aboutsummaryrefslogtreecommitdiffstats
path: root/README
diff options
context:
space:
mode:
Diffstat (limited to 'README')
-rw-r--r--README36
1 files changed, 36 insertions, 0 deletions
diff --git a/README b/README
new file mode 100644
index 0000000..ba383e9
--- /dev/null
+++ b/README
@@ -0,0 +1,36 @@
+To get this steaming pile of excellence working you need
+
+1) an STM32F103 board rated to 72MHz with an 8MHz clock and the PB6 and PB7 pins available
+
+eg
+http://www.ebay.co.uk/itm/132170295715
+http://www.ebay.co.uk/itm/132138287991
+
+2) optional (passive adaptors to get to you HDMI or DVI-D from whatever you have, the apple ones work well)
+
+3) Some way of connecting your HDMI/DVI-D to the board:
+
+HDMI DVI-D Cortex-M3
+
+15 DDC SCL 6 DDC SCL PB6 (with 10k pull-up to 3V3)
+16 DDC SDA 7 DDC SDA PB7 (with 10k pull-up to 3V3)
+17 GND 15 GND GND
+18 +5V 14 +5V (tie to HPD)
+19 HPD 16 HPD (tie to +5V)
+
+The code connects a virtual NS16X50 uart to the I2C bus at address 0x90 (like an SC16IS7XX does),
+and a 24C02 EEPROM containg DDC data for a 1280x1024 monitor at address 0xA0.
+
+The virtual UART is connected to the M3's real uart 1 which is enabled, and to a CDC ACM device
+on USB.
+
+The fake monitor is necessary because recent intel GPUs do hot plug detection and powering up the ports
+in firmware rather than in the driver it also tricks said firmware into configuring the port as HDMI/DVI
+rather than as Displayport - that way the GPIOs that AX can see get connected to SDA and SCL.
+
+The USB interface also offers a DFU device for flashing the firmware without a JTAG/ST-LINK cable,
+which although very slow is good because it synchronises the board reset with the USB host's rescaning
+of the device, so you don't need to replug.
+
+
+J.