[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [bluetooth] RF (our interface)



> > [Ling] Yeah, there are lots of registers, which
> > might be useful for upper
> > layer, but usually not. The RF register provide most
> > of the information for
> > tweaking RF performance. Philsar(Connexant)'s RF has
> > 60 registers, Ericsson
> > has around 20, as far as I know, most of them is
> > unrelevant to upper layer
> > unless upper layer want to show something for
> > debugging or showing status
> > purpose.
>
> So you suggest to let the Baseband control it? But I
> am afraid that if pure hardware control on these
> registers will be done in hardware will not beable to
> support different RF phys? at least we can not
> gurantee that two RF chips have the same address for
> the same registers. So we can just define signals for
> the common registers and bits between all RF devices,
> make the RF bridge map them to the correct address and
> the reset will be controlled by software.

[Ling] Of course, u r right, we needn't let pure hardware control those
registers, we should abstract the signal for commone registers, for JTAG or
SPI interface, the regsiter controll usually need following information,
(1). Address (2). data, (3). direction (r/w), then we can start SPI or JTAG
interface to access the correpsonding interface. The upper layer handle
initialization, and performance tuning mostly. The RF bridge can be designed
based on the funcation model of the interface, pretty straight forward.

> > >
> > [Ling] Right, low power clock(32kHz) also provided
> > for standby purpose. We
> > need take it into account.
> OK lets go for three clocks 1MHz (Tx/RX clock), 32KHz
> (low power clock) and 13->20 MHz system clock.
> need to elaborate more on how to operate these clocks
> and their functionalities
>
> pls check the bluetooth site at opencores I am going
> to update it for each issue we agree upone

[Ling] I will look it up, thanks for your work!

Rgds,
-Ling

--
To unsubscribe from bluetooth mailing list please visit http://www.opencores.org/mailinglists.shtml