What would you measure?

[zozo]

Ok, so I have irregular access to an Audio Precision measurment set-up, and basically any other audio measurent tool known to man through a friend. I have doen all the basic measurements on the pair of modified mk iv's i have (freq response, THDvsfreq, THDvspower, THDvsBIASvspower, FFT at 1K). ALL that these have really told me is that hafler was right with the st-70's bias point and not the mk-ivs, and that the amps are functional, low distortion amps.

My main question is how can you and do you USE measurments to figure out how to make the amp better?

Sean

[Shannon Parks]

Ok, so I have irregular access to an Audio Precision measurment set-up, and basically any other audio measurent tool known to man through a friend. I have doen all the basic measurements on the pair of modified mk iv's i have (freq response, THDvsfreq, THDvspower, THDvsBIASvspower, FFT at 1K). ALL that these have really told me is that hafler was right with the st-70's bias point and not the mk-ivs, and that the amps are functional, low distortion amps.

My main question is how can you and do you USE measurments to figure out how to make the amp better?

Sean

Hi Sean,

Sorry for a slow answer. An audio analyzer setup can really collect a lot of data in a short period of time. The first key is to get some time to use that gear and log some hours on the test bench - the more you learn how to use it the better.

It sounds like you are already on your way. You don't have to start right in trying to optimize the amp, but rather try to understand what the original enginner was trying to accomplish and for what reasons. Simply comparing both channels (or monoblocks) can tell you a bunch. Comparing the data to the original specs is useful, too. We've discovered a sick MKIII and a HF-87 with busted PS this way. And little brain power was needed.

The more technical use would be to use it to collect open loop amplifier data. Take note of where your transformer resonance and where your 120 degree gain is at. Then you balance bandwidth and distortion performance versus those two points. Much quicker than using a scope and sig gen.

Rozenblit addresses this some. If anyone has a good book to recommend, let us know. But many a good amplifier has been an unstable mess because stability wasn't addressed properly. I also think this is the cause of negative feedback bias. Looking at old schematics with amps that used top notch iron, you see low pass filters all over the place. Flatness out to 100KHz looks nice on paper but ends up sounding like an oscillator.

Shannon