[pvrusb2] problems with pvrusb

[Mike Isely]

On Thu, 21 Oct 2010, Jim Peters wrote:

> 
> I have 4 of these tuners, 3 of them are the 24xx series and 1 is a 29xx 
> series, at one time they all worked great and I am not sure what happened to 
> cause this but at some point the 29xx tuner started outputting black and white 
> video. At first I suspected a hardware failure and thought maybe it was in the 
> analog tuner part of the card so I switched it over to using the composite 
> side but this didn't help. As a last resort to chalking it up as a failed 
> piece of hardware I tried it on a windows machine and it works fine there so I 
> am under the impression that this must be a driver/firmware issue. Since my 
> other 3 tuners don't have a problem I assume that there must be some part of 
> the driver/firmware that only affects the 29xx series cards and the problem lies 
> within this, I have no idea where to look for a solution, any help would be 
> appreciated.
> 

This is a very strange issue.  There's really nothing in the driver that 
can cause the video to go B&W.  The only thing that I can come up with 
is that the 29xxx device is being configured with the wrong video 
standard.  Most analog video standards use pretty much the same scheme 
for encoding luminance and sync info - the differences are just 
parmaeters: the number of scan lines, interlacing, and vertical refresh 
rate.  But when color TV first became common so long ago, the standard 
got "forked" and different countries adopted different means for 
encoding a color subcarrier.  That's really the big difference between 
NTSC vs PAL vs SECAM.  The outward effect of this is that if you have 
the wrong video standard set, one likely result is that you'll still be 
"close enough" that the hardware can manage to get a lock on the sync 
pulses and the luminance but will fail to identify the chrominance / hue 
components of the signal.  Result: Black and white video.

Now with that said, everything you do to configure the 24xxx devices 
should be the same as for the 29xxx device.  They do have different 
hardware so I suppose another possibility is that they're all configured 
wrong but the 24xxx devices have chips that are successfully 
autodetecting the video format in spite of what they were told.  Just 
guessing, but those are things I'd look for.

It's also possible that a bug has crept into the saa7115 driver in 
v4l-dvb which has gone unnoticed - that driver is unique for the 29xxx 
devices.  It's not like that hasn't happened before.  To test for that, 
I'd try to boot an older kernel, preferably one that you knew worked 
before.

As for firmware issues, that's pretty unlikely here.  The 29xxx and 
24xxx devices can share the same mpeg encoder firmware so if you got 
that wrong they'd all be busted.  The 29xxx and 24xxx devices use 
completely different FX2 firmware (the file names encode their type so 
there's no collision on systems with both device types).  But the FX2 
firmware is basic to controlling the entire device and if you corrupted 
the 29xxx version then it just isn't going to work at all.  There's 
actually nothing in the FX2 firmware that can interfere with color vs 
B&W video since the video standard info is sent directly to the 
individual chips over I2C and since the video data itself never actually 
gets into the FX2 processor (it's a dog-slow 8051 - no way could it keep 
up with that sort of data stream).


> 
> Operating system 	Mythbuntu Linux 10.04.1 Automatic Daily builds 
> 
> mythtv version .23
> 
> Kernel and CPU 	Linux 2.6.32-25-generic on x86_64
> 
> Processor information 	Dual Core AMD Opteron(tm) Processor 270, 4 cores
> 
> Real memory 	3.75 GB total, 504.48 MB used

A long time ago I acquired a 29xxx device that had a problem where it 
would only produce B&W video.  No matter what I did with it, the damn 
thing would never produce color.  I spent 6 WEEKS trying to figure out 
what kind of pvrusb2 driver bug could cause this and found nothing at 
all.  Then I tried it on a Windows system - and got B&W video.  And 
that's when I RMA'ed the device back to Hauppauge.  However you said you 
tried in Windows successfully so that pretty much proves that the 
hardware is fine.  I'd look hard at the video standard to which it has 
been configured.

  -Mike


-- 

Mike Isely
isely @ isely (dot) net
PGP: 03 54 43 4D 75 E5 CC 92 71 16 01 E2 B5 F5 C1 E8