TBS6209SE packet loss issue identified (DVB-C)

[Delta]

I was excited about to use TBS6209SE to watch some cable.

TBS6205SE/TBS6209SE takes Sony "Dafa"'s CXD6802GL integrated tuning/demod chip which will theoretically deliver better results than the previous generation TBS cards.

However it suffered from a strange packet loss issue that does not happen to carrier-issued DVB-C set-top boxes on the same line.

Tried reinstalling drivers and changing PC/OS. No avail.

Tortured carrier's customer service as I believe my signal strength is substandard.

Overturned the premises to locate the source of interference. Found that when DOCSIS Cable Modem (CM) powers on, DVB-C is f**ked. If CM is off, DVB-C is OK. (This means that CM must be turned off to avoid DVB-C packet loss when using TBS6209SE.)

But I don't want my internet to pass away while watching cable. I want them all safe and sound.

Even contacted government regulators to urge carrier address the cable interference.

They did what I asked for. But little has changed.

Read posts on the forum and got inspired. There's a built-in signal amplifier behind the "RF IN" port (on the card).

Okay, so there is a signal amplifier.

IT AMPLIFIES NOISES!

IT AMPLIFIES NOISES!!

IT AMPLIFIES NOISES!!!

What about to brainlessly switch the RF cable to the only other "RF OUT" port?

It works. It just works.

The signal strength is so weak. But that's okay. We can find a BYO (external) amplifier.

Two weeks of effort for such a hopeless reason. But I can now tamper whatever I want with the cable.

And I hope this helps users who are also bothered by the same (TBS6209SE card packet loss) issue. (The problem is only relevant to electromagnetic wave physics and has nothing to do with software factors such as kernel parameters and drivers.)


Principle explained:

Our RF premises line has a certain signal interference and low signal strength. And I believe this is a common problem for many cable TV users.

Whether it is the operation of STB or CM equipment, small noise interference will be generated. And more terminals lead to greater interference. But these will theoretically not affect the operation of other terminals.

TBS6209SE has an 8-channel tuning function and is designed with signal amplification at the RF IN port, but this is not an ideal solution to resolve signal attenuation. In some cases it helps the user, but in extreme cases it only amplifies the noise to the level that affects normal operation of the tuner. This is what we often call the two sides of things and that absolutely applies to the amplification design of TBS6209SE.

The RF OUT port is working like a signal tap port on the card. Signals passing through it are greatly attenuated, no matter which direction they come from. But this also avoids the amplifier. Therefore, under the premise that the signal strength becomes extremely weak, the packet loss problem caused by noise is also avoided. And now you only need to deal with the SNR problem, which can be simply addressed by an external signal amplifier.

[Delta]

I do wonder why does the operation of DOCSIS CM causes severe signal interference to TBS6209SE.

Our cable system uses the DOCSIS 3.0 protocol for Internet access, which will lock into 8 downlink (DL) frequencies and 2 uplink (UL) frequencies.

Under normal circumstances, the UL transmit power of this CM on our line is about 50 dBmV. During carrier's debugging process, the CM's UL transmit power was increased to 53 dBmV, which resulted in significant increase of TBS6209SE packet loss. This fully demonstrates that the uplink transmission function of CM is a critical cause of interference to TBS6209SE's reception of DVB-C signals. At the same time, the DVB-C set-top boxes on the same line did not experience any packet loss issues.

By consulting carrier professionals, we learned that our uplink frequencies (43 MHz and 50 MHz) are the two frequencies most likely to interfere with line signals, but theoretically should not cause interference to the DVB-C downlink.

The built-in signal amplifier in TBS6209SE may have been designed without considering the interference factors of reverse transmission channels. (Especially with DVB-T, there is no chance of a reverse transmission channel. So it amplifies the signal any way it wants.) It does not effectively suppress out-of-band signals, causing UL signals very close to the transmitter (CM) being amplified, thereby resulting interference.

[crazycat]

Best solution switchable LNA before tuner ICs, some like old PCTV 290e, 292e USB sticks (via GPIO pin on demod or usb controller). LNA enabled for terrestrial reception and disabled for cable.