[linux-dvb] Preliminary support of Hauppauge Nova-T 500Dual(take2)

Steven Toth stoth at hauppauge.com
Wed Sep 20 02:31:41 CEST 2006

hermann pitton wrote:
> Hi,
> Am Dienstag, den 19.09.2006, 16:51 -0400 schrieb Steven Toth:
>>>> In the windows driver we leave the LNA off at all times, but have a registry
>>>> override to allow users to enable it. It's a single LNA for both tuners -
>>>> hence, it has its problems.
>>>> Switching the LNA dynamically in a TV app isn't something the API allows. How
>>>> about a module load param?
>>> As updating an LNA (or something else) is very board specific, I put 
>>> callback "update_lna" in dib3000mc_config. I will call this after starting 
>>> the AGCs to see if the LNA has to be turned on or not (taking the agc 
>>> response into account). For that I would like to know, what is the gain of 
>>> the LNA when turned on? (So that I can adjust the thresholds when to turn 
>>> on or off). Maybe we need to make an average of both channels, not that we 
>>> kill one channel by only taking the other one into account. 
>>> Or is a module_parameter enough? (I don't like this, because the majority 
>>> of users does not even now what is a module, which is good).
>> Image a user that lives somewhere between two broadcast towers. They 
>> receive some muxes at fairly weak levels and some very strong 
>> (environment depending). We never found a good automatic solution to 
>> deal with that within the (windows) driver.
>> If the driver can be made to work reliably for all users in Linux under 
>> all circumstances then that's ideal. :)
>> Otherwise, the user may need some way to dictate LNA behavior.
> looks like it is all because we are poor hackers ;)

On the contrary.

> With reliable data sheets signal strength and/or ber rates should
> indicate when to switch.

Last I looked Dibcom and Microtune hadn't released public datasheets for 
their parts.

If tuner 1 is active on a strong mux, enabling a single LNA for tuner 2 
(so a weak signal can be selected) may end up adding 10db to tuner 1 
leading to problems.

I'm simply saying be careful with implementing policy for a single LNA 
that effects multiple tuners, especially if the user has no choice about 
how the LNA is controlled.

> Looks similar like the situation on analog radio. Guess _nobody_ knows
> how to read auxiliary bits and recently a little more for all tuners
> around to get at least stereo/signal strength ...
A lot of datasheets are available for tuners. I don't know why this 
would be a problem if someone decided to implement it.



More information about the linux-dvb mailing list