadside reference)
between every matrix1 and matrix2 pattern, i.e. 15 times per azimuth.
That dwell × 8 ms × 15 = 120 ms per azimuth × 50 azimuths = ~6 s of
the 18.4 s revisit time was burned on redundant broadside frames.

Pull vector_0 out of the loop and fire it once per azimuth before the
sweep. Each azimuth now produces 1 broadside frame + 30 steered frames
(matrix1 + matrix2 across 15 beam_pos), down from 15 + 30 = 45 frames.
Revisit time drops from 18.4 s to ~12.8 s (31% improvement).

If multiple per-position broadside frames are ever needed, gate them
behind a runtime switch — the comment block flags this.

test_bug16_runradar_shadows_globals updated to mirror the new
1-outside + 2-inside m-counter pattern; 13/13 PASS, full MCU
regression 51/0 + 34/0.
This commit is contained in:
Jason
2026-05-06 12:18:00 +05:45
parent 98ec9cb6a5
commit b215caa294
2 changed files with 35 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files
9_Firmware/9_1_Microcontroller/9_1_3_C_Cpp_Code/main.cpp
+24 -17
View File
@@ -567,12 +567,15 @@ void runRadarPulseSequence() {
// BeamPos/Azimuth/ChirpCount. Local declarations would shadow them and
// freeze the telemetry at 1.
/* Main beam steering sequence. Per beam position we push three ADAR1000
* pattern updates: matrix1, vector_0 (broadside), matrix2 (see
* initializeBeamMatrices for the actual sim-peak directions and the
* sign-convention caveat). The FPGA's chirp_scheduler is firing the
* 3-PRI ladder against whichever pattern is currently loaded; we hold
* each pattern for one full SHORT/MEDIUM/LONG ladder before advancing.
/* Main beam steering sequence. Per azimuth slot we fire one broadside
* reference (vector_0) and then sweep 15 beam positions, each firing
* matrix1 (negative-θ scan) and matrix2 (positive-θ scan). PR-AB.a
* pulled vector_0 out of the inner loop — previously fired 15× per
* azimuth (between every matrix1/matrix2 pair) which dominated the
* 18.4 s revisit time without serving a documented per-pos purpose.
* One broadside frame per azimuth is sufficient as a clutter
* reference; if multiple per-pos broadside frames are ever needed
* they should re-enter behind a runtime switch.
*
* Per-pattern dwell:
* 16 × PRI_SHORT (175 us) +
@@ -581,26 +584,30 @@ void runRadarPulseSequence() {
* HAL_Delay yields to SysTick / IRQs, so unlike the removed
* executeChirpSequence busy-loop the MCU stays responsive to USB CDC,
* UART, and I2C peripherals during the dwell. For drift-immune sync
* we'd wait on an FPGA `subframe_pulse` GPIO instead, but the bitstream
* does not currently bring that signal out to a pin. */
* we'd wait on an FPGA `subframe_pulse` GPIO instead — DIG_7
* (H12→PD15) is wired in the schematic but currently driven by the
* F-6.4 watchdog OR; reassigning it is a tracked follow-up
* (PR-AB.b). */
static const uint32_t BEAM_PATTERN_DWELL_MS = 8u;
// One broadside (vector_0) reference frame per azimuth — replaces the
// 15 in-loop fires that PR-AB.a removed.
DIAG("SYS", "Broadside reference (vector_0) — 1× per azimuth");
adarManager.setCustomBeamPattern16(vector_0, ADAR1000Manager::BeamDirection::TX);
adarManager.setCustomBeamPattern16(vector_0, ADAR1000Manager::BeamDirection::RX);
HAL_Delay(BEAM_PATTERN_DWELL_MS);
m += m_max/2;
for(int beam_pos = 0; beam_pos < 15; beam_pos++) {
HAL_GPIO_TogglePin(GPIOD, GPIO_PIN_9);// new_elevation -- mode 2'b00 input only; no-op in production mode 2'b01 but harmless to keep
DIAG("SYS", "Beam pos %d/15: patterns matrix1/vector_0/matrix2", beam_pos);
// Pattern 1: matrix1
DIAG("SYS", "Beam pos %d/15: patterns matrix1/matrix2", beam_pos);
// Pattern 1: matrix1 (negative-θ scan, peak at -62°..-3°)
adarManager.setCustomBeamPattern16(matrix1[beam_pos], ADAR1000Manager::BeamDirection::TX);
adarManager.setCustomBeamPattern16(matrix1[beam_pos], ADAR1000Manager::BeamDirection::RX);
HAL_Delay(BEAM_PATTERN_DWELL_MS);
m += m_max/2;
// Pattern 2: vector_0 (broadside)
adarManager.setCustomBeamPattern16(vector_0, ADAR1000Manager::BeamDirection::TX);
adarManager.setCustomBeamPattern16(vector_0, ADAR1000Manager::BeamDirection::RX);
HAL_Delay(BEAM_PATTERN_DWELL_MS);
m += m_max/2;
// Pattern 3: matrix2
// Pattern 2: matrix2 (positive-θ scan, peak at +3°..+62°)
adarManager.setCustomBeamPattern16(matrix2[beam_pos], ADAR1000Manager::BeamDirection::TX);
adarManager.setCustomBeamPattern16(matrix2[beam_pos], ADAR1000Manager::BeamDirection::RX);
HAL_Delay(BEAM_PATTERN_DWELL_MS);
9_Firmware/9_1_Microcontroller/tests/test_bug16_runradar_shadows_globals.c
+11 -4
View File
@@ -59,13 +59,20 @@ static void run_buggy(void)
(void)m; (void)n; (void)y;
}
/* Post-fix: same body, no local redeclaration — references globals. */
/* Post-fix: same body, no local redeclaration — references globals.
* PR-AB.a moved vector_0 out of the inner loop, so the m advance is now
* 1 (vector_0, before loop) + 2 × 15 (matrix1+matrix2 in loop) = 31 increments
* per azimuth, instead of the prior 3 × 15 = 45. The wrap behavior
* (g_m wraps on every iteration's matrix2 add) is unchanged. */
static void run_fixed(void)
{
/* PR-AB.a: vector_0 broadside reference, 1× per azimuth (was inside loop). */
g_m += m_max / 2;
if (g_m > m_max) g_m = 1;
for (int beam_pos = 0; beam_pos < 15; beam_pos++) {
g_m += m_max / 2;
g_m += m_max / 2;
g_m += m_max / 2;
g_m += m_max / 2; /* matrix1 (negative-θ scan) */
g_m += m_max / 2; /* matrix2 (positive-θ scan) */
if (g_m > m_max) g_m = 1;
g_n++;