[mythtv-users] IR Blaster - pin 7 instead of pin 4?

[Jay Littlefield]

Hello all,

I have a very strange problem affecting my MythTV system.  I've searched 
the archives for similar items and have not had much luck.  As this is 
becoming a case of spousal embarrassment, I'm hoping someone out there 
can assist me.  :-)

I've built a MythTV system based upon Jarod's Fedora guide. 
(http://wilsonet.com/mythtv/fcmyth.php)  I'm using Fedora 5 and MythTV 
0.20.   I have an ASUS M2NVP-VM motherboard AMD64 machine.  (HDTV 
outputs and Composite TV out included.)  I have a PVR-150 card that is 
working fine, including the IR Receiver.  (The "newer grey" remote in 
Jarod's description.)  The system has been working absolutely fine with 
local broadcast, but once I got DirecTV hooked up in my new apartment, I 
have had problems trying to get the channel changing capability 
working.  The box is a DirecTV D11-500.  This is the last item I need to 
get the system so that I can do timed recordings, and I have struggled 
with it for literally some weeks now.

I stumbled across Andy's site at:  
http://www.eggshellskull.com/lirc/blaster/index.php.  I was hopeful this 
would solve the problem, so I walked through it step by step.  I bought 
an IR blaster from http://www.irblaster.info/ and followed the 
directions on the site, downloding the RPMs from atrpms.net.  
Installation went smoothly, but the blaster does not respond to the 
channel changing script.   I verified the IR blaster works by using the 
digital camera trick while driving it from a 9V battery.  (You have to 
hold it REALLY close to the transmitter, but I could see it light up.)
After several days of frustration, I went onto the motherboard with a 
voltmeter and measured the pins.  I found that instead of using pin 4 
(DTR) as the transmit pin, it is actually pin 7 (RTS) that is toggling.  
When measured against pin 5 (SG), it displays a negative voltage.  
(Approaximately -11V).  I put Andy's  change_chan.sh script in a loop to 
keep the channels changing and measured the pin voltages.  Pin 7's 
voltage starts fluctuating.  On my digital meter, it looks like it jumps 
between about -9.5V and -11V.   As soon as I stop, it returns to a 
steady -11V.  I'm quite sure this is the pin that is being controlled.   
I am attempting to drive the pin using the DirecTV codes from 
http://lirc.sourceforge.net/remotes/directv/G051204 in my LIRC file in 
place of Andy's Motorola box codes.  I'm fairly certain the 
configuration is correct.  It almost appears as if the lirc_serial 
control is driving the incorrect pin, but I have no idea why that would be.

Every piece of documentation I have seen on ir blasters use pin 4 (DTR) 
for the transmitter pin.  I have downloaded the appropriate 64-bit RPMs 
for all 5 LIRC components and have verified the correct kernel version 
(2.6.18-1.2200).  I'm wondering if there's some basic issue I am missing 
with regards to this.  The lirc_serial logs messages that signals are 
being received.  It just seems like they are being routed to a different 
location.

As an alternative, I commented out the mods to rc.local and 
modprobe.conf found on Andy's page and tried to implement the 
configuration found on this page:  
http://www.mythtv.org/wiki/index.php/Controlling_DirectTV_D11_via_USB    
I had all the necessary equipment to try this, but I am running into 
constant timeouts from the serial port.  After multiple retries, it 
simply fails.  I have tried various combinations of baudrates, key 
combinations, etc.  After several weeks of failed efforts, I have begun 
to simply blame the serial port for all failures.  :-)

I'm searched the archives here and can't seem to find anything similar 
to this.  Does anyone have any suggestions as to why this may be 
happening and what suggestions I can try to fix it?  I have the Kernel 
source code RPMs downloaded, but I'd really love to steer clear of 
recompiling the drivers if it's at all possible.

Thanks for your help!

      -- Jay