I've got a question on the following errata:
Initial data bits/clocks of the SSP transmission are shorter than
subsequent pulses at higher frequencies

There is no work around whereas there are some issues for frequencies
between 20MHz and 30MHz.
How can we handle these baudrates with this bug?
I do not understand how to deal with the consequences...

Thanks in adavance

Frederic Start a New Thread

--- In l..., "croquettegnu" wrote:
>
> I've got a question on the following errata:
> Initial data bits/clocks of the SSP transmission are shorter than
> subsequent pulses at higher frequencies
>
> There is no work around whereas there are some issues for frequencies
> between 20MHz and 30MHz.
> How can we handle these baudrates with this bug?
> I do not understand how to deal with the consequences...
>
> Thanks in adavance
>
> Frederic
>
Frederic,

let's assume the first pulse is 10ns shorter than e.g. 16.66 ns at 30
mbit/s. That would be a pulse of just 6.66ns. The device at the other
end, let's say a SPI Flash has to be able to deal with such a short
pulse. If yes, the Errata is of no consequence for you, if not, you
need to go slower until the pulse length matches the required time of
the device connected to the SSP.

Hth, Bob Start a New Thread

Thanks for this clarification Bob!
In my case, I'm just trying first of all to use the SSP in loop back
mode. Do you know if I will have some limitations?
Moreover, I did not succeed in finding some examples on SPI or SSP, do
you have ones?

--- In l..., "lpc2100_fan" wrote:
>
> --- In l..., "croquettegnu" wrote:
> >
> > I've got a question on the following errata:
> > Initial data bits/clocks of the SSP transmission are shorter than
> > subsequent pulses at higher frequencies
> >
> > There is no work around whereas there are some issues for frequencies
> > between 20MHz and 30MHz.
> > How can we handle these baudrates with this bug?
> > I do not understand how to deal with the consequences...
> >
> > Thanks in adavance
> >
> > Frederic
> >
> Frederic,
>
> let's assume the first pulse is 10ns shorter than e.g. 16.66 ns at 30
> mbit/s. That would be a pulse of just 6.66ns. The device at the other
> end, let's say a SPI Flash has to be able to deal with such a short
> pulse. If yes, the Errata is of no consequence for you, if not, you
> need to go slower until the pulse length matches the required time of
> the device connected to the SSP.
>
> Hth, Bob
> Start a New Thread