Variable Phase, Frequency and Amplitude Signal Generator

This sounds doable, but it depends on the precision required. That is: with what precision do you need to know the frequency? Does the frequency need to be continuously variable, or would step increases be acceptable? If so, what size steps?

How accurately does the sine wave have to follow a true sine curve? There are several circuits that approximate a sine wave in steps, then smooth those steps with RC or other networks.

The frequency outputs are the same for both generators and the variation can be at fixed values:- 1Hz, 10Hz, 100Hz,1000Hz, 1666Hz. I can only gauge the presision in terms of frequency from the calibration test results given which puts it at +/- 0.08Hz. Amplitude for the sine wave needs to be between +/- 2mV RMS. Looking at these figures I would expect that the sine wave needs to be fairly accurate in line with the quality of signal produced by a Wavetek 186 signal generator. In theory the tests run at 1666Hz are not so important as they are at the top end of the machines sensitivity. Hope this gives some idea of the requirements.

Many thanks

James

A good basic starting point for electronics is http://www.allaboutcircuits.com/, although this might be more basic than what you are looking for.

I think you should leave the "classical" way of signal generation. You should use a microcontroller to generate the sine wave with PWM sequences. It can provide you the the exact phase shifts, the square waves, maybe trigger pulses for oscilloscope.You can learn about the principle here:

http://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1058&context=eesp

Use a crystal oscillator to drive a programmable divider; use the output of the divider to drive a counter; use the output of the counter to drive the inputs to two PROMs; use the outputs of the PROMs to drive D to As. You can use the output of the D to A directly or use it to drive a variable gain amplifier depending on the specific accuracy needed for the second signal.

Clearly you could simplify the design of the first generator quite a bit: but I think I'd be inclined to make them both the same so that you end up with a more general purpose function generator.

Use a spread sheet to create the files for programming the PROMS.

You may also like to use a small PROM to drive the load inputs of the programmable divider.

I have a Tektronix AFG series Dual Channel function generator that will do what you need. we use LabView to control it but if I remember it included control software. The version we use does a good job at low frequencies and our Metrology lab calibrates it in house.

We deal with several distributors who can get us loaners to evaluate and there is a free demo version of LabView and Signal Express.

Look for a chip with Google like say the ICL8038, which is a bit old, but still around.

Search under "function generator IC"......I found nearly 160,000 hits......

Dear All

I have moved on a bit since my last input and have decided, mainly due to my lack of indepth knowledge with electronics, to take a more pragmatic approach and set up the building blocks to provide the necessary signals and monitoring. So I have a variable signal generator which will give me two channels, one for the square wave reference signal and one for the sine wave with variable frequency and amplitude. I can carefully set the signal levels with a frequency counter and DVM. The last remaining part of the puzzle is simply the phase setting of the sine wave relative to the reference signal. I still need to be able to shift the wave to 90 degrees and 270. I can obtain the 180 shift by reversing the signal input so that is no problem. The simplest(?) solution to the remaining phase shifts would seem to be a C-R circuit but I would ask whether people think this practical?

Trying to delay one signal with respect to the other using a CR network does not sound very practical, especially as you will need to vary the delay dependant on the frequency.

I think you're better off buying one of these function generators from Maplin

http://www.maplin.co.uk/media/pdfs/N42FL.pdf

for £13 and getting the development software for the micro controller on it so that you can figure out how to tap off the square wave you need.