ti-89 wireless link

[jacuinde]

im trying to build a wireless link, so we can use it for educational purposes heh, i see that the I/O port isnt a RS-232 one like i thought, so im going to build this schematic
Vdd
---
|
< 10 Kohms
>
|
V Diode
|
out--O---------- input
|
< 470 ohms
>
|
NPN \__/\/\____ output
/ 33 Kohms
|
__|___
Gnd
any help welcome, the purpose of this is using cheap 433mhz transmitters, i already have all the stuff to transmit data with that, im just stuck at the ti protocol things

[jacuinde]

guess no one gives a dam about this, i have made progress with the protocol information i read from ticalc, any help welcome, i just need something to see im sending the right stuff to the ti-89

[Liazon]

idk, much about hardware, but Kermmartian on Cemetech studies EE so he might be able to help. He's more of a z80 calc guy, but I'm sure he can help.

[benryves]

The most efficient way of doing this would be to use a pair of microcontrollers on each end that act as a bridge between the calculator's own protocol and a protocol that is suitable for radio transmission with your transmitter/receiver pairs (low speed serial)?

[DigiTan]

Yeah definitely check into a micro controller. The most basic of basic of one-way radio can be done using a $6 state machine like shown below. The receiver end will have about double the parts, and even more to build a transceiver.

It really all depends on the RF equipment you bought. The picture below generates a self-clocking transmission, but the downside is your transmitter needs to have at least x8 the max bitrate of the TI data. (x2 for clocking, another x2 because there's 2 lines to encode, another x2 for sampling reasons). A micro controller could do all this in less space and work with slower transmitters. The big questions so far are:

• What's your budget?
  
• Can you acquire microcontrollers?
  
• Are you building it for your own custom protocol or for TI's? (For TI means more hardware requirements)
  
• What are the bitrates for your RF devices?
  
• Will both your transmitter-receiver pairs operate at the same frequency? (Same frequency is slower)
  

[Liazon]

Digitan you've got skillz that's for sure. Mind summing up what that diagram means?

[DigiTan]

No prob. Going from left the right, you have your TI outputs. You can wire both of those to the D input of a D flip-flop. Both D flip-flops also have a clock signal going into them that causes them to reject any TI output changes until proper time. The recommended rate is x2 or more the TI's max bit rate. Maybe even x3, x4 or x5 to be really safe.

Right of that are two XOR circuits--one for both wires. The same clock also goes into these. Any time you XOR data with a clock, it generates a format called Manchester that carries the clock signal with it.

The last part is a selector/multiplexer. It flips back and forth: carrying the tip's signal half the time, and other half carrying the ring's signal. The flipping back in forth is controlled by a clock that runs at half the previous clock's rate.

When it's all said and done...

If the TI sends: tip = high, ring = high. The output goes...
0101 (repeating)

If the TI sends: tip = low, ring = low. The output goes...
1010 (repeating)

If the TI sends: tip = high, ring = low. The output goes...
0110 (repeating)

If the TI sends: tip = low, ring = high. The output goes...
1001 (repeating)

And these four bits repeat until the TI data changes. The useful part is that the 1st and 2nd bits are always mis-matched no matter what. Same for the 3rd and 4th. That effect is caused by the clock from earlier, and the receiver can detect that. The main disadvantage is the low speed and size. Also, the receiver can detect the clock just fine; but it can't tell the tip's bit from the ring's bits (custom software would have to do that).