I guess this Outlaw stuff is catching. Could not a smiley face with a cowboy hat?

Most people will think this is Outlawish, but I use a Mesa/Boogie Stategy 400 tube amp. Yes, it was a guitar amp, and by changing some of the tubes (now at 12 6L6's), could be used for Home stereo. I think it sounds great. I have never been able to audibly hear it reach any ceiling that I could take. Maybe because I use full range speakers each with two 10" woofers a 6" midrange and the Heil Air Motion Tweeter. Then there are four of those along with two 15" subs. So you can see that in a regular size room I can't turn it up two much.

------------------
MeanGene's Home

[This message has been edited by MeanGene (edited January 16, 2003).]

[This message has been edited by MeanGene (edited January 16, 2003).]

I think I heard an ad on the radio the other day. The Tubes are playing an exclusive show up in Marin County somewhere...

Do you think an impulse based test like an MLS system would show this, by avoiding the steady state tone as used in more conventional tests?

Charlie:

I don't think the MLS tests would necessairly be of any more use than a conventional distortion analyzer and ocilloscope (I''m assuming you are talking about electronic measurements here, not acoustic ones). If you zoom in enough at the crossover point of the wave on a solid state amp, you can sometimes see the 'glitch'. With high sensitivity speakers, it can be heard as a 'roughness' and 'courseness' to low level signals. In extreme cases, I have heard the sound actually dropping out as the wave transitioned the crossover point - it sounds a lot like a low bit rate digital recording, without dither added (sort of like telephones can sound if the sound was digitized at some point). Since on a tube amp each half of the push-pull stage is essentially operating class "A" for a significant amount of it's power range, there is no crossover notch to be generated. By the time it does cross into class "A/B", any crossover glitch is a small proportion of the total signal level. A solid state amp generally is designed to have very little overlap of the positive and negative swings of the wave. Of course a solid state amp can be constructed that has a richer biasing, even pure class "A". This is almost never done however except for very high-end (and expensive) amps due to the costs and amount of heat generated. This heat becomes even more of an issue in a multi-channel amp intended for home theater.

[This message has been edited by soundhound (edited January 20, 2003).]

Yep - I'm talking about using an MLS system to generate the 'signal' and analyze the result as captured from the amp outputs, without generating any messy sound waves.

I was actually more wondering about TIM than crossover distortion - crossover distortion is pretty easy to observe with a digital scope and some care. What do you think? I'll probably try it sooner or later - just for kicks.

EDIT - It just occurred to me that the levels of TIM (IIRC) are pretty low percentage wise, so I wonder if MLS would be sensitive enough to detect it without super expensive capture devices...

I suppose one could run multiple samples to increase the sensitivity, I think so anyway.

[This message has been edited by charlie (edited January 20, 2003).]

I've never used the MLS system to specifically look for TIM distortion, so you will be charting some interesting territory.

Yeah, I'm sure someone somewhere has done it. I've looked at tests of other (non-speaker/microphone) stuff like soundcards and amps using MLS, but I'm not sure they were specifically looking for TIM.

It seems like the 'chirp' would tend to excite a different response than a steady tone if there is a significant time factor present though.

I don't think there is any 'storage' mechanism involved in TIM, unlike a resonance in a speaker. But the test would prove this one way or another.

Actually I was referring to the 'delay' in getting the global negative feedback around to the input again. If this is a significant source of distortion it seems like an impulse test of some sort would excite it, maybe?

Yes, It would seem to catch that latency. BTW, it's that 'always lagging the signal' nature of negative feedback (even though it's extremely small) that makes me shy away from large amounts of it - or entirely