TBS6903, latest driver and CrazyCat's StreamReader.DLL

[vma]

After spending hours trying to understand where the bug was in my software, I discovered that the bug is not within my software at all.



With the latest 1.1.0.6 driver, StreamReader.DLL will not work correctly on the secondary tuner. It will connect, lock, send DiSEqC-commands, read signal level, etc. Even the constellation diagram works. But as soon as you try to stream or decode the TS packets, you will only receive corrupt ones.

Using a TV app that relies on BDA drivers, the secondary tuner works perfect.

When downgrading the driver to 1.0.0.5, StreamReader.DLL will work as expected and everything is fine.

There must be some bug between driver and StreamReader.DLL.

Bear this in mind before upgrading the driver!

Regards,
Vitor

 

[vma]

Just now I got a big scare, thinking I broke tuner 2 of my TBS6903.

I was testing the VMA Stream Reader with a new "old style" blindscan and wanted to compare it against EBSpro. So I had my blindscan running on tuner 2 (DiSEqC-1.0 on port 2) when I started EBSpro on tuner 1 (DiSEqC-1.2). As soon as the dish started to rotate to position itself, tuner 2 stopped receiving valid data.

I stopped everything and to my surprise, after restarting only my software, tuner 2 was still not working correctly.

Reinstalling the driver did not help, either (in fact something strange happend: the tuners got swapped so tuner 1 became tuner 2 and vice versa).

A reboot did not help.

I thought that I broke the card for good and decided to give it one last try: instead of rebooting, I switched off the computer for 10 minutes (while I was doing something else).

After restarting the computer, tuner 2 was working ok again. What a relief!

But I wonder what happend? Did the DiSEqC-1.2 command somehow overload the tuner 2? Why did I need to shutdown the computer? So that all caps got drained and the card could be started over? Somehow this incident is making me wonder if I should not stress the card so much (two parallel blind scans with DiSEqC commands)?

 

[s-band]

I don't use DiSEqC but have seen a few similar things. I have one each TBS6903, 6983 and 6903x installed and usually run multiple instances of Crazyscan with desktop shortcuts linked to the 6 channels.
1. If one half of a 6903/6983 is receiving and streaming and I scan using the other channel of the same card the streaming stops.
2. If an instance of CS crashes, usually due to operator error and I shut the PC down, on re-start the 6 channels can shuffle. That means that the shortcuts apparently link to the wrong card/channel. Actually, the card/channel data in the .ini files has somehow changed.
Maybe the something similar happens if there is a DC glitch caused by DiSEqC?

 

[vma]

I think something like this must have happened.

Also, perhaps the card has some kind of RAM memory, that gets initialized and if corrupted will show these symptoms. Just rebooting doesn't fix it, because the contents of the RAM are retained. You need to switch off the computer to wipe the RAM contents and when you start again, the driver initializes the card's RAM correctly again.

At least that would be my (optimistic) interpretation of what happened...

 

[CrazyCat]

no any RAM 6983, 6903/5/8 cards have dual tuner/demod chips and each streamreader instance reinit demod on start.

P.S. 6903x/6909x have FBC tuner/demod chip STiD135

 

[vma]

Then what could have happened? Is there any risk of breaking the TBS6903 when driving two DiSEqC-1.2 motors simultaneously?
Also, why did the card only get to normal state after switching off the computer, instead of a simple a reboot?

 

[deeptho]

Then what could have happened? Is there any risk of breaking the TBS6903 when driving two DiSEqC-1.2 motors simultaneously?
Also, why did the card only get to normal state after switching off the computer, instead of a simple a reboot?

Probably the risk is low to none. The main challenge for the card is when the motor is powering up, and this requires large currents,
which the card must be able to handle. It is unlikely that the two dishes start moving at exactly the same time.

As others have said: These problems are often software problems (drivers, firmware...). It is also possible
that your dish was still pointing in the wrong position and then moved with a new diseqc command.

Note that the cards have an extra wire that you can connect to the power supply of your computer to provide more juice
directly, instead of through the pic-express connector

 

[vma]

I have the extra wire connected and tuner 2 is connected to a disewc switch. This got corrupted and was not fixed by a simple reboot or drive installation. I had to switch off the computer.

After switching it off for two minutes and rebooting the card was again working normally.

 

[deeptho]

I have the extra wire connected and tuner 2 is connected to a disewc switch. This got corrupted and was not fixed by a simple reboot or drive installation. I had to switch off the computer.

After switching it off for two minutes and rebooting the card was again working normally.

Yes: that is a bug in the firmware or some hardware glitch (less likely). All it takes is writing
wrong data in a chip register, e.g., because an address bit gets flipped.

 

[ptergrand]

Does anyone know why 2 tuners are not working?