tell the world that TeleMetrum works!

diff --git a/bdale/blog/posts/TeleMetrum_Working.mdwn b/bdale/blog/posts/TeleMetrum_Working.mdwn
new file mode 100644 (file)
index 0000000..5d1643e
--- /dev/null
+++ b/bdale/blog/posts/TeleMetrum_Working.mdwn
@@ -0,0 +1,40 @@
+[[!tag tags/rockets]]
+Keith and I have been plugging away for the last couple weeks 
+turning on and testing the various 
+bits of [TeleMetrum](http://altusmetrum.org/TeleMetrum),
+and I can now very happily report that we have all the hardware 
+working!  Amazingly, we got almost everything right in the 
+hardware design.  Only two things need to be changed. 
+
+We need to be able to control the chip select line on the the SPI 
+non-volatile memory chip, which means we need one more GPIO line to it 
+from the processor.  To free up a line, we gave up the ability to put 
+the accelerometer in to self test mode.  Two trace cuts and two jumpers 
+to the current board implement this change.
+
+The last thing we got working just this evening was the ability to sense 
+the presence of an ejection charge igniter.  We need another resistor in 
+each of the two sense circuits to form a voltage divider so that we don't 
+overload the ADC input.  Not sure how we missed that in our many design 
+reviews!  Oh well.  To ensure robust ignition of the electric matches 
+we prefer to use for ejection charges, we charge a 1000uF cap to a voltage 
+higher than our ADC inputs can handle.  Fixing this problem on the current 
+boards requires changing the values of two resistors, then tacking extra 
+resistor to one end of each of the existing ones and then wiring the other 
+ends of the new resistors to ground.  Since the resistors in question are 
+all 0402 surface mount parts, this is a royal pain...  but with my finest 
+dental-pick soldering iron tip, and working under the inspection 
+microscope, it is indeed possible.
+
+Alternatively, if the ability to sense igniter continuity isn't important,
+just leaving off R14 and R15 at least prevents the 3.3V processor from going
+whacko due to having 5V applied to two ADC inputs!
+
+I'm really impressed at how well these first two boards reflow soldered in the
+electric skillet in my basement came out.  Getting the right amount of solder
+paste on the board is something I'm likely to get better at with practice,
+but so far we've only had a few (very obvious during inspection) shorts to
+clean up, and a couple poorly soldered pins.  All in all, I couldn't be
+happier with how this has gone... and I'm really looking forward to flying
+one of these boards sometime soon as an auxiliarly payload to gather some
+initial flight data!