My R5A30.2 setup has both a PVR-250 tuner and a DCT-6200 tuner connected via firewire (TI chipset). I am seeing two occasion problems:

1. 0-length files from the DCT-6200/FW tuner with the corresponding warning in mythbackend.log (15 seconds without video...). This happens very infreqently.

2. Incorrect channel change. Occasionally the dct-6200 will tune to the wrong channel. When I am in the room to see it happen, it occurs with the second of 2 back-to-back recordings on that tuner (west wing followed by desperate housewives, in this case). The LCD display of the dct-6200 shows the digits flash by, as if it's going to change channels, but then it glitches and reverts back to the channel it was already on instead of changing to the new channel.

Do others see this happen ocasionally? If I move to a firewire card with an NEC chipset will these problems go away completely? Finally, I've seen people mention that there is a published list of "good" firewire chipsets -- can someone point me to that list? I've looked for it but I haven't been able to find it. Thank you!

Marc

My firewire has always been very reliable. I did start having problems a couple of weeks ago, but turns out the cable company had moved some of the channels around and mythtv didn't know about it. I'm not sure what chipset I'm using. It's the onboard firewire from my Abit AS8. I think it's an intel chip, but not sure. I've never seen a definitive list of good firewire cards/chipsets,of course I may have just missed it since mine was working I haven't really looked.

_________________
Have a question search the forum and have a look at the KnoppMythWiki.

Xsecrets

X, thanks for replying -- appreciate the help. On my system "/proc/pci" contains the following (amongst other stuff)

Would you mind publishing what yours shows? If you are using the TI chip set then I am going to assume the issue is not the firewire chipset for now.

Ealier my firewire card and my on-the-motherboard sound were sharing IRQ 7. I disabled the on-board sound and added a sound card I had hanging around and now the FW card is on it's own IRQ. I need to run for a while to see if this helps since the problem is intermittent. Also, my motherboard uses the SIS chipset, so I've had to do quite a few things to get to the point where it is stable (boot with "noapic" and "noagp"; can't use "-u1" on hdparm; using the nvidia AGP driver instead of agpgart.). It wouldn't surprise me if, in the end, it has something to do with the SIS chipset....

Thanks again!

Marc

from lspci

0000:02:01.0 FireWire (IEEE 1394): Texas Instruments TSB43AB23 IEEE-1394a-2000 Controller (PHY/Link)

sorry my /proc/pci only has bus info not any info on what the devices are.

_________________
Have a question search the forum and have a look at the KnoppMythWiki.

Xsecrets

X, thanks for posting that info. It looks like your TI chipset may be a different version than mine. I'm going to see how my system does over the next week now that the firewire card no longer shares an IRQ with the sound card and see how it goes. If there are still problems I'll try substituting a new firewire card to see if that makes a difference.

My other observation is that the "misfires" with channel changing occured under fairly heavy load -- while recording 1 HD and 1 SD show while simultaneously watching the HD show that was being recorded. I'll also try some load testing when I get some time next weekend.


Marc

well I can wath hd record two hd and one sd with no problems. The only thing that bogs my system down is commflag so I have turned that off.

_________________
Have a question search the forum and have a look at the KnoppMythWiki.

Xsecrets

X,

I suspect it's the SIS chipset. I say this because I have read about varous problems with the SIS chipset and I had to make a lot of tweaks to get rid of various errors like "APIC error on CPU0"; "drive_cmd errors on boot with hdd"; "entire system hangs after 12-24 hours of uptime"; etc. For each problem I eventually found a posting somewhere that described ithe problem I was having and a good work around, but it was typically referenced as being caused by chipset issues.

I built my machine from an ECS motherboard and 2.8 P4 I had "hanging around" . With hyper-threading enabled I find CPU utilizaton stays quite reasonable -- always seem to hvae at least 25% available and typically 40% available, but it probably makes sense to purchase new hardware for this setup at some point.

Thanks agian for your help!

Marc

well if you are using commercial flag it's not the cpu that maxes out it's the I/O I can have plenty of cpu left on my PIV 3.0G HT but video skips and stutters like crazy when commflag is running particularly on an hd show.

_________________
Have a question search the forum and have a look at the KnoppMythWiki.

Xsecrets

X, I don't use comm flagging or transcoding. The only "batch" job I can think of that runs (other than mythfilldatabase) is mythlink -- runs at 5 and 35 past the hour.

Marc