Ethan Winer Null Test

0 Members and 1 Guest are viewing this topic. Read 2320 times.


Re: Ethan Winer Null Test
« Reply #60 on: 21 Nov 2018, 06:37 pm »
Years ago Hafler built an amplifier, the XL 280 that they claimed passed a null test from 20 to 20K Hz.

Our investigation showed that they had added a trimpot that flattened the frequency response at the very high end of its range.  However they did this by providing a bit of feed forward compensation that made the measured high end appear flat while adding a significant very high frequency resonate peak centered around 40K Hz.

This made the amp susceptable to out of band HF garbage such as cartridge mis-tracking, out of band distortion products from equipment ahead in the chain, etc.  The results was a unit that measured really good but sounded bright.

Passing a null test only measured in the audio range is not a useful criteria for perfect performance.


Thanks for that information Frank, and for also noticing the null tester's frequency response wall/filter and its
manipulation of the signal. Rise time, of course, being altered by the frequency response restriction.

Cheers and great Thanksgiving.

« Last Edit: 21 Nov 2018, 09:29 pm by Steve »


  • Jr. Member
  • Posts: 181
Re: Ethan Winer Null Test
« Reply #61 on: 9 Dec 2018, 08:25 am »
It may be difficult to pull off, but I think the place to do a null test like this is at the speaker level, recording the speaker playing a specific track with one cable hooked up, and then again with another cable hooked up, or maybe a different dac or whatever. This would test the cable's interaction with the complete system. As long as we were careful not to accidentally nudge the microphones or speaker position, I'll bet we could tell within 0.1 decibel if there were any actual spectral shift in the output.
If it's less than that, at least it gives us an idea of what level of subtlety we are dealing with. Even if there's no measurable spectral shift but a slight change in overall gain that could be telling.