Re: TI-H: RF Linking


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

Re: TI-H: RF Linking



Matthew H. Fogle wrote:

> I'e been thinking about what has been said recently and I agree with you
> I think. I have thought up a design that would use one freq and yes that
> will solve a lot of problems but then it will HAVE to have a software
> driver for it. The only way to make it compatible with old link cables
> is to use two freqs because of the async connection. But could we make a
> TSR that would convert this non-ti protocol to TI protocol as far as the
> calc knew I wonder?

Sorry. TSR's on the 8x calcs can only be "hooked" to the clock interrupt
which has a frequency of about 200 Hz (I clocked mine to 185 Hz). This
means the interrupt routine is called at most 200 times a second. This is
nowhere enough to translate the transmission.

I think there are only two reasonable options: use a microcontroller or two
distinct frequencies. Both have their pros and cons. Microcontrollers take
up quite little space and I wouldn't be surprised if there was a
microcontroller on the market that already had most of the needed RF
circuitry built-in (at least IR ones exist). The MCU would run a piece of
code which encoded the data about the status of the link port wires in some
way and the MCU at the other end would decode this. The MCU's would sample
the link port at given intervals, or perhaps only respond to changes via
interrupts. Error correction etc. would only be a matter of software (or
actually firmware). I've never used microcontrollers so I really don't know
them very well, however I'm 100% sure that some sort of a RF link can be done
relatively easy with a MCU.

The other approach is to use discrete components and probably two
frequencies. This is maybe simpler to design (for an electronics expert)
but probably takes up more space and possibly costs more.

-Ozone

--
*** Osma Suominen *** ozone@clinet.fi *** http://www.clinet.fi/~ozone ***


References: