jcx - I agree with almost all of what you say. But I don't see how you can say that because someone looks at long term thermal compression that they are aware of the instantaneous effects as well. Unless they specifically mention it I have to assume that they are unaware of it. You can't assume anything else.
The audibility is an open question. There is no data either way. But I will say this as I have done the calculations and some measurements. The effect of thermal modulation will be 1000 times greater in a 1" tweeter than it is in a 1" compression driver. It is unlikely to be as big an effect or difference in a woofer.
The point that I will make is this, at some level it has to be audible as it is certainly there. Where is that point, I do not know, nor does anyone else.
To make the distinction clearly consider a thought experiment that is idealized to make a point. We have a voice coil in a magnetic field such that the heat from this coil can be drawn off completely in lets say 10 ms. Thus 10 ms. after an impulse has passed the temperature of the coil has gone back to ambient. This situation will have NO thermal compression. But now send in a sine wave that has a period longer than 10 ms. As the current rises in the coil the resistance will change instantaneously, but the thermal radiation will take some time. This instantaneous change in resistance will cause and instantaneous reduction in the current thus flattening the waveform in a nonlinear manner (odd order actually.)
In this example we can see that there is thermal modulation when in fact there is no thermal compression. The two things are clearly different.
I have tested tweeters with high levels of sound and there are frequency response changes much more rapidly than the thermal compression effects occur and in different ways. In other words, the frequency response changes very quickly reaching a steady state change only after some period of time usually 1 second or more. If thermal compression were the only effect then the frequency response changes would be constant simply growing in amplitude until they reached thermal steady state. This is not what happens.
In the above example consider a "dynamic" noise (random noise with a modulated envelope) with dynamics on the order of the thermal time constant of 10 ms. This signals envelope would be modulated by the TM (thermal modulation) effect that I propose. I have long wanted to do this test, just as I have long wanted to do a subjective test of its audibility. But then it dawned on me: "Why would I do that?" If it is not audible at all then clearly it is not audible in a system with 1000 time lower TM than another system. Only if the lower efficiency system had audible TM would it be interesting, but then only to a system designer who designs that kind of system - I don't. If it is an audible effect then high efficiency drivers are the solution. The fact that high efficiency loudspeakers are often stated as having great dynamics is clear evidence of this issue. Is it proof - not at all. Do I need proof - not at all. Basically it is a moot point as far as I am concerned. An interesting problem whose solution is already known and implimented.
Sound tracks are audio reproduction. I listen to far more sound tracks than I do two channel audio.
Decent write-up. Done well after I first stated looking at the problem. In fact we had talked about this effect here back in the late 2000's.
Anyway, I am not trying to claim any credit for "inventing" this issue. It doesn't really matter to me. But that it is NOT the same thing as thermal compression, which people have studied for decades, should now be very clear. Thermal Modulation (TM) and Thermal Compression (TC) are almost completely unrelated. This fact is almost completely misunderstood.
Last edited:
Actually it's dated 2008- sounds like late 2000s to me. I still remember the discussions- it was always intuitively easy to picture the shorter-duration thermal effects as opposed to steady state efficiency changes.
Ok so we have some relevant information that no longer needs discussion, defending or whatever....Earl finds well made compression drivers when properly mated to waveguide devices to be THE definitive high frequency transducer of today's state of engineering.
While I can't say that I personally agree, I can respect the position.
While I can't say that I personally agree, I can respect the position.
Hmm both the BMS4550 and 4552ND show appreciable extension up to almost 20k. I'd wager that when EQ'd properly I wouldn't even notice that and that's on relatively young ears.
I have never heard a CD in a domestic environment either. Anything overloaded sounds like crap and I hear you about Cinema's playing stuff too loud. Thankfully the place near us sounds good at the levels they play, although sometimes I would like them to turn it down just a little. I really would be interested in trying out a CD, but for it to make sense in my system I would need it to be ideally capable down to around 8-1kHz. The 4552ND would be the best choice, due to its compact size, but it wouldn't exactly be inexpensive as an experiment.
I too, like Earl, use my system for lots of things other than just music. HT, multichannel computer games and music. Even though music is always more enjoyable through better systems, the other two benefit just as much from having excellent clarity and image presentation.
Perhaps you could come up with a sound bite that accessibly illustrates the difference for those whom you believe don't understand it. Then you could really make use of your 'rolling eye' emoticon if there remains any further misunderstanding.
Last edited:
I believe even a 2445J diaphragm has response a little up past 20khz on axis with about a 10 - 15db loss of sensitivity compared to midband. This may not be true of a 2446 diaphragm which has about 3 db better amplitude response from 10 up to 17 khz than the 2445 diaphragm, but then drops like a rock.
I believe the TAD 4001 has a little extension up to 20khz at a reduced sensitivity.
I believe the TAD 4001 has a little extension up to 20khz at a reduced sensitivity.
I did my tweeter tests back in Thailand so that would date it at the Mid-00's as I have not been back there since 2007. We talked about it here around 2008. At that time no one accepted my contention so I don't know how "intuitive" it was.
That is not so easy to do because a model has to be made and then files have to be processed - and remember this is a nonlinear process so you can't just do a convolution. I was about to do this when I decided that there was no point. I long ago gave up publishing papers just to publish something.
When I was using a TAD4001, I found that on a CD waveguide the response drops quite rapidly above 8 kHz. It has response above 10kHz on-axis if you use a non-CD horn - beaming. Of course all this can be EQ'd.
In my old age I often have trouble with intelligibility when watching TV on a normal TV. When I watch on my HT I never have a problem with intelligibility.
Last edited:
What bassmid and what WG/tweeter driver combo ?
But you're satisfied with 15" PA midrange ?
You should try it with some high quality midrange (as is B&C 6md38 or some of the JBL 10" midranges) that has high sensitivity. Just to have a listen how it should sound - needless to say "if done properly".
Last edited:
Well what one could perhaps say is that typical thermal compression occurs over time as the general, and average, heating up process of all the thermally conductive parts within the motor structure.
The thermal compression that you're getting at though can occur when all of the motor structure is nominally very cool. What you're trying to say is that when a transient of high current flows down the voice coil, that it will heat up the wire within the voice coil by perhaps a hundred degrees or so and cause a sudden increase in the resistivity of the metal. The transient carrying far more energy than could be conducted away through the motor structure during it's occurrence, but after it is over (thankfully its periodic) the coil quickly returns to its steady state temperature and the resistivity then quickly drops back down.
Linkwitz has done some work on this where he uses tone bursts of high amplitude on two different tweeters. First of all it shows the typical thermal compression that we are used to seeing on the poorer of the two. But due to the stimulus used and the way in which he measured things, I would expect any instantaneous heating effects to be present too (ie by distorting the shape of the reproduced wave form).
Spatial distortion
I did not post that link to contest anything you said.
I remember reading that discussion back then and found it really enlightening.
I also remember you used the metaphor of a light bulb to describe the "instantaneity" of the thermal effect.
with 1.4" - 2" driver working from 500..700 to 8k or higher 4 ways is enough imo.
Last edited:
- Status
- Not open for further replies.