THE SPEAKING COUNTER
The speaking counter uses 8 bytes of storage to contain 8 binary-coded decimal (BCD) numbers with a capacity of 99,999,999. The code for maintaining this counter is fully in place, but the logic for playing the numeric messages only foes up to 999 at the present time. I’ve had it count up to 275 so far. There is still work to be done on the delays and connecting between the messages involved.
“one” “hundred” “seventy” “one” has 4 messages for 171.
PIC PROGRAMMER 100 TIMES FASTER
To program the PICs, I use a Qbasic language pgm of my own, and an older PC laptop. The pgm sends control signals via the printer port and some TTL buffering to the PIC chip. The main signals are the clock and data lines which contain the
programming commands and the data of the PIC assembler pgm to be loaded.
I encountered a difficulty in that I was forced to enter one instruction, then to go out of programming mode and back in for the next instruction. I believe I was not allowing enough time (10 msec) after the “program one cycle” commands.
Well I made some changes and included the necessary delay and now I can just stay in programming mode.
DOS BATCH FILE COMPLETELY AUTOMATIC
I discovered that ther is a Basic command called SYSTEM. It closes all open files and returns control to the operating system. Before I used the command STOP.
But for this I had to intervene to get back into DOS. Now the batch file is completely automatic.

When you are in the heat of a project, you don’t want to waste time cleaning up, so you just push stuff aside. But finally you get to a point where you can’t even move around in your shop, there’s so much stuff on the floor. Then you have to make a tactical retreat and put everything back where they belong. That’s where I am now and it’s really helping me with my project.

I have been having a lot of fun placing boxes and things
around the floor and seeing how well SENSBOT does.
Sometimes it gets itself stuck. But it does very well
and sometimes great when it goes into a dead-end hall area
anbd manages to turn around and come out again all
by itself.
SENSBOT has a rotating turret driven by a stepping motor
controlled from the computer program in the PIC microcontroller.
In general my turret sweeps 90 degrees, and I have settled on
5 segmnents of detection. Extreme left and extreme right,
slightly left and slightly right, and direct ahead. My sensor is the
Sharp GP2Y0D340K. This infrared sensor is set to activate a
signal when object is less than 40 centimeters (about 15 inches?)
from sensor. I bought 4 of these sensors from junun.org at $6
each. I use only one in the rotating turret.
My program algorithm is as follows.
If object sensed on extreme left, stop right wheel for brief time.
If object sensed on extreme right, stop left wheel for brief time.
If object sensed slightly to left, stop right wheel for longer time.
If object sensed slightly to right, stop left wheel for longer time.
If object is directly ahead, I alternate between a hard right
and a hard left.
This robot is not set up to go in reverse. The PIC sends signals
to 2 transistors which control the on and off to the wheel motors.
The sensors are quite complex. An infrared signal is sent out
at some specified freequency, and a phototransistor or photodiode
detects signal. The receiver though will filter out any other signal
that is not at the specified frequency. This prevents false signals
from other sources.

Thanks to Bob Neidorff of NEMES for help with
293 H-bridges which worked well with
my DC motor control. However things did not go smoothly until
I put in the diodes to curb the spikes
from the motor coils, etc.

I am coming to grips with problems of torque. The wave pattern
is the easiest to drive a stepper. These are unipolar stepper
motors. So I went to a half-step pattern which should give me
more torque – I hope. There is still another pattern that I can look into.
With the DC motors there seems to be no problem with torque.
Stepper motors for locomotion do have some advantages.
To reverse direction, I just send the patterns in the opposite
sequence. To stop, I just stop sending the signals.
An example of wave pattern is as follows:
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
Each of the one’s represents energizing one of the coils
in the stepper motor. You just repeat sending this pattern over and over
and the motor steps ahead. The pattern for the half-step
is more complicated, but basically energizes 2 coils
on some of the patterns.
Of course, I am sending these patterns to 2 stepper motors,
so I can make one stop or go in reverse while the other
goes forward.

A program called INBTWEEN.BAS takes the hex file produced by my ASSEMBLER and prodeuces a file of 16 character strings, all containing 1’s & 0’s.
Ther second program WORKRND2.BAS reads the string file and sends this data along with the necessary programming COMMANDS to the PIC. Note that I still have to enter & leave programming mode each time I write an instruction to the PIC. But I have cleaned up the code and the processing. I have a variable I use for delay loops. Up to now I have been using 1000, but I have found by experiment that the program works at 100, which has speeded things up tremendously

The program uses the timer0 with an interrupt. The pre-scaler is set for 1:256.
My interrupt routine subtracts from a register everytime there is an interrupt. When my register is zero, I make my turn. I get about 8 to 10 seconds delay.