I bought one of the PCTV HD Ultimate Sticks, and returned it for one of the pro's, since seeing your progress in working on this. The Ultimate stick is labeled an '880e' usb device.

Thanks for replying. Please post if you take the time to try out the driver and let me know how it went.

Mrechberger confirmed the "Ultimate" version is a completely different hardware design using a new em28xx chip that is unsupported:

http://www.linuxtv.org/pipermail/linux-dvb/2007-November/021908.html

I'm actually curious about the device that is one step lower than the "Pro", which I believe just known as the "PCTV HD Stick". If it's an 801e then I will update the Wiki to reflect that so people can just buy the cheaper device. If it's the same hardware but a different USB ID, I can add the device ID so it starts working.

Cheers,

Devin

Well, after installing, the device was detected fine. After a while though, it became hot to touch, which I found a little worrying. Ran it with mythtv, and it seems to be working well. I need to learn the ins and outs of mythtv though (first dvb card, so all a pretty new experience).

I'm running Fedora 9, kernel 2.6.25.


Thanks for all the work you've put into this.

Christopher

When you said it was "hot to the touch", do you mean the entire device or just the F-connector that the antenna was plugged into (which is normal)?

Thanks for the feedback. I will take a thermal reading of the device under load and compare it to the Windows driver to be sure it isn't introducing any adverse effects.

it was the entire stick that felt hot. I'm going to try a few more things with it when the weekend is finished (lots of back-work to take care of), but I'll let you know if anything comes up.

Thanks again for all your hard work in getting this device up and running.

Christopher

That's good to know.

It's possible I'm not putting the CX25843 to sleep when I'm in digital capture mode. I'll go over the GPIOs tonight.

Ok, I did some more testing. Add the following to dib0700_devices.c at line 1131:

/* Put the CX25843 to sleep for now since we're in digital mode */
dib0700_set_gpio(adap->dev, GPIO2, GPIO_OUT, 1);

So it should end up looking like:

        dib0700_set_gpio(adap->dev, GPIO4, GPIO_OUT, 1);
        dib0700_set_gpio(adap->dev, GPIO7, GPIO_OUT, 1);
        dib0700_set_gpio(adap->dev, GPIO2, GPIO_OUT, 0);
        msleep(30);

        /* Put the CX25843 to sleep for now since we're in digital mode */
        dib0700_set_gpio(adap->dev, GPIO2, GPIO_OUT, 1);

        /* GPIOs are initialized, do the attach */
        adap->fe = dvb_attach(s5h1411_attach, &pinnacle_801e_config,
                              &adap->dev->i2c_adap);

I'll update my patch a little later so others don't have this issue

Edited at 2008-08-29 01:33 am (UTC)

I've applied your patches to my MythTV machine last night and everything seems to be working great.... I've had the same heat problem that the other poster has already mentioned, and I'm applying your patch for that right now. I won't be able to feel it to see if it worked or not until after I get off work though.

Didn't notice any problems with anything else DVB related in my machine, although I'm not running anything else with a dbi0700.

Great job on getting this device working, it is very much appreciated.

-Ken

Yes, please definitely be sure to apply the second patch mentioned above, and report back your results. The second patch should make the device run much cooler.

Thanks,

Devin

I have a 801e SE, and other than the USB ID being 2304:023b instead of 2304:023a, it seems to be exactly the same. Hardware specs appear to be the same, and indeed I've got some ATSC video up right now, by just adding one line to recognize the 023b ID.

As for the stick: I don't have your second patch in yet, but I ran the stick for a few minutes on a friend's Windows box, and it was quite hot on there too. I think this stick just runs hot!

Hello,

Thanks for the feedback. It's good to finally know what the deal is with the non-Pro stick (I had expected as much).

The device does run hot under Windows as well, from what I have seen. In what would seem to be a bit of irony, I think even the Pinnacle driver for Windows always runs both the demodulator and the cx25843 at the same time. And my original GPIO programming was based off the USB trace of the Windows driver.

Please try out the second patch and let me know if it runs cooler (it's a one line change).

Couple of additional questions:

Does the device have a small port on the side that looks like a USB mini port? On the Pro Stick, this is where an A/V cable would be connected.

Although the product does not come with a remote control, does it appear to have an infrared window on the device?

Would you be willing to pop the cover off and take photos of the PCB and send them to me? In particular, I need to know if there is CS5340 chip which on the Pro device is used for FM radio? The two PCBs are not soldered together and separate quite easily.

Would you be willing to work offline with me to get a USB capture of the Windows driver with SniffUSB? My concern is that while ATSC may be working, I don't know if the GPIOs are being programmed properly for any missing hardware (like the FM radio support)

http://www.pcausa.com/Utilities/UsbSnoop/default.htm

I would really like to get support for both devices into the mainline, but I am hesitant to just check in the additional USB id without knowing I'm not going to burn out everybody's devices after a month of use. :-)

"The device does run hot under Windows as well, from what I have seen. In what would seem to be a bit of irony, I think even the Pinnacle driver for Windows always runs both the demodulator and the cx25843 at the same time"
That is pretty funny 8-) I always love when the Linux driver for something ends up actually being better than the Windows one.

"Please try out the second patch and let me know if it runs cooler (it's a one line change)."
Yeah that cooled it down a fair amount. It seems to me it was already cooler (without this patch) than I recall it being on my friend's Windows box.

"Does the device have a small port on the side that looks like a USB mini port?"
Nope.
"Although the product does not come with a remote control, does it appear to have an infrared window on the device?"
It *appears* to, but (see the photos below) there's nothing underneath -- just an LED.. heh.


"Would you be willing to pop the cover off and take photos of the PCB and send them to me?"
Yep. It DOES in fact have the CS5340. These'll take several seconds to load...




"Would you be willing to work offline with me to get a USB capture of the Windows driver with SniffUSB?"

I can certainly test stuff out, but I don't have any Windows installs handy. I think it's the same stick with IR and the extra input deleted though.

"I would really like to get support for both devices into the mainline, but I am hesitant to just check in the additional USB id without knowing I'm not going to burn out everybody's devices after a month of use. :-)"

I can certainly test things out. I have a mythtv setup with a BT878 right now. I need a fairly large antenna to get more than 2 channels over the air, so I intend to hook the PCTV up to cable until then, and I think there's quite a few analog-only channels on there. So 1) I can have mythtv give it plenty of exercise and 2) If I burn it out it's not a big deal, I have a working tuner already setup.

I had to do a 'hg revert -d 8/36/08' on the v4l-dvb clone for it to patch and compile correctly, otherwise the stick would not initialize (load the firmware I think as I could see it with lsmod).

I don't know if this was just me or not but I am running a brand new install of Ubuntu 8.04 i686 with kernel 2.6.24-19-generic.

Anyhow works now! Thanks for doing the work on adding 801e support, this is great!

The hg diff was based off of a patch from last week, and I believe device support for a couple of other devices got checked in (which would create a conflict).

The i2c fixes went into Patrick's branch yesterday and a pull has been submitted to Mauro. The 801e fixes are being submitted to him tonight, so everything should be in the head sometime in the next couple of days.

Until then, you're going to have to stick with what you've got...

Also, make sure to apply the one line the patch mentioned in the previous comments.

No problem and done. Just though that I would mention this step as it caused me an hour of confusion.

Thanks again!

No, I appreciate the feedback, and it might help someone else trying out the patch.

This morning the final changes for the 801e got pulled into http://linuxtv.org/hg/~pb/v4l-dvb/ so I'm just now waiting on Mauro to do the pull into the mainline v4l-dvb.

Looks like we're almost there...

I also own an 801e SE and a Windows Vista PC that I'm willing to run a USB capture on it for you. Do you need me to just plug it in and capture Windows initializing it, or do I also need to launch Media Center and tune in some ATSC broadcasts?

-Mike

If you could help out in getting a capture, I will happily add the device profile for inclusion in the 2.6.28 kernel.

Another user reported that "it's working for him" after just changing the USB ID in the driver, but I want to be sure the GPIOs are the same before I check in a change.

Here's what to do:

1. Get SniffUSB 2.0:
http://www.pcausa.com/Utilities/UsbSnoop/default.htm
2. Install the drivers and the software that came with the device
3. Make sure the software is working by doing a scan for ATSC stations
4. Start SniffUSB and "attach" to the device
5. Start capture in SniffUSB
6. Hit the "replug device" button in Sniff USB
7. Launch the viewing application
8. Tune to a single ATSC channel
9. As soon as you see video, stop capture in sniffusb

In particular, you want to do steps 5-9 in as little time as possible, to minimize the size of the capture file. If you're quick, the file will be around 50 MB.

You can either make the file available to me for download, or you can take one extra step by downloading "USBReplay" and running the script through parser.pl first from a Linux box:

wget http://mcentral.de/hg/~mrec/usbreplay/raw-file/486a40310c52/parser.pl
perl parser.pl capture_file > foo

Then send me the file "foo" (it will be *much* smaller)

My email is "devin.heitmueller" at "gmail.com"

Thanks!

Devin

You got it, Devin!

The parsed file has been sent.

-Mike

Great. I'll look at it this weekend and if all is well I'll submit a patch.

Thanks for your help,

Devin

Please forgive me if this isn't the right place to pose this.

I've downloaded today's mercurial v4l-dvb which looks like it has the merge (thanks!).

After installing, when I reboot or unplug and replug the stick, I get this in dmesg:

dvb-usb: found a 'Pinnacle PCTV HD USB Stick' in warm state.
dvb-usb: will pass the complete MPEG2 transport stream to the software demuxer.
DVB: registering new adapter (Pinnacle PCTV HD USB Stick)
DVB: Unable to find symbol s5h1411_attach()
dvb-usb: no frontend was attached by 'Pinnacle PCTV HD USB Stick'
input: IR-receiver inside an USB DVB receiver as /class/input/input4
dvb-usb: schedule remote query interval to 150 msecs.
dvb-usb: Pinnacle PCTV HD USB Stick successfully initialized and connected.
usbcore: registered new interface driver dvb_usb_dib0700

When I start mythtv-setup, and begin to setup a new card, I see an unknown dvb device. I'm guessing that is to do with the lack of a frontend being loaded.

Is this the way it should work? Do I maually need to modprobe a frontend driver? If so, which one, if not, do you have any suggestions?

Thanks,
Ray

Ok, I guess you didn't see the post that said "don't update yet!"

http://devinjh.livejournal.com/146548.html

The code in the v4l-dvb will currently panic your kernel if you were to actually try to tune. I submitted a patch Sunday night but Mauro hasn't merged it yet.

Regarding the problem you reported, start by rebooting your PC. It's possible that not all the modules are in sync. Doing a "make install" will install all the new modules but it won't reload any modules already loaded.

If that doesn't work, email me at devin.heitmueller at gmail.com and I will work with you to isolate the problem.

Regards,

Devin

Thanks for letting me know that it'll panic.

I tried unplugging the usb, rebooting, make rminstall, reboot, make install, reboot and re-plugging. It still has the issue of not autodetecting the frontend.

I then manually modprobe s5h1411 and xc5000, replugged the stick and both dmesg and mythtv-setup look correct now.

Now that I can at least manually load the drivers, I'll revert and apply the patches you provided earlier. I'll figure out/remember how to automatically load the modules by preloading when dib0700 loads.

Thanks again!

Ok, well that's not good. I will look at the dependencies code, since it appears to be working here fine.

What distribution and kernel version are you running?

Also, regarding the panic, the fix is pretty simple and you can do it by hand:

Open v4l/dib0700_devices.c
Find the xc5000_tuner_attach() function
Move the following lines so that they appear before the return statement:

       /* FIXME: generalize & move to common area */
       adap->fe->callback = dib0700_xc5000_tuner_callback;

I'm using Gentoo, still at 2.6.23-gentoo-r3, likely not a very popular release anymore, so please don't spend too much time on my account.

Modified the xc5000_tuner_attach() function. I'll test once I can pry the remote from my wife's hands for a little while.

Just a follow up...

After fixing v4l/dib0700_devices.c and manually adding frontends to dib0700 dependencies in modules.dep, reboots are clean, capture card detected in myth, Comcast in Fremont, CA cable qam-256 scanning detects, tunes, plays live tv, and records unencrypted tv without issue.

Thanks again,
Ray

That's good to hear. Thanks for the feedback.

I'm still wondering why you had to do something special to make the deps work. It's on my todo list to investigate later this week.

Devin

Hello.

I accidentally bought 340e, and what it seem to share same firmware/driver with 801e.

http://bahathir.googlepages.com/340e.tgz

I am willing to try any patch with my system.

GNU/Linux Slamd64 (Slacware 12 x86_64 clone)
linux kernel v2.6.27
gcc 4.4.0

The modules/firmware loaded successfully, and the green LED lit.

But, I got this dmesg error during scanning.

DiB0070 I2C write failed


Here is my patch against 2.6.27 vanilla


--- orig/dib0700_devices.c 2008-10-13 13:33:58.000000000 +0800
+++ dib0700_devices.c 2008-10-13 14:00:31.000000000 +0800
@@ -1119,6 +1119,7 @@
{ USB_DEVICE(USB_VID_LEADTEK, USB_PID_WINFAST_DTV_DONGLE_STK7700P_2) },
/* 35 */{ USB_DEVICE(USB_VID_HAUPPAUGE, USB_PID_HAUPPAUGE_NOVA_TD_STICK_52009) },
{ USB_DEVICE(USB_VID_HAUPPAUGE, USB_PID_HAUPPAUGE_NOVA_T_500_3) },
+ { USB_DEVICE(USB_VID_PINNACLE, USB_PID_PINNACLE_HYBRID_PRO_STICK_340E) },
{ 0 } /* Terminating entry */
};
MODULE_DEVICE_TABLE(usb, dib0700_usb_id_table);
@@ -1126,7 +1127,7 @@
#define DIB0700_DEFAULT_DEVICE_PROPERTIES \
.caps = DVB_USB_IS_AN_I2C_ADAPTER, \
.usb_ctrl = DEVICE_SPECIFIC, \
- .firmware = "dvb-usb-dib0700-1.10.fw", \
+ .firmware = "dvb-usb-dib0700-1.20.fw", \
.download_firmware = dib0700_download_firmware, \
.no_reconnect = 1, \
.size_of_priv = sizeof(struct dib0700_state), \
@@ -1374,7 +1375,7 @@
}
},

- .num_device_descs = 4,
+ .num_device_descs = 5,
.devices = {
{ "DiBcom STK7070PD reference design",
{ &dib0700_usb_id_table[17], NULL },
@@ -1391,6 +1392,10 @@
{ "Hauppauge Nova-TD-500 (84xxx)",
{ &dib0700_usb_id_table[36], NULL },
{ NULL },
+ },
+ { "Pinnacle 340E",
+ { &dib0700_usb_id_table[37], NULL },
+ { NULL },
}
}
}, { DIB0700_DEFAULT_DEVICE_PROPERTIES,

--- orig/dvb-usb-ids.h 2008-10-13 14:19:41.000000000 +0800
+++ dvb-usb-ids.h 2008-10-13 13:43:31.000000000 +0800
@@ -206,4 +206,8 @@
#define USB_PID_ASUS_U3100 0x173f
#define USB_PID_YUAN_EC372S 0x1edc
#define USB_PID_DW2102 0x2102
+
+/* 340e support */
+#define USB_PID_PINNACLE_HYBRID_PRO_STICK_340E 0x023d
+
#endif


Thank You

Mohammad Bahathir Hashim.

Sorry I forgot to include the reference sites :)

http://www.gkall.nl/pinnacle-pctv-hybrid-pro-stick-340e.html
http://www.linuxtv.org/wiki/index.php/Pinnacle_PCTV_Hybrid_Pro_Stick_340e

Mohammad Bahathir Hashim

Unfortunately, the 340e has an xc4000, which is totally unsupported with Linux.

As far as I know, nobody has the datasheet, and nobody has taken the time to reverse engineer the difference between the xc4000 and the xc5000. Since I'm in the United States, I don't have access to a signal that I could use to reverse engineer it (it does dvb-t only whereas I only have access to ATSC).

Other than the xc4000 though, everything else *should* be the same.

Sorry.

Devin

Edited at 2008-10-13 02:13 pm (UTC)

I got it working with the 10/11 v4l-dvb sources. The patch would only apply to 8/26 sources, but wouldn't make a usable 2.6.27 module. I still needed to comment out SAA52?? stuff(./v4l/.config) so it would compile with 10/11 sources. It works though. Any eta on analog / radio support? I'm more interested in those abilities than dvb.

I'm a DVB newbie, so it took a while to figure out to use scan on kaffeine's us-ATSC-* confs to get a channel list. Kaffeine's current scanning is broke by all accounts. But I have at best 13 channels locally for ATSC / DVB. 3 don't render with vlc. 1 has no sound. And all seem to cop out after 20 minutes for some reason. It starts getting choppy then bumps a lot of erros like this:

Errors: [00000458] main decoder error: decoder is leaking pictures, resetting the heap
(while running vlc on a .conf of an individual channel. Doesn't matter which channel, they all get to that point eventually.)

And various libdvbpsi errors. For most channels. I'm not sure if this is normal. And may just be something odd with my local carriers. Like I said, a dvb newbie. But I bought the card for analog and radio. To avoid buying another TV and to get some of my old VHS tapes to DVD. Oh well for now I guess.

Thanks for making a driver. Even if it took me two days to figure it out.

Hello there,

Please don't use the patch referred to in the previous post. Everything is now merged into the v4l-dvb mainline.

Analog support is not present because it doesn't exist in the dvb-usb framework. I am slated to work with mkrufky to add such support, at which point making it work with the Pinnacle 801e will be straightforward. No ETA on that though, as it's relatively low priority (sorry).

I added ATSC scanning support to Kaffeine 0.8.7. If you're using Kaffeine 0.8.6 (which still ships with most distros), scanning is not going to work. Also, there's a known issue with 0.8.7 where it works with over the air ATSC, but not with cable QAM.

I don't think vlc supports ATSC (but I could be wrong and they might have added support since I last looked). If you're not getting sound, it might be because the MPEG decoder used by vlc doesn't recognize the ATSC AC3 sound extension (differs from how it's done in DVB).

Please throw away my patch and use the current v4l-dvb source. It's possible that will fix your issue with the picture cutting out after 20 minutes (there was a thermal bug not addressed in the patch, but the fix is later in the comments section as well as in the v4l-dvb mainline).

Regards,

Devin

Confirmed: current v4l-dvb 10/13 seems to not have the cut out. Although I was using fresh sources from 10/11 when I had the issues. hg pull --update (with 55 updates)

I seem to have fewer issues on my 64 bit desktop than my 32 bit laptop. But I still get a number of PSI errors for some channels.

libdvbpsi error (PSI decoder): PSI section too long
And a bunch of TS discontinuity messages. Some channels run relatively error free. Others generate nothing but errors.

VLC seems to be the only one I've gotten to show most dvb channels. Mplayer only seems to play one channel based on the output of czap as it's channels.conf. Scan found 13 channels on my Desktop and only 10 on my laptop. And at least two channels generate mostly errors with no output on each. Locally it would be Pentago (military channel) that has no audio. And [0008] and [0009] seem to be the local channels that don't yield any media output. Not that I'm worried about it, windows only shows 8 channels.

The device does get hot in both OSes. Steering wheel in the summer(texas) sun hot. But not enough to cause 2nd degree burns (yet). Not that I can tollerate touching the metal cable joint for very long. I'll just assume that this is normal.

Thanks again,
James

First, let's talk about the thermal:

Yes, both the Xceive xc5000 and xc3028 do run a bit hot, regardless of the OS. This is normal. They improved the thermal properties in the xc3028L chip.

There was a bug in the patch I had posted on this site where the device in general ran hotter than it should (resulting from an improperly configured GPIO that had both the digital demodulator and analog decoder running at the same time). It was actually a bug in the Windows driver's GPIO config that I faithfully copied from the USB trace. But it's fixed in the v4l-dvb mainline.

There is an outstanding issue where the tuner is never powered down even when the device is idle. This causes the tuner's f-connector to be hot even when you are not actively using the device. I have a patch but it affects all xc5000 based devices so it is still pending with the original author.

Regarding your reception issues, are you *sure* that you just don't have a crappy signal for those channels? Have you tried it with the device under Windows keeping the same antenna orientation? That's certainly what it sounds like.

If you send me the output of running azap for a few seconds for each of the channels in question, I can tell you how good the signal is.

I got mplayer working for me. I much prefer mplayer. And at least one channel is signal deprived. But all others appear to work better on my 64 bit desktop than my 32 bit laptop. The desktop being 1.9GHz dual core. The laptop being 2GHz single core. The desktop having 1GB of ram. the laptop having 512MB of ram. At a minimum mplayer has audio for Pentago (military channel), where vlc does not.

I don't consider thermal an issue. I wouldn't want to leave the device unattended with it touching a flamable source (carpet). But for my current uses, it does seem to run cooler in Linux than under Windows. Not to imply cool. So for my purposes (linux) it is a non-issue. It does concern me that it might affect overall device life, but hopefully not. I'm not as worried for this $100 device versus my $300 300GB USB HDD seagate drive that never quite worked right. And is in essence a $300 paper weight at this point. Still a 300GB HDD, and still works to a VERY limited extent. But essentially a BRICK for all dependable intents and purposes.

I'm not that concerned at signal strength. I'm sort of halfway between Austin and Waco, and only 4 of 13 channels seem to suffer from signal strength(or other issues). Everything seems to work better on the desktop.

At a minimum debian seems a bit weird. For some reason my 32 bit laptop install, which is up to date for updates, seems to have a 0.8.3 version of kaffiene. Were my 64 bit desktop, which is almost 1GB of updates out of date, seems to have a 0.8.7 version of kaffeine. Even though they're both running debian / stable. I guess I'll stop at this point as it's not a driver issue per say. But I'd really like to have analog support. I have a number of VHS tapes and a quality VHS player. Some old college home made tapes, and other odds and ends that are not available on DVD or blueray. Part of the justification for this purchase was to avoid the purchase of a camcorder that supports analog inputs.

On my desktop I have a noticeable lack of issues (64 bit / 1GB Ram / 1.9GHz dual core CPU). Where my laptop (32 bit / 0.5GB ram / 2GHz single core CPU) has more ussues. It might just be bus power. My laptop may be under powered on the USB bus. Or it might be video card driver related. My laptop has an ATI (integrated) video card. My desktop has an nVidia (intergrated) video card.

- James

I live in the country so I'm not really close to stations. But I'm using a big trucker V antennae and not the stock one that came with it. Wired up to some metal cohangers and dangling from the curtain rod at the window. Actually I have a splitter there too, so I'm using the big trucker V and the stock attennae. Seems to work better than either by themselves. It may be signal strength related. I seem to have better success during certain times of the day. And the weather this past week really made things weird.

So what's the eta on radio/analog support? 1 month? 6 month? 1 year? 6 years? Maybe never? Just curious since I'm not about to leave certain projects on hold for over a year. Not when there's many craigslists options for less than a tank of gas. And recording quality from analog in windows kind of sucks. Not sure if that's the packaged encoding software or not. But if that's how it's gonna look in linux, I might be better off getting another card. DVB in linux looks great though. Especially with a Full HD PBS station, which seems to have the best overall signal.

- James

Unfortunately, there isn't an ETA on analog support for this device. Adding support for analog requires some changes to the basic DVB infrastructure, and mkrufky doesn't have ETA on that yet. Once that infrastructure is in place though, getting analog working with the 801e is pretty straightforward.

If you're looking to see this working in the next couple of months, you might be better served just to take it back and get a $50 HVR-950 or the older variant of the device (Pinnacle 800e).

Regards,

Devin

Edited at 2008-10-21 05:53 pm (UTC)

Bad news I guess. One month almost to the day (bought 10/10/2008, died 11/9/2008) and the device no longer functions. The little led light no longer comes on. In windows it recognizes it as a usb device, but does not associate it with a known driver. And the led doesn't come on even in windows. Previously installed, driver and all, so the device is most likely toast. I guess I wont be waiting for an analog driver after all. Unless my only recourse is shipped replacements for the next 23 months of the warranty.

from /var/log/syslog

Nov 9 21:47:57 soundbox kernel: dib0700: loaded with support for 8 different device-types
Nov 9 21:47:57 soundbox kernel: usbcore: registered new interface driver dvb_usb_dib0700
Nov 9 21:49:01 soundbox kernel: usb 1-8: new full speed USB device using ohci_hcd and address 11
Nov 9 21:49:01 soundbox kernel: usb 1-8: device descriptor read/64, error -62
Nov 9 21:49:01 soundbox kernel: usb 1-8: device descriptor read/64, error -62
Nov 9 21:49:01 soundbox kernel: usb 1-8: new full speed USB device using ohci_hcd and address 12
Nov 9 21:49:01 soundbox kernel: usb 1-8: device descriptor read/64, error -62
Nov 9 21:49:02 soundbox kernel: usb 1-8: device descriptor read/64, error -62
Nov 9 21:49:02 soundbox kernel: usb 1-8: new full speed USB device using ohci_hcd and address 13
Nov 9 21:49:02 soundbox kernel: usb 1-8: device not accepting address 13, error -62
Nov 9 21:49:02 soundbox kernel: usb 1-8: new full speed USB device using ohci_hcd and address 14
Nov 9 21:49:03 soundbox kernel: usb 1-8: device not accepting address 14, error -62
Nov 9 21:49:03 soundbox kernel: hub 1-0:1.0: unable to enumerate USB device on port 8

Which version of the code were you using? The original patch? The original patch plus the fix described above to deal with the thermal problem? Or the latest version from v4l-dvb.

Also, what is your usage pattern? Are you capturing 24/7 with the device? Is it usually idle? What program are you using with it? (MythTV/Kaffeine)

Obviously this is *really* bad and I want to better understand the circumstances so I can prevent this from happening to others.

Devin

James

**PLEASE** contact me via email so we can discuss this matter further. I need to get to the bottom of this so it doesn't happen to anyone else.

devin.heitmuelller at gmail.com

Thanks,

Devin

I got the 801e working with the dvb drivers. So what happened? Is there analog support yet? Should I stop using the driver, so that mine doesnt die?



-N

Hello,

It seems to have been a one-time thing. I know of lots of users who run the device 24x7 without any issue.

Regarding analog, the issue is not with the 801e specifically. The dvbusb framework that 801e relies on does not have any analog support (and it will be a huge effort to add). For example, I just added analog support for the HVR-950q and it took me six hours a day, every day for the last six weeks.

There is some renewed consideration for doing the work for the dvbusb framework, but if we decide to do the work it won't be done anytime soon.

Devin

Devin Heitmueller of Kernel Labs is currently considering writing the driver for analog part of dib0700 devices:

"Some discussion I have been having with Andrej Falout has prompted me
to consider finally adding analog support for dib0700 based devices?
As a result, I am interested in trying to figure out how many people
care about this. In particular, this is a rather large project, so I
want to get an idea how many people would actually benefit from this
support (and might be willing to donate a couple of dollars toward the
effort).

I've got all the specs and information required to do the work - it's
just a matter of deciding whether it is worth spending three or four
weeks of my time that could be spent doing stuff that I would almost
certainly consider more interesting.

If you are interested in seeing this work, please reply here, or email
me privately. Also, include what dib0700 based device you have (and
what video decoder and tuner chips it uses if you happen to know).

--
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com"

If you are interested in this functionality, please let Devin know by sending a message to Linux Media Mailing List :
<linux-media@vger.kernel.org>

Um, I'm not sure who decided to post the above reply, but "Devin Heitmueller of Kernel Labs" is the same person as "Devin Heitmueller of devinjh.livejournal.com"

:-)