LinHES Forums
http://forums.linhes.org/

LIRC delay and then inactivity problem
http://forums.linhes.org/viewtopic.php?f=2&t=18205
Page 1 of 1

Author:  pilesofspam [ Tue Apr 22, 2008 8:34 am ]
Post subject:  LIRC delay and then inactivity problem

I've got a LIRC blaster and receiver (homebrew) on the same serial port. On a cold boot, both work great. After about a day, sometimes 2, my remote still works with a few seconds of delay- like the commands are often missed, but sometimes queued for a second or two before they happen- like you're on the main screen and hit the down arrow to go to the next menu item, and it waits a full second (up to 4 seconds) before this happens. I could understand if the machine was busy doing something else, but its not. What's more, is that the keyboard still has instant response.

In another day, the remote doesn't work at all. If I restart the frontend (with ctrl-alt-backspace) it works fine again.

The irblaster never stops working. As far as I can tell, the receiver is working fine as well, but mythfrontend stops listening. Anybody got any ideas? I'd just bit the bullet and buy a streamzap, but I've got the homebrew receiver set up and wired through the walls. I don't know if the streamzap receiver could be extended to 40'.

Author:  tjc [ Tue Apr 22, 2008 5:05 pm ]
Post subject: 

Possibly "Know Issue #5" from the R5F27 Hints. The ir_kbd_i2c keyboard module is very badly behaved.

Author:  pilesofspam [ Wed Apr 23, 2008 1:03 pm ]
Post subject: 

Thanks for the response. I started tracking down this problem about a week ago and ended up deleting that module since I don't use the analog side of my HD5500 anyway- whether my hd5500 actually has an SAA7134 is up for discussion.

Results? No changes. LIRC still misbehaves after a day or so. I haven't reinstalled the module because it doesn't seem to be affecting anything else.

Has anybody had luck extending the streamzap receiver wire?

I've got a few other fixes I could possibly implement:

1. Automatically restart the frontend every day at like 4AM
2. Convince the front end to let go and re-grab lirc. I'm obviously talking conceptually here as I haven't done any research. Is this possible?
3. Get a streamzap, hack it longer.


I think I'm giving up on the root cause. If anybody has advice on the best way to implement the above, let me know!

Author:  pilesofspam [ Tue Aug 26, 2008 11:02 am ]
Post subject:  LIRC lockup workaround

I started this thread a long time ago, and I still have this issue under R5.5 . and I finally have a workaround as planned- I got a streamzap. I then took it apart, and desoldered the IR sensor in the PCB, and soldered in three conductor 50' wire- resoldering the IR sensor to the other side of that cable. It works great, and does not slow down or lock up. Note that the streamzap IR sensor component is not just a phototransistor, but also a demodulator (I think its filter is centered around 38kHz).

The IR blaster is still fired off of com1.

Author:  cliffsjunk [ Wed Aug 27, 2008 12:31 pm ]
Post subject:  Re: LIRC lockup workaround

pilesofspam wrote:
Note that the streamzap IR sensor component is not just a phototransistor


I too fiddle with electronic hardware design. I built a home brew LIRC
receiver starting with a photodiode that worked ok, but finally realized
that I really needed a receiver module for the AGC and for the similar
problems that the 38khz carrier gets around.

I had problems with a curtain getting warm when the sun shone on it in
the summer. This swamped out the signal.

Page 1 of 1 All times are UTC - 6 hours
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/