programmer, Using Apple II for frequency generator

Hello group,

I want to program a loop to generate varying frequencies from
1hz to 500khz, out the gameport. This is to test the resonant
frequency of a Tesla coil...

The tightest loop I can think of:

loop STA gameport
add JMP LOOP


to decrease frequency, have a BASIC program
insert a NOP, JMP LOOP at add....

loop STA gameport
add NOP
JMP LOOP

then move add up by one byte:

loop STA gameport
add NOP
JMP LOOP

using stock 6502 (1mhz) what is the highest
frequency I can generate? What is the formula
to convert the above codes output to Khz?

Rich

STA absolute takes 4 cycles.
JMP absolute takes 3 cycles.

The 6502 operates at (I think) 1.024 MHz, yielding 146285 repetitions
per second, or 146KHz.

You can improve this by doing:

loop STA gameport
STA gameport
STA gameport
[...]
JMP loop

With enough STAs, the JMP goes into the noise, and you get 256KHz.

--
Send mail to XXXX@XXXXX.COM (Andy McFadden) - http://www.fadden.com/
CD-Recordable FAQ - http://www.cdrfaq.org/
CiderPress Apple II archive utility for Windows - http://www.faddensoft.com/
Fight Internet Spam - http://spam.abuse.net/spam/ & http ://spamcop.net/

Thanks.... I'll have to experiment with the Transwarped IIe to see if I
can
break the 256khz barrier... I wonder if the TWIIe slows down for
gameport
access... I can check with my frequency meter, and the miniassembler
to enter code.

R

It shouldn't.

-michael

Parallel computing for 8-bit Apple II's!
Home page: http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it is seriously underused."

I've seen some people use the BIT opcode when messing with the speaker
-- probably used to preserve the value of the accumulator. It also
takes 4 cycles. Not sure if that's worth anything. Here's a way to
shave off a cycle using a relative branch (2 cycles unless you jump a
page boundary)

start: clc
loop: sta gameport
bcc loop
which is then 6 cycles instead of 7. If the previous version yieded
146khz, this should make it 170khz.

I believe branch instructions require 2 cycles if the branch is not taken,
3 if it is taken, or 4 if the branch is taken across a page boundary.

http://www.6502.org/tutorials/6502opcodes.html

--
Send mail to XXXX@XXXXX.COM (Andy McFadden) - http://www.fadden.com/
CD-Recordable FAQ - http://www.cdrfaq.org/
CiderPress Apple II archive utility for Windows - http://www.faddensoft.com/
Fight Internet Spam - http://spam.abuse.net/spam/ & http ://spamcop.net/

> I believe branch instructions require 2 cycles if the branch is not taken,

Whoops... silly me. I misread my own reference.

http://www.obelisk.demon.co.uk/65C02/reference.html #BCC

Sorry about the confusion there.

-B

...


Seems like you may want to run an external circuit with maybe a 555
timer that you can trigger and adjust from the II. I am certain someone
here has access to aproject book that may have such a ciecuit. Just
remember to isolate your power sources and measuring equipment, a tsla
coil can produce a large voltage spike (not to mention current) that may
be unhealthy for your Apple II.

Thankx,
Ed

Perhaps a card with an 6522 could be an alternative.
I'm fairly sure that Rich has some lying around.

Use PB7 as the output, program timer 1 to count down as fast as you
like and set it up for continuous operation. This will output a square
wave.

IIRC the timers operate at system clock rate so that should be fast
enough. Better check the data sheets, however, as my memory is a bit
hazy ;o)

bye
Marcus

BIT messes with the status bits, while STA leaves everything alone.

Any absolute reference is going to be a 4-cycle op (3 to fetch, + 1).

You can't jump *anywhere* with a 6502 in less than 3 cycles, regardless
of the opcode.

-michael






--

-michael

Parallel computing for 8-bit Apple II's!
Home page: http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it is seriously underused."

You remember correctly.

A 6522 clocked at 1MHz can generate a 500kHz square wave.

The problem with using anthing clocked at 1MHz to generate frequencies
this high is that your frequency resolution is *terrible*. The highest
frequency you can generate is 500kHz, the next highest is 333kHz, the
next highest is 250kHz, etc. See the problem?

Much better to build a little tuned circuit oscillator which will have
infinite resolution and can produce a sinewave output. A 3:1 tunable
range shouldn't be hard to do.

Using an Apple as an arbitrary frequency generator gets more interesting
at lower frequencies (like audio).

-michael

Parallel computing for 8-bit Apple II's!
Home page: http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it is seriously underused."

This "could" be a problem. Rich will know if ;-)


Of course you are right. But one would have to build something first.


I bet the designers of the 6522 had audio output in their minds, too.
I re-read the datasheet today and got the impression the timers are
seldomly used to their full potential.
And I doubt that they were often used for such high frequencies.

bye
Marcus

He wanted to use it to "tune" a Tesla coil--he will want to use
a continuously variable frequency to do that.


Sometimes you have to do that...


They also included the ability to rotate a "serial" register
under timer control, allowing some control of harmonics.

-michael

Parallel computing for 8-bit Apple II's!
Home page: http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it is seriously underused."

The serial shift register is known to be buggy however, at least in
early 6522s. That's why the Commodore VIC 20 (and in consequence the
Commodore 64) have such a god awful slow serial bus if you don't run a
fastloader.

--
Linards Ticmanis

I wouldn't expect much problem from a circular shift register. ;-)

I presume the problem is with the handshaking related to the shift
register function...or is it?

-michael

Parallel computing for 8-bit Apple II's!
Home page: http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it is seriously underused."