THE ATARI PUNK CONSOLE
 aka The stepped tone generator by Forrest M Mims

The Stepped Tone Generator is a simple dual oscillator circuit with a single pulse wave output. The first oscillator drives the pitch of the output oscillator which divides the pitch by increasingly smaller amounts as the pitch increases depending on its own pulse width (steps) setting. The name "Atari punk console" came about as some builders say the sound resemble's that of early Atari games consoles.

Basic demo (no effects added)

The original circuit shown right - is pretty raw if R1 gets to low in resistance the circuit stops and its output only suitable for a 8ohm speaker but these little points can easily be sorted. The circuit is made up of two simple circuits ~ A Astable multi vibrator (oscillator) triggering a Mono-stable (one shot pulse generator) which acts as a frequency divider and pulse width modulator.
many others have tweaked this circuit giving it a line out etc some of the best are
Kaustic machines Apparently came up with the Atari Punk name
Ge-tLow-Fi A really good electronics / circuit bending site
Beavis Audio Some great little 555 / 556 timer projects

One of the drawbacks of the original circuit is is sucks a bit of current (10 - 20mA) which is not to much on its own but once you start to add LED's, modulators and other circuit bits your batteries don't last very long! The purpose of this page is to look at how the circuit works why it uses so much current and how I came to adapted the circuit for a low power consumption.(0.5 - 1.5mA instead of 20mA !) A schematic of this circuit is at ther bottom of the page.

 STGForrest M Mims III page 26 "Timer, Op Amp & Optoelectronic circuits and projects" Volume 1, Master publishing,Inc, 2004.
WWW.FORESTMIMS.COM
555 / 556 Timer basics

It is really not necessary to under stand how the 555 / 556 timer works in order to make the stepped tone generator but its worth having to understanding if you want to modify the circuit. Basically a 556 timer is a dual timer, it has to equiverlant 555 timers in one chip package IE you can make the circuit with 2 555 timers or one dual 556 timer SEE HERE for an example.
HERE is a good expanation of how the timer's work. Below is an expanatiove of how the Stepped tone generator works scroll down for my 2 revised APC circuits.		 556 pinouts
The first stage Astable oscillator

When the output is high the voltage through R1, VR1, R2 charges C2 until a voltage of 2/3rds of the supply is across the threshold pin 2, then the output goes low, C2 now discharges through pin 1 (discharge). When the voltage across the threshold goes below 1/3rd supply voltage because the trigger pin6 is conected to the theshold pin the output now goes high and the cycle start again and again IE its self oscillates.

The crucial part of this for the stepped tone generator is that VR1 controls only the positive pulse time, High resisatance = long positve pulse and therefore lower pitch and low resisance = short postive pulse IE high pitch.

The down time of the pulse is constant and short its this that trigger's the mono stable pulse output. Long negative pulses can make the mono stable jump octave more suddenly and with pulse width change a crucial part of the tonal character its good to make it short.
The Mono-stable Output stage

The mono stable output of the stepped tone generator only produces a pulse when its receives a negative going trigger at its trigger input (pin 8).

A negative going trigger at pin 8 causes the output to go high The voltage through VR2 and R3 then starts to charge up C3. When the charge on C3 reaches 2/3rds the supply voltage the threshold triggers setting the output low C3 now discharges through R3 into pin 13 (discharge) which connects to ground when the output is low. The output stays low until a new negative going pulse is received at the trigger pin8 .

While the output is high the circuit ignores triggers at its input until the output goes low. This is how it divides the trigger signal. VR2 controls the pulse length by changing the time it takes to charge up C3
The stepped output follows the frequency of the astable oscillator when VR2 resistance is 0
As the resistance of VR2 increases the output pulse gets longer and the tone changes but the frequency is the same.
As the resistance of VR2 is increased further the pulse width is longer than the frequency and the output steps to half the frequency of the astable oscillator.
A major characteristic of the APC is it divides higher frequency more than lower frequencies at the same setting the pulse width also changes.
   
The audio demo below is of my circuit using the original NE555 bipolar timer. The sound is a little richer - fatter to my ears Above is my 1st revision of the original circuit its offers nothing really new its just my way of doing it and produces the tones i like. I have added a resistor (R1) in line with VR1 to limit the charge on C2 so the sound never cuts out. I use a 33nf cap for the frequency of the Mono~stable stage it make it step down in octave less at high frequencies.R3 restricts the current draw of the Mono~stable stage. R4 and R5 form a potencial divider to give a safe line output slightly lower than other design's but it works better for me (never clips my sound card input.)

The original circuit uses a NE556 timer which uses allot of current i measured 10 - 20mA on my circuit. The circuit draws more current when VR2 is at its lowest resistance IE the shortest pulse (devides the astable frequiency by 1). I use a TS555 timer which draws 1.2 - 8mA again more when VR2 is low resistance. Below is a circuit that overcomes this problem.
The audio demo below is of my circuit using a TS556 low power timer. produces the same effect but the click's are a little louder as it jumps octaves
atari punk stepped tone generator

The astable oscillator only draws 0.3mA on its owns which is very low so its the monostable circuit that draws the most current (about 7.7mA at the shortest pulse width). The problem is how to get the Monostable to divide the astable frequency by 1 without drawing a lot more current, one way is to use a 20k resistor in series with VR2 and uses a low value capacitor for C3 lets says 470pf which sort of works but then the monostable can only produce a short pulse-width and does not divide mid-low frequencies at all and sound's very different.

I have got round this problem by having a fixed charge on C3 through a 680k resistor (R3) which on its own produces a large pulse-width, to have control over the pulse-width i have used the voltage control pin 11.

The voltage control pin affects the internal reference voltage of the timer. R6 (10k) and Q1 -pnp transistor apply a bias voltage to the voltage control pin VR2 (which is now a log pot) is connected to ground at 0ohms resistance the pulse output is at it shortest (divides the astable input frequency by 1) and its current consumption is 1.5mA.
When VR2 is at maximum resistance the pulse-width is at its longest and the current consumption drops to 0.5mA (tested on a bread board)

There is a little down side - its shortest pulse is not as short as the original so it can not following the frequency of high pitches on my test circuit i measured a highest frequency of 3.3Khz were as the original can go up to 7khz in my tests. and so im not fully there yet but i have low power and the same tone ! pretty good so far im guessing using a differant transistor (i only have 2n3906's at itme of writing) thansk for reading and happy building.

 555 internal diagram
A demo of the low powered circuit
Pretty much the same to my ears ! and batteries last a lot longer

I hope to develop the circuit more but this is it for now More information on the 555 / 556 timers can be found HERE and HERE
Web Design Matt Newlove