triggered function generator circuit

Simple enough, I could even do it with discrete transistors. The triangle would be a little more complicated but nothing too hard.

..Or you can also use a chip, which works basically the same way.

It would be nicer if you had the instantaneous CV output for the squarewave generator, or any sort of prior signal source before just a generic squarewave. Otherwise you have to frequency > voltage convert it, then a square/triangle VCO or something.

Is the output clocked to the input square wave, or what is the function of input to output frequency?

Tim

-- Deep Fryer: a very philosophical monk. Website:

This is a music circuit. The square waves are coming out of a top octave divider which keeps all the notes in relative tune. I want to turn these into triangle/sawtooth and variable duty cycle pulses with the slope/duty cycle being controlled by a pot. The input and output frequencies have to be identical. I've been looking at the MAX038 function generator which has a phase detector input. I think it can be configured as a PLL that will synch the chip's oscillator to an input square wave from the top octave divider. The chip can produce all the waveforms I want with duty cycle adjustment I need. I don't know much about PLLs and I wonder if the chip will be able to capture and lock on to the wide range of frequencies I'm going to send to it. Maybe not.

May I ask a stupid question, why do you want to do it. Is that like a school assignment, that is immutable? If so, then I have no suggestions, but if not, then maybe there are other approaches.

I read in sci.electronics.design that snipped-for-privacy@prodigy.net wrote (in ) about 'triggered function generator circuit', on Fri, 4 Nov 2005:

The way to do a sawtooth with variable tooth slopes is to charge a capacitor from one constant current source and then discharge it through another current source. Switch sources with your square wave and two CMOS analogue switches. Adjust the currents to change the slopes.

I'm thinking of a fast clock, some kind of programmable logic, and one or more DAC's (with suitable low-pass filters). I guess you would also need an ADC to convert the pot control voltages to digital form.

Maybe that is too expensive or too complicated for your application?

Just a thought.

--Mac

We will be wondering too if you don't say what the "wide range of frequencies" is exactly. Following the app note, it can be seen that the PDO->Fadj PLL allows a lock range of +/-70% about center frequency, or

A to G# ~1:1.9 frequency range