KWorld PE3552T PCIE Dual DVB-T (1 Viewer)

[debaser_uk]

Hi,

I have one of these KWorld's in my new HTPC and I'm having a terrible time with it. Has anybody else got one of these working OK? I've tried the drivers that came with it, and the drivers from the KWorld website, but it's a pig of a card. I'm getting random system freezes, whilst watching live TV, or when the TVServer is recording anything at all. In fact, whenever the TVService is using the card, I can get a system freeze at any time. Which isn't making me very popular at home. This is with a clean install of XP Pro 32bit with SP3 and all updates.

That'll teach me to worry about my energy efficiency, I just retired two twinhan PCI single tuner cards from my system, they served me well for 3+ years, I've had the KWorld for a week, and I'm about to bin it and buy something else.

So, if anyone else is using this card succesfully, can you let me know please?

Kworld Product Deatials: PCI Express Dual Digital TV Card | PE3552T

 

[Mike_S]

I've been having similar trouble as you. I've been thinking of binning it and buying another. Did you get yours sorted? if so how?

 

[madcart]

Similar issues also - did either of you figure out how to fix this?

 

[pilehave]

I have the MSI TV@nywhere E-Duo card, which I think is the same hardware as the KWorld - just branded differently. You might wan't to try these drivers?
Maybe it'll work
MSI Europe ? Computer, Laptop, Notebook, Desktop, Mainboard, Graphics and more

You might also wanna try to move the card to another PCI-E slot if you can...

 

[madcart]

Thanks - I think it is the same hardware.

Removing nvidia nTune seemed to fix a lot of stability issues.

The tv card now works reliably with either the tv@nywhere driver, or the latest kworld driver after a reboot.
However if the pc goes into S3 sleep for a long period (ie 10 10 minutes or more) the tv server no longer detects a signal, and i have to do a full reboot to get it working again. This happens with all the drivers.

Its as if the card shuts itself down in S3 sleep, and doesn't start up again when the system resumes. I dont know too much about PCI-E or if this is a real proposition or not. I only have oen slot so can't move the card..

 

[mm1352000]

Hi there

Problems after resume is a common problem, but it is not always caused by the TV card. I recommend you try:
1. Set up a static IP address.
2. Reinitialise TV Server on resume.
3. Delay the start of TV server (to give the tuner driver time to reload).

 

[madcart]

Thanks for this advice.
I had been reinitializing the tuner, and adding a delay didn't fix things either. TV Service seems to recognize the card ok, but the card provides no signal. Presume this is different to the driver not being loaded?

I've discovered that restarting the TV card using devcon.exe always gets it working again.
The DevCon command-line utility functions as an alternative to Device Manager

However to do this, TV Service must be down.
So i have a script that takes down tv service, restarts the card and brings the tv service back up again.
This generally works ok - I configured the TV service power scheduler to run this script on wakeup. TV Service restarts itself and the TV card, and i am now able to record TV from the suspend state.

Its not totally reliable though - if i suspend the PC and wake it back up straight away manually, restarting the TV card often fails (or results in the BSOD). TV Service now throws an exception when it tries to access the card - i assume this is what happens when the driver isn't loaded?
My suspicion is that in this case the card is still functioning correctly and is reinitializing itself when i try and restart it, which it doesn't like.

Curious about static ip suggestion - implication is that obtaining a new ip from DHCP screws things up?

 

[mm1352000]

Hi again

I've discovered that restarting the TV card using devcon.exe always gets it working again.
The DevCon command-line utility functions as an alternative to Device Manager

However to do this, TV Service must be down.
So i have a script that takes down tv service, restarts the card and brings the tv service back up again.
This generally works ok - I configured the TV service power scheduler to run this script on wakeup. TV Service restarts itself and the TV card, and i am now able to record TV from the suspend state.

Hmm, that indicates a tuner driver issue with waking up. Do you remember how long you delayed the TV Server startup when you were testing?

Its not totally reliable though - if i suspend the PC and wake it back up straight away manually, restarting the TV card often fails (or results in the BSOD). TV Service now throws an exception when it tries to access the card - i assume this is what happens when the driver isn't loaded?
My suspicion is that in this case the card is still functioning correctly and is reinitializing itself when i try and restart it, which it doesn't like.

I'd guess the same - either the driver didn't fully unload or it reloaded by itself. Perhaps you can work around this by adding some "smarts" to your devcon script. For example, rather than just blindly restarting the device, try updating then checking its status. The status might tell you whether it is doesn't need a restart...

Curious about static ip suggestion - implication is that obtaining a new ip from DHCP screws things up?

Yes, exactly. The TV Server and MySQL processes usually bind to the IP address on your network adapter. If your network adapter wakes up sometime after the DHCP lease has expired then it will receive a new IP address, and the address the TV Server and MySQL were bound to will be invalid. They won't like that...

 

[madcart]

Hi,

Do you remember how long you delayed the TV Server startup when you were testing?

I added a delay of 20 seconds I think - though i didn't play around with this too much.
After initialization, if the tv service can't tune in, it will never tune in, but there is no other error in the log that i saw. However from what you are saying, if tv service reinitializes itself before the driver has been reloaded then this is what I might expect to see?
I'm pretty sure that i restarted tv service when it was in this state (several minutes after wakeup from sleep), but i'll double check this. If restarting the tv service fixes the issue, then it seems like i just need to get the delay correct.

Perhaps you can work around this by adding some "smarts" to your devcon script. For example, rather than just blindly restarting the device, try updating then checking its status.

I'll play around with updating and checking the status of the device tonight also - i'd only ever checked the status before...

If your network adapter wakes up sometime after the DHCP lease has expired then it will receive a new IP address, and the address the TV Server and MySQL were bound to will be invalid. They won't like that...

Hm - don't think i'm suffering from this. I'd assume that if this were the case, MP wouldn't be able to connect to MySQL when it starts up, and hence wouldn't retrieve the EPG?

Many thanks again for your tips and advice!
I'll try these out tonight and let you know what happens..

 

[madcart]

I checked a few things last night:

* Reinitializing or restarting TVService never 'fixes' the TV card once it has failed to detect a signal, no matter how long i wait for the card to reload. Restarting the card using devcon is the only way to get it working again.

* devcon.exe status always reports that the card is running, no matter what state it is in. devcon.exe update requires the .inf file for the card, and i'm not sure where that is located.

* Even devcon claims that the card filed to restart, and a reboot is required, it seems to work ok.

I actually wonder if the issues with the card are to do with my motherboard struggling to come out of S3 sleep. If i put the system into and out of sleep in regular succession (ie whilst testing this) eventually it fails to come out of sleep at all, and needs a full reboot.
I read in a different thread that this is likely to be due to the power supply having a low voltage across one of the terminals.