DIY amp testing....

[markzb]

Hi guys.

As I have a pc based oscilloscope that I use for my work.....(fixing cars). I would be grateful if some of you could help me out in how to go about using it for testing amplifiers.

I gather that for starters..you need to inject a signal into the device you want to test...in my case a preamp and a power amp.

My first question is...do any of you know of a discrete high performance circuit that can be used as a signal generator?

The advantage I gather would be that you could tweak and adjust it with your scope making sure that the square or sine wave is symetrical before connecting it to the pre or power to be tested.

Does any such circuit exist? Anyone been there done that?

I recall from the old forum that some use a pc sound card using some freeware. What sort of symetrical precision do these produce? I am assuming here of course that we do want our generator to be as symetrical as possible! Hope thats a correct assumption!

So.....anyone got any ideas how to obtain a suitable means of signal generation?

Many thanks for reading

Mark

 

[Andrew L Weekes]

The standard technique for generating low distortion test signals (sine waves) is the Wein bridge oscillator.

A search on the web should bring up a few circuits and there's plenty in the semi-manufacturers data sheets too, if you can find the right ones.

For a good start, try here: -

http://sound.westhost.com/project22.htm

The Wein bridge needs some kind of amplitude stabilisation, and there's two ways to do it. The old-fashioned and very effective way is to use a self-heating thermistor in the feedback path, but the thermistor needed is extremelt expensive, so a small lamp (or several in series) can be used instead, if you can fnd one suitable.

The other way is to use a FET, but my experience is this is often less satisfactory and harder to get right and achieve the low distortion required.

If you're not measuring distortion, then even relatively high figures are not likely to be a problem - you won't see it on a 'scope trace, but if you are measuring it, you need to start with as pure a signal as you can achieve, or it will limit your measurement floor.

A PC sound card is a great solution, Les put up a link to some great software, that I can't find at present, maybe the old forum has the link?

The downside is you need to have the PC close to the amp etc, not always practical - and make sure you don't feed in any nasty signal to the sound card, during my early PSU tests, I went through 3 SBLive's before I decided maybe I should design a proper buffer / i/.p protection circuit .

Andy.

 

[martin clark]

Following Andy's leads above I'd just add that the lightbulb is the way to go. For an opamp-based oscillator, 12v 40mA 'pilot' lights are the way to go - they're pennies each, and work reliably.

If you want low distortion below c.200Hz, use three or four in series; the cooling rate needs to be long compared with the period of oscillation. I built a power oscillator to deliver 20W around a LM1875, and opamp and these bulbs, and could get <0.03% THD+N. at 50Hz @ peak output.

BTW that was for a DIY PS-Powerplant-alike for source components. Anyone interested in the details...?

MC