Differences

This shows you the differences between two versions of the page.

--- packet:mmdvm-tnc [2023/10/08 12:45] – [Building and flashing from source (hard(ish) mode)] m0lte
+++ packet:mmdvm-tnc [2024/04/14 11:26] (current) – [Testing notes] m0lte
@@ Line 14: / Line 14: @@
 ===== Preparation =====
-Burn Raspberry Pi OS Lite (tested on 32 bit) to a micro SD, boot it
+Tested on 32 bit Bullseye Raspi OS
-`sudo raspi-config` → interfaces → serial shell disabled, serial port enabled
+Verified to NOT work on 64 bit
+Verified to NOT work on Bookworm, only Bullseye, due to /sys/class/gpio/export deprecation
+Burn Raspberry Pi OS Bullseye 32 bit Lite to a micro SD, boot it
+''sudo raspi-config'' → interface options → serial port → serial shell disabled, serial port enabled
 Then:
@@ Line 26: / Line 32: @@
 # Fetch the flashing tool
-curl -OL https://www.bi7jta.org/files/MMDVM_Firmware/repeater_HAT/stm32flashV5
+curl -OLs https://www.bi7jta.org/files/MMDVM_Firmware/repeater_HAT/stm32flashV5
 chmod +x stm32flashV5
-</code>
-===== Flashing from the web =====
+sudo reboot
-Unplug the board from the header
-Run the following:
-<code>
-curl -O https://mmdvm-tnc.s3.eu-west-2.amazonaws.com/mmdvm-f4-simplex.hex
-while ! sudo ./stm32flashV5 -v -w mmdvm-f4-simplex.hex -R -i 20,-21,21:-20,-21,21 /dev/ttyAMA0; do echo Again; done
 </code>
-Carefully but swiftly (!) plug the board into the header
+===== Building and flashing from source =====
-This is required because of some capacitor in the wrong charge state and there may be some workaround coming.
-===== Building and flashing from source (hard(ish) mode) =====
 <code>
@@ Line 65: / Line 57: @@
 <code>
-curl -O https://www.bi7jta.org/files/MMDVM_Firmware/repeater_HAT/stm32flashV5
+mv ../stm32flashV5 .
-chmod +x stm32flashV5
 while ! sudo ./stm32flashV5 -v -w bin/mmdvm_f4.hex -R -i 20,-21,21:-20,-21,21 /dev/ttyAMA0; do echo Again; done
 </code>
+Carefully but swiftly (!) plug the board into the header. If it fails, unplug the board for a while and leave it before trying again later. A minute or two should be fine.
+This is required because of some capacitor in the wrong charge state and there may be some workaround coming.
 ===== Testing notes =====
@@ Line 96: / Line 92: @@
 <code>
 sudo python3 kiss-ax25-ui-batch.py /dev/ttyAMA0 115200 A0AAA B0BBB 100 "Hello world" 100 1
+</code>
+This will listen:
+<code>
+sudo python3 kiss-listen.py /dev/ttyAMA0 115200
 </code>
 sudo required for this unless your user is in the dialout group.
+This diff hacks kiss-ax25-ui-batch.py to:
+  - set persist=255
+  - divide supplied time interval by 10
+<code>
+diff --git a/kiss-ax25-ui-batch.py b/kiss-ax25-ui-batch.py
+index 11b3a75..2ebcbf8 100644
+--- a/kiss-ax25-ui-batch.py
++++ b/kiss-ax25-ui-batch.py
+@@ -120,6 +120,11 @@ TFESC = int(0xDD).to_bytes(1,'big')
+ TFEND = int(0xDC).to_bytes(1,'big')
+ KISS_PORT = 0
+ KISS_COMMAND = 0
++
++kiss_output_frame = bytearray(FEND) + bytearray(int(0x02).to_bytes(1,'big')) + bytearray(int(0xFF).to_bytes(1,'big')) + bytearray(FEND)
++#print(kiss_output_frame)
++port.write(kiss_output_frame)
++
+ KISS_TYPE_ID = (KISS_PORT * 16) + KISS_COMMAND
+ KISS_TYPE_ID = KISS_TYPE_ID.to_bytes(1,'big')
+@@ -191,6 +196,6 @@ for i in range(0, frame_count):
+        # print(kiss_output_frame)
+        frame_time = len(kiss_output_frame) * 10 / int(sys.argv[2])
+        port.write(kiss_output_frame)
+-       time.sleep(frame_interval)
++       time.sleep(frame_interval/10)
+ print('\nDone.')
+ GracefulExit(port, 0)
+</code>
 ===== Serial debugging =====