On Tue, 31 Oct 2000, Scott Newell wrote:
> >Actually, no. Only the IrDA specification needs to be met. And it says
> something
> >to the effect that the pulse may be as short as ~2us (I think it is 1.4 or
> >1.6us) and as long as 1/4 bit time. So, if you make the pulses short, then
> the
> >same setup works for 9600 baud and 115.2kBaud.
>
> On the transmit side, sure. What about the IR pulses coming at the dongle?
>
>
> >No. I suspect you can control the pulse widths produced by a JetEye
> dongle. (I
> >personally never have tried). I just know from experience that it is
> >unnecessary.
>
> I still don't understand. Say I point an IR led at the JetEye. The led is
> pulsed once so that it emits for 1.41�s, meeting the IrDA spec. How long
> will the JetEye output a '0' bit to the PC serial port?
>
> Seems to me it's going to put out one bit to the PC, and that bit duration
> depends on the baud rate.
That's what I kept saying too, but it worked. We tried 9600, 38.4k and 115.2k
and they all worked. Your objection is similar to the one I made. "It can't
work!" But it did. I didn't measure the pulse widths in this instance. I just
used the sniffer. In another instance, I did measure the pulse widths and
noticed they were close to the minimum allowed. In this case, I did not look at
all of the baud rates. So perhaps what's happening is that the width of the
start pulse (remember, a logic `0' will generate a burst of IR) is decoded by
the receiveing JetEye dongle to generate the appropriate pulse width.
If I get a chance (I'm not working on IrDA at the moment) I'll measure these
pulse widths and try to determine what's happening.
Scott
_______________________________________________
Linux-IrDA mailing list - [EMAIL PROTECTED]
http://www.pasta.cs.UiT.No/mailman/listinfo/linux-irda
