Ever miss an obvious solution?

Thread Starter

Veracohr

Joined Jan 3, 2011
772
Have you ever missed an obvious solution to an engineering problem?

I'm starting to design an audio line mixer for myself, and I decided to make some distortion measurements of the final product when the time comes. I had a solution to do this (in lieu of obtaining super-expensive measurement equipment), which has been in the back of my mind for a while. It involved making a very narrow notch filter to remove the test frequency, and amplifying the leftover signal. This is a pretty standard solution, yes?

I did some Google searching today about distortion measurement, and came across a solution so obvious I felt like an idiot for not thinking of it:

Difference amplifier.

As in: (DUT output) - (signal source) = DUT distortion+noise (and amplified)

How did I not think of SIMPLE SUBTRACTION before jumping to creating a filter narrow enough to remove a single frequency without affecting its harmonics? :confused:

Also, as an added bonus, this search made me aware of some opamps (referenced in the articles I found) which have SUPER low distortion figures, and aren't terribly expensive. Quite useful for audio!
 

crutschow

Joined Mar 14, 2008
34,285
Yes, it's easy to miss the best solution to a problem. That's why you don't stop at the first solution that comes to mind. You keep trying to come up with other ways to solve the problem (including doing a lot of research on the issue) and don't give up until you are pretty sure you have all (or most) of the possible solutions. Then you pick the one that appears best. Even then you keep refining the design to improve it as much as feasible. Sometimes it can be a long and tedious process.

You also have to be careful that the simplest solution actually works in practice. The problem I see with your simple difference amplifier approach is that the signal being subtracted must have exactly the same gain as the signal from the amplifier and that can be difficult to achieve in practice. For example if you are looking for a distortion value of say 0.01% for high quality audio gear then the two gains must differ by less than that value to avoid false distortion reading from gain error in the difference measurement. Have you considered that?
 

Thread Starter

Veracohr

Joined Jan 3, 2011
772
Top