Circuit : Rodney Byne, UK
Email :

Description
The multifunction generator in this circuit, is capable of producing 7 different waveforms, sine, square, triangle, positive pulse, negative pulse, positive ramp and negative ramp waveforms.
Widerange Multifunction generator
This hobby prototype was constructed out of curiosity after researching the internet to find out what types of non-commercial or home made function generator circuits were available. In the main, usually three-waveform versions predominate. Square, triangle and sine.

It appeared that my aim of including sawtooth waveform generation was lowish on the public priority need, so I looked up the document offerings from Ray Marston of the US. His electronics practical experience is renowned.

In his op-amp cookbook series Nuts and Volts, I noted in part three, the figure 17 description 100Hz-1kHz ramp/rectangle generator with variable slope-M/S ratio, was an eye-catching multifunction generator and suited my experimenting purposes perfectly.

To my surprise, this particular design offered more variety of waveforms than I realised until I had built my prototype to explore Ray's design capabilities. In fact I use two of his circuits together, which generate no less than seven waveforms. This article is the result of my findings.

Build notes
Audio oscillators come as single supply and split supply versions. Mine is split supply, so for the convenience of portability, I use two 9 volt PP3 batteries as +9V and -9V. This total 18v arrangement with the common junction as 0V technical ground, in my view offers greater waveform amplitude headroom before clipping or saturation.

As per the photo, my finished test jig assembly is not a traditional circuit breadboard, but otherwise constructed in a workmanlike makeshift fashion. It uses a wooden base material, two secured springy toolclips to hold the batteries and a small veroboard sub-chassis mounted on a paxolin tagboard for the two main IC bases and associated components which form the core circuit. There is also a small extra veroboard, housing a lamp-stabilised sinewave oscillator secured to the base.

The split-supply power rails are parallelled to the two circuits and the source +/- battery leads are switched on and off together as per the system drawing.

The finished assembly uses six multiturn pre-set pots which should be wired the right way round. By convention, the fully anticlockwise end where the adjustment screw is, should face 0v. This ensures that when approaching signal maximum, amplitude is in the clockwise direction like a volume control.

In order to absorb any external vibration causing microphony to the stabilising lamp filament through its connecting wires, as per the photo it is suggested to lay the lamp down on a small cushion of thin sponge on the pcb.

Bend at right angles the two thin wires from out of the lamp base and pass them through two tiny holes made in the sponge, before soldering them to the pcb.
Use a bit of black tape to firmly hold the exposed lamp legs to the upper surface of the sponge. Output level stability of the sine wave should so be unaffected by vibration and stay constant after unit testing.

Circuit notes & testing results
The frequency range is approx 45hz to just over 1khz.

To reduce battery drain currents, the two ICs 741 shown were replaced by TLO71. Approx battery currents: +9v = 11mA, -9v = 12mA

In the absence of a lab bench oscilloscope for testing, I use software on a laptop. This is a satisfactory substitute but it must be remembered that the maximum signal into the microphone socket shouldn't exceed 100mV for safety. I recommend throughout 77.5mv RMS which is -20db also known as "neg twenty" when set by each amplitude pre-set and viewed on the spectrum analyser part of the program.

On slow clockwise adjustment of the variable slope-Mark/Space ratio pre-set, the following waveforms were observed in this order;

Waveforms

Firstly at the ramp output; negative going slope sawtooth, changing to uniform triangle, changing to positive going slope sawtooth.
Secondly at the rectangle output; positive going pulse, changing to a 1:1 M/S squarewave, changing to negative going pulse.
Thirdly at the sine board output; a single frequency with prompt startup and stable 1khz f1 sine wave, having its f2 rel harmonic at -55dB and f3 rel harmonic at -65dB.
It is possible to fine tune f2 distortion down to rel -60db, but the sine oscillator startup behaves erratically and is unpredictable to settle consistently at the same level.

After the individual control operations are observed, this completed unit should form a good analogue waveform learning experience.