Audio discussion with a friend

Discussion in 'Random Thoughts' started by ultrabike, Nov 22, 2017.

Tags:
  1. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    So a coworker of mine and I discussed a few things about audio in general. Too general to fit in a particular thread IMO, and in a way sort of random. So here are some of the outcomes from our brief talk:

    1. Tube and Transistor Distortion: As many here know, tubes in general are measurably more linear than transistors. These has been told to a lot of people when discussing amplifier performance. So the question naturally is why in general transistor based amplifiers tend to have less distortion than tube amplifiers? The answer to that is related to the fact that in general transistors have way more open loop gain in the audio band, and allow for more feedback to correct for non-linear distortion behavior. Tubes, while more linear, do not exhibit as high an open loop gain and cannot provide the same level of feedback to correct for their own non-linear residuals.
    2. Class A vs other topologies: In audio, the vast majority of devices are class AB or some sort of class D. Many hi-fi systems are class A and mention this as a desirable feature. In general, and due to feedback and powerful filtering techniques, some class AB and D (or D-like) amplifiers can measure better than class A. It is perhaps possible to make a class A amplifier measure better with some feedback though. The question then is, are there situations where class A is a must? The answer is yes. There are relatively high bandwidth/rate applications with stringent out-of-band noise limitation requirements that make class A almost the only possible solution. Open loop gain goes down with frequency and therefore so does feedback. Switching noise is also a problem sometimes. Note however, that this is an inconvenience and it's being overcomed with newer technologies. Audio is not a high bandwidth/rate application.
    3. Instability: With feedback comes instability. If the feedback (error) signal changes polarity then instead of correcting errors, the feedback makes the system blow up. This happens often with op-amps that are not unity gain stable or even multiple stage amplifiers where poles get moved around to the point that things go boom. Careful design with simulation and testing is needed making sure one is using parts as intended. Which argues against random op-amp rolling and similar things.
    Anyhow, just random thoughts that came up from a discussion with a friend. I write this crap not only for others, but for me as well since I tend to forget crap and from time to time I read things in the interwebz that throw me off balance.

    Take note that I'm not talking about this being better than that. Nor am I talking about subjectives. Just some info that can help when dealing with misconceptions and the topics of discussion with my friend. Some folks already know well what I'm talking about. I did, and I forget.

    BTW, my friend designed and build his own mega-watts class A tube amplifier to power his planar magnetic full size speakers. He likes tubes because he finds them engaging. I think he had to rewire his house to power his monster amp. He also said that he did not want to deal with laying out a board and deal with some of the details that come with smaller op-amp based designs (LOL! but he rewired his house so go figure). He likes to hand wire shit. He understands distortion and other stuff mentioned here from our discussion.
     
    Last edited: Nov 22, 2017

Share This Page