Hauppauge 2250 - second tuner is only half working

Hauppauge 2250 - second tuner is only half working

[seth]

Hello,

     I've run into a problem with my capture card and wanted to solicit
ideas on how to further pinpoint the problem and/or hopefully correct
it.  At the risk of being to presumptious/terse - my problem is the
second tuner on the card is unable to lock on any frequency between
609mhz and 693mhz (while the first tuner does just fine).

     I recently purchased a bunch of hardware and assembled a HTPC.  I'm
running a fairly vanilla mythbuntu on it.  I have only the one
capture card installed, and i'm using comcast cable for input.  Other
than a handful of various apt packages i've added, i've compiled from
source saa7164-dev branch (was on the -stable at one point),
lcdproc-0.5.2 w/ imonlcd-0.3.patch, lirc (trunk), scte65scan, and a
couple other tools.  Unfortunately for me, I ran across the problem
right after I did a big db merge on my mythtv's channel &
dtv_multiplex tables (I was joining schedules direct, scte65scan
results, and dvb-app scan results).  Naturally, i started
troubleshooting from that point under the assumption i screwed up the
tables.  I eventually worked my way down to where i'm stuck at now -
the second tuner mysteriously is unable to lock into some
frequencies.

     I ran full scans (us-Cable-Standard-center-frequencies-QAM256) on
both adapters - then sort -t : -k 2n -k 6n the output on both and
diffed and ended up with the 609 and 693mhz numbers from above - all
frequencies in that range are not tuneable for adapter1.  It
represents 109 tunable channels on adapter0 that just don't seem to
work on adapter1 (there are about 250 that do work on both adapters
properly).  I spot checked at least a dozen channels in both the
problem frequency range and in the good range (i used mythtv live tv
and azap w/ dvbtraffic).  dvbtraffic showed hundreds of lines of
small bandwidth channels on the problem frequency range on adapter1

(For brevity sake, i simplified my scan list down to two channels to make
it easier to test)

channels.conf contains:
KOMO:555000000:QAM_256:1984:1985:3
SYFY:669000000:QAM_256:2112:2113:7

seth@dh101:~/development/dvb-apps/dvb-apps_hg_co/util/szap$ ./azap -a 1 -c
channels.conf komo
using '/dev/dvb/adapter1/frontend0' and '/dev/dvb/adapter1/demux0'
tuning to 555000000 Hz
video pid 0x07c0, audio pid 0x07c1
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 1f | signal 0186 | snr 0186 | ber 00000000 | unc 00000000 |
FE_HAS_LOCK
status 1f | signal 0186 | snr 0186 | ber 00000000 | unc 00000000 |
FE_HAS_LOCK
^C
seth@dh101:~/development/dvb-apps/dvb-apps_hg_co/util/szap$ ./azap -a 0 -c
channels.conf komo
using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0'
tuning to 555000000 Hz
video pid 0x07c0, audio pid 0x07c1
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 1f | signal 0190 | snr 0190 | ber 00000000 | unc 00000000 |
FE_HAS_LOCK
status 1f | signal 0190 | snr 0190 | ber 00000000 | unc 00000000 |
FE_HAS_LOCK
^C
seth@dh101:~/development/dvb-apps/dvb-apps_hg_co/util/szap$ ./azap -a 1 -c
channels.conf syfy
using '/dev/dvb/adapter1/frontend0' and '/dev/dvb/adapter1/demux0'
tuning to 669000000 Hz
video pid 0x0840, audio pid 0x0841
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 | 	<<<<
I've let it sit for 5+ min without lock
^C
seth@dh101:~/development/dvb-apps/dvb-apps_hg_co/util/szap$ ./azap -a 0 -c
channels.conf syfy
using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0'
tuning to 669000000 Hz
video pid 0x0840, audio pid 0x0841
status 00 | signal 0000 | snr 0000 | ber 00000000 | unc 00000000 |
status 1f | signal 0190 | snr 0190 | ber 00000147 | unc 00000147 |
FE_HAS_LOCK
status 1f | signal 0190 | snr 0190 | ber 00000000 | unc 00000000 |
FE_HAS_LOCK

dmesg output:
[74217.307056] saa7164 driver loaded
[74217.307106] saa7164 0000:03:00.0: PCI INT A -> Link[AE3A] -> GSI 16
(level, low) -> IRQ 16
[74217.307944] CORE saa7164[0]: subsystem: 0070:8891, board: Hauppauge
WinTV-HVR2250 [card=7,autodetected]
[74217.307950] saa7164[0]/0: found at 0000:03:00.0, rev: 129, irq: 16,
latency: 0, mmio: 0xe5000000
[74217.307957] saa7164 0000:03:00.0: setting latency timer to 64
[74217.308710] saa7164[0]: i2c bus 0 registered
[74217.308886] saa7164[0]: i2c bus 1 registered
[74217.309069] saa7164[0]: i2c bus 2 registered
[74217.343459] tveeprom 0-0000: Hauppauge model 88061, rev C3F2, serial#
6254250
[74217.343462] tveeprom 0-0000: MAC address is 00-0D-FE-5F-6E-AA
[74217.343464] tveeprom 0-0000: tuner model is NXP 18271C2_716x (idx 152,
type 4)
[74217.343466] tveeprom 0-0000: TV standards NTSC(M) ATSC/DVB Digital
(eeprom 0x88)
[74217.343468] tveeprom 0-0000: audio processor is SAA7164 (idx 43)
[74217.343470] tveeprom 0-0000: decoder processor is SAA7164 (idx 40)
[74217.343473] tveeprom 0-0000: has radio, has IR receiver, has no IR
transmitter
[74217.343475] saa7164[0]: Hauppauge eeprom: model=88061
[74217.645028] tda18271 1-0060: creating new instance
[74217.649054] TDA18271HD/C2 detected @ 1-0060
[74217.900894] DVB: registering new adapter (saa7164)
[74217.900898] DVB: registering adapter 0 frontend 0 (Samsung S5H1411
QAM/8VSB Frontend)...
[74218.197734] tda18271 2-0060: creating new instance
[74218.201853] TDA18271HD/C2 detected @ 2-0060
[74218.456641] DVB: registering new adapter (saa7164)
[74218.456645] DVB: registering adapter 1 frontend 0 (Samsung S5H1411
QAM/8VSB Frontend)...


I'd really appreciate any help or guidance on this problem as i'm fully
perplexed by it.

-Seth

Re: Hauppauge 2250 - second tuner is only half working

[Steven Toth]

> I'd really appreciate any help or guidance on this problem as i'm fully
> perplexed by it.

Hey Seth,

I ran the same tests on my cable system (channel 103) on 669Mhz and had no 
issue, and my snr's reported as (0x172 and 0x17c).

One possibility is that you're overwhelming the frontend. Try adding a small 
mount of attenuation to the signal for test purposes.

Hard to believe but this is where I'd start looking.

-- 
Steven Toth - Kernel Labs
http://www.kernellabs.com

Re: Hauppauge 2250 - second tuner is only half working

[seth]

>> I'd really appreciate any help or guidance on this problem as i'm fully
>> perplexed by it.
>
> Hey Seth,
>
> I ran the same tests on my cable system (channel 103) on 669Mhz and had no
> issue, and my snr's reported as (0x172 and 0x17c).
>
> One possibility is that you're overwhelming the frontend. Try adding a
> small
> mount of attenuation to the signal for test purposes.
>
> Hard to believe but this is where I'd start looking.
>
> --
> Steven Toth - Kernel Labs
> http://www.kernellabs.com
>
>

Thank you for reply!  Hearing that the same frequency works on another
card is pretty positive confirmation in my mind that this is a
hardware/setup issue.  I tried stopping by a local radio shack last night,
but wouldn't you know they no longer carried simple attenuators.  Looks
like i'll be picking one up online (or maybe ill lookup a schematic online
and try building a simple one).

On a side note - Thank you very much for hacking on the saa7164 - other
than this frequency glitch its been working great for me!

Re: Hauppauge 2250 - second tuner is only half working

[Andy Walls]

On Wed, 2009-08-19 at 07:12 -0700, seth@cyberseth.com wrote:
> >> I'd really appreciate any help or guidance on this problem as i'm fully
> >> perplexed by it.
> >
> > Hey Seth,
> >
> > I ran the same tests on my cable system (channel 103) on 669Mhz and had no
> > issue, and my snr's reported as (0x172 and 0x17c).
> >
> > One possibility is that you're overwhelming the frontend. Try adding a
> > small
> > mount of attenuation to the signal for test purposes.
> >
> > Hard to believe but this is where I'd start looking.
> >
> > --
> > Steven Toth - Kernel Labs
> > http://www.kernellabs.com
> >
> >
> 
> Thank you for reply!  Hearing that the same frequency works on another
> card is pretty positive confirmation in my mind that this is a
> hardware/setup issue.  I tried stopping by a local radio shack last night,
> but wouldn't you know they no longer carried simple attenuators.  Looks
> like i'll be picking one up online (or maybe ill lookup a schematic online
> and try building a simple one).


You can build a bridged-T, they are pretty simple.

                Z
                  2
      +--------/\/\/\-------+
      |                     |
      |     Z         Z     |
      |      0         0    |
 V ---+--/\/\/\--+--/\/\/\--+---- V
  i              |                 o
                 >
                 > Z
                 >   1
   |             |              |
   =             =              =


Assuming that the source impedance and load impedance (not shown) are
also Z0 (e.g. 75 ohms) then

Z   = a Z
 1       0

      Z
       0
Z   = --
 2     a 
    
 

V
 o      1
-- =  -----
V         1
 i    1 + -
          a

                           V
                            o
Power Gain in dB = 20 log ----
                           V
                            i


So pick an attenuation that you want to achieve, and that tells you "a".
"a" will tell you the resistor values you need.  Then try and pick ones
that are close.

All this of course assumes I created and solved my system of equations
without making an error. 


Regards,
Andy

> On a side note - Thank you very much for hacking on the saa7164 - other
> than this frequency glitch its been working great for me!

Re: Hauppauge 2250 - second tuner is only half working

[Steven Toth]

> On a side note - Thank you very much for hacking on the saa7164 - other
> than this frequency glitch its been working great for me!

You're welcome! :)

-- 
Steven Toth - Kernel Labs
http://www.kernellabs.com

Re: Hauppauge 2250 - second tuner is only half working

[Steve Harrington]

I have the same problem as Seth.  My system will tune RF channel 85 and 
below will not tune RF channel 91 and above on the second tuner only.  
First tuner works fine on all channels as does the PC HDTV 5500 also in 
the system.  My normal configuration is a 4-way splitter from the cable 
into the house.  One leg goes to a 2-way that feeds the two cards in the 
system.  With this configuration the normal result is:

Frontend /dev/dvb/adapter2/frontend0 tuned in 1 seconds.
Channel 80:    Standard=ATSC, Modulation=QAM_256
       SNR=0X0172, SIGNAL=0X0172

for channel 80 and:

Unable to set frontend /dev/dvb/adapter2/frontend0:frequency=669000000, 
modulation=QAM_256

for channel 103.

After reading Steven Toth's reply I tried adding 1 and then 2 2-way 
splitters before the 2250 input. No joy.  I also tried feeding the cable 
directly into the 2250 with no splitters.  Again - no joy.
Any other ideas?

Re: Hauppauge 2250 - second tuner is only half working

[seth]

I RMA'ed that tuner - awaiting a replacement. I found that adding and/or removing all splitters made no difference (wasn't able to get a hold of a proper attenuator - did try the splitter with a cap on one side method without luck).

I did find shortly after a kernel upgrade, a few modprobes, a couple reboots, some firmware file moves, and a dist-clean & rebuilding the modules - that 2nd tuner started working for the "gap" frequencies. I fired up mythtv and switched over to watch the second tuner and it was working fine for about 30s, then the system spontaneously rebooted. I played around a bit, but all subsequent attempts to get the "gap" frequencies to work again failed (right back to same problem as before - it won't get a lock, but first tuner was working fine).  I don't know if it was something with the state of my system just after a new kernel was installed or maybe it was just coincidence (maybe the signal attenuation was just right for a little bit there). Either way that spontaneous reboot seemed like maybe the card / tuner hardware itself might be defective.

I still had a few days left to make good on my 30 day warranty so I figure I'd give that a try. I'll let you know in a week or so if the new card is any better.

-Seth

------Original Message------
From: Steve Harrington
Sender: linux-media-owner@vger.kernel.org
To: linux-media@vger.kernel.org
Subject: Re: Hauppauge 2250 - second tuner is only half working
Sent: Aug 25, 2009 2:01 PM

I have the same problem as Seth.  My system will tune RF channel 85 and 
below will not tune RF channel 91 and above on the second tuner only.  
First tuner works fine on all channels as does the PC HDTV 5500 also in 
the system.  My normal configuration is a 4-way splitter from the cable 
into the house.  One leg goes to a 2-way that feeds the two cards in the 
system.  With this configuration the normal result is:

Frontend /dev/dvb/adapter2/frontend0 tuned in 1 seconds.
Channel 80:    Standard=ATSC, Modulation=QAM_256
       SNR=0X0172, SIGNAL=0X0172

for channel 80 and:

Unable to set frontend /dev/dvb/adapter2/frontend0:frequency=669000000, 
modulation=QAM_256

for channel 103.

After reading Steven Toth's reply I tried adding 1 and then 2 2-way 
splitters before the 2250 input. No joy.  I also tried feeding the cable 
directly into the 2250 with no splitters.  Again - no joy.
Any other ideas?
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html



Sent via BlackBerry from T-Mobile

Re: Hauppauge 2250 - second tuner is only half working

[Steven Toth]

> After reading Steven Toth's reply I tried adding 1 and then 2 2-way
> splitters before the 2250 input. No joy.  I also tried feeding the cable
> directly into the 2250 with no splitters.  Again - no joy.
> Any other ideas?

Most likely this is dependent on what frequency the other tuner is tuned to, 
especially since Seth indicated it worked for a short time. And, again, I don't 
see the issue but two other people do.

RMA is probably not the answer.

When you next get a chance to test please use azap and keep track of what 
frequency the first tuner is currently tuned to even if tuner#1 is technically 
no longer streaming. I suspect varying the frequency on tuner#1 will vary your 
test results.

-- 
Steven Toth - Kernel Labs
http://www.kernellabs.com

Re: Hauppauge 2250 - second tuner is only half working

[Steven Toth]

On 8/25/09 6:59 PM, Steven Toth wrote:
>> After reading Steven Toth's reply I tried adding 1 and then 2 2-way
>> splitters before the 2250 input. No joy. I also tried feeding the cable
>> directly into the 2250 with no splitters. Again - no joy.
>> Any other ideas?
>
> Most likely this is dependent on what frequency the other tuner is tuned
> to, especially since Seth indicated it worked for a short time. And,
> again, I don't see the issue but two other people do.
>
> RMA is probably not the answer.
>
> When you next get a chance to test please use azap and keep track of
> what frequency the first tuner is currently tuned to even if tuner#1 is
> technically no longer streaming. I suspect varying the frequency on
> tuner#1 will vary your test results.
>

OK, I can repro the issue.

Tuning tuner 1 to 669 works, then tune tuner2 to 669 no lock. Set tuner #1 to 
579 is locks, then tuner2 automatically also goes into lock.

So, it depends on where tuner#1 was previously tuned to.

I'll look into this.

Save yourself the trouble of the RMA if it hasn't already shipped.

-- 
Steven Toth - Kernel Labs
http://www.kernellabs.com

Re: Hauppauge 2250 - second tuner is only half working

[Steven Toth]

On 8/25/09 7:23 PM, Steven Toth wrote:
> On 8/25/09 6:59 PM, Steven Toth wrote:
>>> After reading Steven Toth's reply I tried adding 1 and then 2 2-way
>>> splitters before the 2250 input. No joy. I also tried feeding the cable
>>> directly into the 2250 with no splitters. Again - no joy.
>>> Any other ideas?
>>
>> Most likely this is dependent on what frequency the other tuner is tuned
>> to, especially since Seth indicated it worked for a short time. And,
>> again, I don't see the issue but two other people do.
>>
>> RMA is probably not the answer.
>>
>> When you next get a chance to test please use azap and keep track of
>> what frequency the first tuner is currently tuned to even if tuner#1 is
>> technically no longer streaming. I suspect varying the frequency on
>> tuner#1 will vary your test results.
>>
>
> OK, I can repro the issue.
>
> Tuning tuner 1 to 669 works, then tune tuner2 to 669 no lock. Set tuner
> #1 to 579 is locks, then tuner2 automatically also goes into lock.
>
> So, it depends on where tuner#1 was previously tuned to.
>
> I'll look into this.
>
> Save yourself the trouble of the RMA if it hasn't already shipped.
>

I was able to repro the issue once however during patching the issue went away, 
never to return - regardless of whether the patch was active or not. I even ran 
a series of cold boots to try and repro the behavior but I cannot.

I have seen the issue and I believe it exists, I just cannot get a reliable repro.

If you can test tuner 1 on selected frequencies then test tuner 2 always against 
channel 103 (669MHz) and find a reliable repro case then I'll take another look.

Annoying.

-- 
Steven Toth - Kernel Labs
http://www.kernellabs.com

Re: Hauppauge 2250 - second tuner is only half working

[seth]

Well my card is out the door already.  So it'll be a week or so till i can
try again. I'll give it a pretty thorough run down when i get the new
card, maybe I can dig up a repro.

This is probably just a red herring, but FWIW I had never cold booted the
machine (except monday morning when i yanked the card).  I warm booted
plenty, but i frequently would run full us-Cable scan's on both tuners. 
Some time last week when repo's pushed out 2.6.28-15, i had at least one
warm boot in there where i had the modules/firmware missing.  I
reinstalled (dist-clean, make, make install), rebooted, and tried again
and found it was working (well, for a little while until that spontaneous
reboot).

-Seth

> On 8/25/09 7:23 PM, Steven Toth wrote:
>
> I was able to repro the issue once however during patching the issue went
> away,
> never to return - regardless of whether the patch was active or not. I
> even ran
> a series of cold boots to try and repro the behavior but I cannot.
>
> I have seen the issue and I believe it exists, I just cannot get a
> reliable repro.
>
> If you can test tuner 1 on selected frequencies then test tuner 2 always
> against
> channel 103 (669MHz) and find a reliable repro case then I'll take another
> look.
>
> Annoying.
>
> --
> Steven Toth - Kernel Labs
> http://www.kernellabs.com
>
>

Re: Hauppauge 2250 - second tuner is only half working

[Steven Toth]

On 8/25/09 10:18 PM, seth@cyberseth.com wrote:
> Well my card is out the door already.  So it'll be a week or so till i can
> try again. I'll give it a pretty thorough run down when i get the new
> card, maybe I can dig up a repro.
>
> This is probably just a red herring, but FWIW I had never cold booted the
> machine (except monday morning when i yanked the card).  I warm booted
> plenty, but i frequently would run full us-Cable scan's on both tuners.
> Some time last week when repo's pushed out 2.6.28-15, i had at least one
> warm boot in there where i had the modules/firmware missing.  I
> reinstalled (dist-clean, make, make install), rebooted, and tried again
> and found it was working (well, for a little while until that spontaneous
> reboot).

Fair enough.

Another user is preparing remote access so I can take a look at the problem 
first hand. With any look we'll see the issue and have a patch by the time your 
replacement board comes back.

- Steve

-- 
Steven Toth - Kernel Labs
http://www.kernellabs.com