|
From: Jay J. <cu...@ch...> - 2009-01-11 03:12:54
|
I am setting up a MythTV box for a friend, and he needs an "IR blaster" to control an HD to SD converter box. I am using Fedora 10, and the "provided" lirc RPMs from Fedora. I figured out how to get lirc to simultaneously support coexistance with his Hauppauge 350 (lirc_i2c) and serial (lirc_serial), with two lircd processes. The receivers on both work fine, and the Hauppauge receiver works fine with MythTV. However, transmitting is not working. Even when I remove lirc_i2c and just run the serial lirc, the transmitter is unable to get the controlled device to respond. So, after trying "this and that" to no avail, I took my two transmitters (one a home built using the "improved" two-transistor circuit, off of the site, and one purchased over to my Windows box) and installed each, one at a time in turn, and tried them with WinLIRC, and both work fine from Windows (so, the transmitter hardware is fine). But neither transmitter works using the lirc I have installed under Fedora 10 . On the two-transistor circuit, I have a visible LED in series with the IR LED (series so that I was sure the IR LED wasn't connected backwards), and the visible LED does flash when I use irsend on the Linux box. However, none of the devices I try to control respond to the Linux box -- and both transmitters control both devices fine from WinLIRC. In both cases I used identical configuration files for WinLIRC and Linux lirc, and I have tried both "raw" and standard config files. (Currently using the "raw" file). So, the config file is fine, too. So, I see only two possibilities: Either DTR on the COM port on the Linux box is busted (seems really unlikely, especially given the visible indication from the other LED), or a software problem (perhaps inverted pulses from Linux LIRC irsend or something I need to provide to modprobe lirc_serial?, wrong frequency, or some such). FYI, I also noted that I was unable to get the transmitter to do anything at all on COM2 on the Linux box, even when providing serial port parameters to modprobe. Perhaps it wasn't compiled with that support? Is there any way to tell how it was compiled? I also tried unloading lirc_i2c and lirc_dev and doing the modprobe for lirc_serial, and adjusting lircd appropriately, to no avail -- wanted to make sure lirc_i2c wasn't interfering. I searched the mailing list archives, and didn't see anything that seemed to apply. Suggestions? Right now my prime suspect is lirc_serial . Here is the info from my most recent test (where I eliminated lirc_i2c): [root@mythsoby1 ~]# uname -a Linux mythsoby1 2.6.27.9-159.fc10.i686 #1 SMP Tue Dec 16 15:12:04 EST 2008 i686 i686 i386 GNU/Linux [root@mythsoby1 ~]# rpm -q -a | grep lirc lirc-0.8.4a-1.fc10.i386 lirc-devel-0.8.4a-1.fc10.i386 lirc-remotes-0.8.4a-1.fc10.i386 lirc-doc-0.8.4a-1.fc10.i386 lirc-libs-0.8.4a-1.fc10.i386 [root@mythsoby1 ~]# lsmod | grep lirc lirc_serial 15072 0 lirc_dev 14232 1 lirc_serial # dmesg lirc_dev: IR Remote Control driver registered, major 61 lirc_serial: auto-detected active low receiver lirc_dev: lirc_register_plugin: sample_rate: 0 [root@mythsoby1 ~]# ls -ld /dev/lirc* lrwxrwxrwx 1 root root 5 2009-01-10 20:31 /dev/lirc -> lirc0 crw-rw-rw- 1 mythtv root 61, 0 2009-01-10 20:31 /dev/lirc0 srw-rw-rw- 1 root root 0 2009-01-10 20:34 /dev/lircd1 (Note: when I started lircd, I specifically indicated output=/dev/lircd1. I realize this isn't "typical"). lircd --driver=default --device=/dev/lirc0 --output=/dev/lircd1 --pidfile=/var/run/lircd1.pid # irsend -d /dev/lircd1 SEND_ONCE Tivax.raw POWER POWER POWER # tail /var/log/messages Jan 10 20:58:42 localhost lircd-0.8.4a[8781]: accepted new client on /dev/lircd1 Jan 10 20:58:43 localhost lircd-0.8.4a[8781]: removed client Thanks in advance for any help. (Otherwise, I suppose the next step will be to hook a 'scope up and see what the signals coming out of the serial ports actually look like -- but that is a bit of work...) JRJ --- Jay R. Jaeger The Computer Collection cu...@ch... |