Re: PVR x50 corrupts ATSC 115 streams

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


David Engel wrote:
On Tue, Feb 17, 2009 at 11:05:40AM -0500, Steven Toth wrote:
Does anyone know what might be going on?  These very same tuner cards
worked fine in the old system (Intel P4 3.0GHz CPU and Abit IC7
motherboard) for close to two years.
Determine whether this is an RF issue, or a DMA corruption issue:

Ahh, I didn't even think of RF.  I didn't have any RF problems in the
old system (that I know of) so that didn't even cross my mind.  I
actually was, and still am, more afraid of DMA issues, though.

1. Check the RF SNR of the digital cards using femon, anything odd going on when the PVR250 is running? Does it fall out of lock or SNR dip dangerously low, bursts of BER's?

Here are some 10 second captures from femon.  Card1 and Card2 are ATSC
115s.  Card4 and Card5 are PVR x50s.

Card1 recording by itself:

status SCVYL | signal fe90 | snr e4dc | ber 000000b8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e682 | ber 00000038 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe60 | snr e682 | ber 00000078 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe80 | snr e624 | ber 000000a0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e682 | ber 000000d8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe70 | snr e598 | ber 00000140 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe90 | snr e654 | ber 00000078 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e6b2 | ber 00000090 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe70 | snr e654 | ber 00000000 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe60 | snr e654 | ber 000000d8 | unc 00000000 | FE_HAS_LOCK

Card1 and Card4 recording:

status SCVYL | signal fe80 | snr e5c6 | ber 000000b8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe90 | snr e682 | ber 00000100 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e53a | ber 000000c8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e568 | ber 00000130 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e624 | ber 000000b8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe80 | snr e654 | ber 000000c0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe20 | snr e6b2 | ber 000000c0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e682 | ber 00000028 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe60 | snr e654 | ber 000000d8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe90 | snr e624 | ber 000000c8 | unc 00000000 | FE_HAS_LOCK

Card1 and Card5 recording:

status SCVYL | signal fe40 | snr e624 | ber 00000120 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe80 | snr e654 | ber 000000d0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e598 | ber 00000238 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe90 | snr e682 | ber 000000c8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal feb0 | snr e654 | ber 00000068 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe90 | snr e654 | ber 00000118 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal feb0 | snr e6e0 | ber 00000038 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe50 | snr e4dc | ber 00000060 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe50 | snr e44e | ber 00000058 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fea0 | snr e5c6 | ber 000000d0 | unc 00000000 | FE_HAS_LOCK

Card2 recording by itself:

status SCVYL | signal fe20 | snr e53a | ber 00000130 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe20 | snr e53a | ber 000000b0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe30 | snr e3c0 | ber 00000128 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe20 | snr e3c0 | ber 000000a8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe40 | snr e568 | ber 00000060 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe00 | snr e4dc | ber 00000058 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe10 | snr e53a | ber 00000098 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fde0 | snr e4dc | ber 00000118 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe10 | snr e568 | ber 00000168 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fdf0 | snr e53a | ber 000001a0 | unc 00000000 | FE_HAS_LOCK

Card2 and Card4 recording:

status SCVYL | signal fe40 | snr e4ac | ber 000000d8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe50 | snr e624 | ber 000001b8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe70 | snr e598 | ber 000000a8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe40 | snr e5c6 | ber 000000b0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe40 | snr e53a | ber 000000d8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe50 | snr e50a | ber 00000098 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe10 | snr e50a | ber 000000a0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe10 | snr e598 | ber 000000b0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe30 | snr e5f6 | ber 000000c0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe60 | snr e568 | ber 00000180 | unc 00000000 | FE_HAS_LOCK

Card2 and Card5 recording

status SCVYL | signal fe70 | snr e4ac | ber 000000b8 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe50 | snr e598 | ber 00000190 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe40 | snr e598 | ber 00000118 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe80 | snr e5c6 | ber 00000008 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe50 | snr e568 | ber 000000c0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe70 | snr e5c6 | ber 00000178 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe40 | snr e53a | ber 00000120 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe70 | snr e50a | ber 00000168 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe60 | snr e598 | ber 000000b0 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal fe40 | snr e5c6 | ber 000000f8 | unc 00000000 | FE_HAS_LOCK

I don't see anything significant there.

The BER looks buggy in this driver, assuming these readings are false.

Me neither, other than the dvb cards are reporting bit-errors. This feels like a driver bug not a generic card problem. If these ber's area real then this will account for your video issues, but that would be unrelated to the PVR250.

It'd personally remove the PVR250's and get your DVB statistics to report 0 for BER before continuing.


2. Move the two cards as far apart as possible in the slots in the system and repeat the test above, any better?

What happens?

This will require rmoving cards.  The old system had 5 PCI slots and I
had a small ethernet NIC between the pair of 115s and x50s.  The new
system only has 4 PCI slots so both pairs are in adjacent slots.  I'll
try pulling the x50s one at a time this evening when I get home.

Fold some thick cardboard and place it in between the PVR and the DVB boards, this will block some RF which may be coming from the encoder directly into your demod.

Does this help?

- Steve

--
video4linux-list mailing list
Unsubscribe mailto:video4linux-list-request@xxxxxxxxxx?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/video4linux-list

[Linux Media]     [Older V4L]     [Linux DVB]     [Video Disk Recorder]     [Linux Kernel]     [Asterisk]     [Photo]     [DCCP]     [Netdev]     [Xorg]     [Util Linux NG]     [Xfree86]     [Free Photo Albums]     [Fedora Users]     [Fedora Women]     [ALSA Users]     [ALSA Devel]     [SSH]     [DVB Maintainers]     [Linux USB]     [Yosemite Information]

Add to Google Powered by Linux