antti@sistudio.com wrote:
>
> Hi
>
> If you want 4 Pins to output 62.5 Khz
>
> and at the same time 2 more pins to output adjustable frequency, than
> that might be a little bit tricky, if you need the 62.5 to be very
> stable (ie no jitter)
>
> also it important how many steps you need from 100Hz to 5KHz.
> .
>
> hm it is doable, yep the 4 outputs should be D0..D3 of on port to
> consume time, I think even at 4MHz clock you can run it fully as
> polled software loop ie no interrupts.
>
> You in loop until it is time for 62.5 KHz edge then you inc (port)
> and deceement a counter (for the variable frequency ouptut) when
> it overrolls you update the second var freq output, and reload counter.
>
> not giving 100% that it goes at 4MHz, but with very programming I suppose
> it doable at 8 or 10MHz clock you would not have any problems.
>
> I wrote a dual FM genrator for PIC, it runs two Frequency synthesizer
> algorihtms both up to 30 KHz, at 4MHZ PIC clock.
>
> antti
>
> pls ask if you need more help:)
>
> -- Silicon Studio Ltd.
> -- http://www.sistudio.com


Antti- appreciated.  You are the only one who gave me any concrete
answers to start out with.  Everyone is interested in plans, however, if
I can't simplify the circuit into a PIC, then no one will be able to
build the plans easily (which I will post anonymously at John Bedini's
site (http://www.nidlink.com/~john1/tesla.html)!  The chip need not be
ultra frequency stable, simply because the resonant frequency of the HOH
atom has memory and that it approaches its own resonance, again, you
would have to be looking at the patent to notice this.

rem, it should not be too bad

I will start out with some code I think will work.  Crawling before
walking:

from david tait, here is some short code that amused me

        LIST    P=16C84
        MOVLW   0
        TRIS    6
        OPTION
LOOP    SLEEP
        INCF    6,F
        GOTO    LOOP
        END

causes RB0-RB7 to increment 00000000-->11111111
which means we are halfway there, so to modify this statement, we only
need the fastest pin going at f, which we can tweak with a variable RC
clock, and we only need to tris 4-pins as 4-bits.  Rem 4 resistors sums
this, so


                LIST    P=16C84

                CLRF    PORTB
                BSF     STATUS, RP0
                MOVLW   0xC0            ; RB(6-7) as inputs  from pushbuttons
                                        ; RB(4-5) as the two other counted outpu
ts
                                        ; RB(0-3) as the 4-bit up count outputs
                MOVWF   TRISB
                OPTION
PUHARICH        INCF    6,0             ; this is where I have my question- how
to increment
only RB(0-3), and so I assume 6 is port B and 0 is for RB(0-3)  but this
is where the 4-bit count is satisfied for the fusion drive circuit, next
the Tesla Switch circuit pins RB(6-7)
                DECFSZ  reg,1           ; you said decrementing a counter,ok, so
 I think you
mean subtract until a carrier number decided by the bounceless switch
routine is met for toggling RB(6-7) at the lower frequency, I will
constantly subtract F from the reg and if this breaks to zero I it calls
routine Tesla for RB(6-7) which means I have to set the reg as the
carrier switch.  I think this keeps the 4-bit count sequence constant!
                GOTO    PUHARICH
TESLA           INCF    6,C             ; toggle RB6
                COMF    6,C             ; have to complement the RB7 from RB6 bu
t don't know how !
                END



Once the relationship between the registers is properly fixed, I then
need to revamp this whole set of code and put in the bounceless switch
routine to determine the reg variable.  Note that even though the reg
setting part from the bounceless switch may not work at first, the
fusion drive pins RB(0-3) should still kick, and there should be some
sort of output on the Tesla Switch pins RB(6-7) pins I hope.

You know, for a someone who has lost his instruction set assembly
knowledge, I think I can study up on this.  Apparently, this program
does not need a whole lot of subroutines.  I am doing this thing out of
compassion for everyone, not for money, for free. I suppose time waits
for no one, so that is why I will try and "wing it" in the here and now.

P.S. I ignored the 3 second delay on and off for the 4-bit sequence, but
I suppose it could be implementable later somehow, but it is not
extremely critical.