Michael Buesch wrote:
> I analyzed the bouncing of the estimated TX power
> output versus the desired power.
> 
> Here's the hacky script to parse dmesg.
> It takes a number as arg. It's the PHY number.
> It requires printk times to be enabled in kconfig.
> 
> Looks pretty good. Though, it begins to bounce at the end.
> Not sure what happened there.

I hadn't done any plots, but I see the same kind of steady values followed by a 
seeking that lasts 
quite a while. At first, I thought the convergence needed damping, but that 
only made it worse. That 
4:1 ratio in the effects of the two parameters is hard to track. The 
fluctuations at the end may be 
due to changes in the temperature of the chip, causing variations in the output 
power for a given 
set of attenuations.

Larry
_______________________________________________
Bcm43xx-dev mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Reply via email to