kstrain said:
I am confused but I am not confusing average SPL with peak SPL 🙂
I understand that 90 dB average SPL means 95 dB or 105 dB or perhaps even 115 peak SPL depending on the dynamic range of a given recording
kstrain said:
excellent question!
because "a norm" is what "high fidelity" is all about, what "audio engineering" is all about
a norm, a standard, a model
for example "20 Hz - 20 kHz"
or it isn't?
best,
graaf
post scriptum
it is closer to "122 dB(C)" of Dr Geddes or "peaks of 122-125 dB" of Rybaudio
BUT according to National Research Council this is "far beyond normal listening levels"
graaf said:
it is closer to "122 dB(C)" of Dr Geddes or "peaks of 122-125 dB" of Rybaudio
BUT according to National Research Council this is "far beyond normal listening levels"
kstrain said:
Well this site is ceratinly no place to find "normal"!! I'm not normal, I seek the extreme situation - a theater experince that is better than a commercial one. Thats hardly normal. I mean does the average person have this, or even seek this? I don't think that normal is relavent here.
That said I also find the extremes played out here sometimes annoying. I mean I want the 6 sigma sound system, but I don't want to pay a 10 sigma price tag. Value and efficiency are very important to me while seeking the extreme.
kstrain said:
Correlating perception with measurable data can be very difficult.
I, for one, have never said the total problem with a dome tweeter was thermal compression. Its ONE potential problem. But I would tend to say that its lack of polar response control is the bigger problem at lower levels. Then when one EQs it (or designs it) for flat axial response this creates a real problem with its perceived frequency response. And it could also be that the dome does have audible nonlinearity - it would not be hard to do in a small device like that where you have very limited excusion capability.
In short there could be a number of reasons why you don't like a particular driver.
thermal compression
I'm curious if anyone has info on the thermal time constant for driver compression effects... since the argument seems to revolve around, seemingly arbitrarily, peak @ 20+dB above average levels of 95dB(C) as being suspect, with no information on dwell time at those levels. What is the "duty cycle" of typical music peaks of this level? 1%? 10%? variable all over the map? Seems I've seen data on this sort of thing (re: peak vs. average power for various types of music) but I don't recall where.
IOW, what is the compression/distortion factor vs. time @peak level relationship for the various drivers discussed? Hard to correlate continuous spl power handling data with peak data wrt compression/distortion artifacts if one doesn't know the thermal dissipation rate, signal power slew, etc.
John L.
I'm curious if anyone has info on the thermal time constant for driver compression effects... since the argument seems to revolve around, seemingly arbitrarily, peak @ 20+dB above average levels of 95dB(C) as being suspect, with no information on dwell time at those levels. What is the "duty cycle" of typical music peaks of this level? 1%? 10%? variable all over the map? Seems I've seen data on this sort of thing (re: peak vs. average power for various types of music) but I don't recall where.
IOW, what is the compression/distortion factor vs. time @peak level relationship for the various drivers discussed? Hard to correlate continuous spl power handling data with peak data wrt compression/distortion artifacts if one doesn't know the thermal dissipation rate, signal power slew, etc.
John L.
graaf said:
It's not weighted. Loudspeaker measurements seldom (if ever) are mesaured or displayed that way.. as far as I know I should add.
Electronics performance often are though when it comes to noise and dynamic range. A non weighted measurement of noise for example takes no notice on where the energy is in the band of interest and therefore two components can measure the same but sound differently noisewise.
/Peter
Re: thermal compression
Say 6dB to 20dB between average and peak but extremes can be more or less dynamic.
Google "crest factor music".
/Peter
auplater said:
Say 6dB to 20dB between average and peak but extremes can be more or less dynamic.
Google "crest factor music".
/Peter
Re: Re: thermal compression
Yes, I'm aware of crest factor, but I don't believe that's the entire story here. I'm interested in how long at peak level (maybe the "Q" of the peaks) vs. the dissipation rate of the thermally sensitive driver parts, voice coil/former etc. of the power for dynamic music signals. Might provide some insight into true failure modes for music as opposed to test signals.
John L.
Pan said:
Yes, I'm aware of crest factor, but I don't believe that's the entire story here. I'm interested in how long at peak level (maybe the "Q" of the peaks) vs. the dissipation rate of the thermally sensitive driver parts, voice coil/former etc. of the power for dynamic music signals. Might provide some insight into true failure modes for music as opposed to test signals.
John L.
gedlee said:
Yes, it appears that is the simplest explanation and likely matters most. All the domes I used had about the same polar response and were applied in about the same way. Swapping among very similar devices could have been predicted to make very little difference.
It is too easy to get trapped into thinking too narrowly by what is popularly believed or practised, then again there is a whole industry doing that.
Ken
ps. I'd almost added "nothing on this forum" as an aside to my question about what is normal, and that is a good thing too.
gedlee said:
Indeed. After all, the tag line for the site is "projects by the fanatics for the fanatics".
Sheldon
Re: Re: Re: thermal compression
I don't think that the discussion is about failure modes, but thermal modulation.
You asked about time constants - consider this. The voice coil will heat instantaneously as current is sent through it - there is no time constant. Its mean temeperature has time constants (basically two, one for the VC and the other for the magnet), but not its instantaneous temperature. How much resistance change is there from this instantaneous temperture change you should ask. It depends on the amount of metal, its temp rise versus BTU (there is a name for this but I forgot) and its temp. coef. of R. The two properties of the material are constants and tend to be very similar for metals of practical use in voice coils. Thus the single factor for the instantaneous resistance modulation is the mass of the voice coil conductor. This is virtually unaffected by the structural time constants because they are so much longer.
The longer term rise in the voice coil temperature as well as the motor structure, tend to be on the order of several milliseconds to seconds. The instantaneous modulation is what I am curious about as it has not really been looked at in any detail that I know of.
auplater said:
I don't think that the discussion is about failure modes, but thermal modulation.
You asked about time constants - consider this. The voice coil will heat instantaneously as current is sent through it - there is no time constant. Its mean temeperature has time constants (basically two, one for the VC and the other for the magnet), but not its instantaneous temperature. How much resistance change is there from this instantaneous temperture change you should ask. It depends on the amount of metal, its temp rise versus BTU (there is a name for this but I forgot) and its temp. coef. of R. The two properties of the material are constants and tend to be very similar for metals of practical use in voice coils. Thus the single factor for the instantaneous resistance modulation is the mass of the voice coil conductor. This is virtually unaffected by the structural time constants because they are so much longer.
The longer term rise in the voice coil temperature as well as the motor structure, tend to be on the order of several milliseconds to seconds. The instantaneous modulation is what I am curious about as it has not really been looked at in any detail that I know of.
thanx.. I think
Not sure about "instantly heating when current is applied" but I'll agree that a "mean temperature" is established, making the initial ramp to mean temp irrelevant to the delta values.
I believe the term for the temp rise vs. btu is "specific heat capacity". By failure modes, perhaps I misspoke, as I was lumping increased distortion and/or outright catastrophic failure together as non desirable responses.
Yes, instantaneous modulation and its effects above the mean temp is most likely what I was trying to convey.
I discussed integral cooling of vc wires with Dick Pierce 10 odd years ago using electroformed fluid filled hollow windings with one way valves pumped by vc motion, to which he expressed interest, but no funding was available and it came to naught. Mostly for pro woofer setups, but might also work in other domains.
Thanx.
John L.
Not sure about "instantly heating when current is applied" but I'll agree that a "mean temperature" is established, making the initial ramp to mean temp irrelevant to the delta values.
I believe the term for the temp rise vs. btu is "specific heat capacity". By failure modes, perhaps I misspoke, as I was lumping increased distortion and/or outright catastrophic failure together as non desirable responses.
Yes, instantaneous modulation and its effects above the mean temp is most likely what I was trying to convey.
I discussed integral cooling of vc wires with Dick Pierce 10 odd years ago using electroformed fluid filled hollow windings with one way valves pumped by vc motion, to which he expressed interest, but no funding was available and it came to naught. Mostly for pro woofer setups, but might also work in other domains.
Thanx.
John L.
Re: Re: Re: Re: thermal compression
I'm not so sure about that. Agreed that the thermal mass of the voice-coil is small (order 10K/J). The time constants for the redistribution of heat out of the voice-coil should be quite short. Otherwise the tweeter would be too fragile. The shortest time constant we care about (i.e. where there is any singnificant audio energy) is > order 100 microseconds, and there can be significant cooling - mainly by convection in ferrofluid - on timescales not much longer than that. I can only comment on tweeters with ferrofluid as those are the only ones I've measured at all.
A proper measurement focusing on sub ms timescales would, indeed, be interesting, the difficulty is distinguishing impedance changes due to large signal magnetic effects from resistance rise due to heating. I could not figure out how to do that (at hobby level).
Ken
gedlee said:
I'm not so sure about that. Agreed that the thermal mass of the voice-coil is small (order 10K/J). The time constants for the redistribution of heat out of the voice-coil should be quite short. Otherwise the tweeter would be too fragile. The shortest time constant we care about (i.e. where there is any singnificant audio energy) is > order 100 microseconds, and there can be significant cooling - mainly by convection in ferrofluid - on timescales not much longer than that. I can only comment on tweeters with ferrofluid as those are the only ones I've measured at all.
A proper measurement focusing on sub ms timescales would, indeed, be interesting, the difficulty is distinguishing impedance changes due to large signal magnetic effects from resistance rise due to heating. I could not figure out how to do that (at hobby level).
Ken
Re: Re: Re: Re: Re: thermal compression
Well guys I have studied this problem and there are several time constants. The mean temperature is affected by the two longer time constants of the voice coil and the magnet. The voice coil being much shorter than the magnet. I am excluding ferrofluid here since I have never seen any data using it. I'm sure that its time constant would be very low compared with not using it.
The time constants for the voice coil and magnet are so long that they can only react to the long term average of the signal - they do not modulate the signal at signal rates.
When I was looking at this I asked a materials physicist about the heating and he gave me the answer that you question. The voice coil wire will begin to heat instantly and its temperature will rise depending on the materials specific heat and its mass. The instantaneous wire temperature change will be virtually unaffected by the long time constants of the voice coil structure and magnet, which will have a dominate effect on the mean voice coil temperature. The part that interests me is the instantaneous wire temperature changes that happen at the signal frequencies and depend solely on the mass of copper of the wire.
IF - and this needs to be tested - the instantaneous Re changes audibly affect the sound then speakers with more copper - bigger voice coils - will have a significantly smaller change and their sound will be affected the least. This is a very strong implication that bigger is better, which I have somehow always found to be the case. The sound just appears to be more dynamic less - well less impressive for lack of a better word.
As far as the audibility goes, the nonlinear distortion studies would imply that this instantaneous Re change would be inaudible. Thats why I say that it needs to be tested.
Obviously manufacturers seek to minimize the copper in the voice coil - its expensive. But this may not be the best thing to do for the best sound.
auplater said:
kstrain said:
Well guys I have studied this problem and there are several time constants. The mean temperature is affected by the two longer time constants of the voice coil and the magnet. The voice coil being much shorter than the magnet. I am excluding ferrofluid here since I have never seen any data using it. I'm sure that its time constant would be very low compared with not using it.
The time constants for the voice coil and magnet are so long that they can only react to the long term average of the signal - they do not modulate the signal at signal rates.
When I was looking at this I asked a materials physicist about the heating and he gave me the answer that you question. The voice coil wire will begin to heat instantly and its temperature will rise depending on the materials specific heat and its mass. The instantaneous wire temperature change will be virtually unaffected by the long time constants of the voice coil structure and magnet, which will have a dominate effect on the mean voice coil temperature. The part that interests me is the instantaneous wire temperature changes that happen at the signal frequencies and depend solely on the mass of copper of the wire.
IF - and this needs to be tested - the instantaneous Re changes audibly affect the sound then speakers with more copper - bigger voice coils - will have a significantly smaller change and their sound will be affected the least. This is a very strong implication that bigger is better, which I have somehow always found to be the case. The sound just appears to be more dynamic less - well less impressive for lack of a better word.
As far as the audibility goes, the nonlinear distortion studies would imply that this instantaneous Re change would be inaudible. Thats why I say that it needs to be tested.
Obviously manufacturers seek to minimize the copper in the voice coil - its expensive. But this may not be the best thing to do for the best sound.
Re: Re: Re: Re: Re: Re: thermal compression
The instantaneous changes can be calculated easily enough by assuming for these time scales that the system is adiabatic. One problem in studying this experimentally, is that it's going to be difficult to control for a single variable like coil size, as it will also have mechanical effect on the acoustical properties of the driver.
Sheldon
gedlee said:
The instantaneous changes can be calculated easily enough by assuming for these time scales that the system is adiabatic. One problem in studying this experimentally, is that it's going to be difficult to control for a single variable like coil size, as it will also have mechanical effect on the acoustical properties of the driver.
Sheldon
I cannot beleive the "instantaneously heated" description when current flows though the wire, which by normal definition would mean a contant temperature is reached like an ideal step function. We know that this is not possible in the real world.
Edit:
Sorry for the above, I obviously missed part of the explanation. Getting close to bet time here.
Edit:
Sorry for the above, I obviously missed part of the explanation. Getting close to bet time here.
gedlee said:
I think it may be a matter of weight more than cost, at least for tweeters and smaller drivers. I've never given it much thought, but coil DCR need not be directly related to impedance, as that should be determined by the motor design and number of turns on the coil. Do designers size the wire so that the DCR comes in a little lower than the lowest impedance? Diminishing returns, compared to weight, in going larger?
Sheldon
Sheldon said:
Design implications have to follow objectives. If the objective were to increase the mass of copper to minimize thermal modulation then this could be done, but has a lot of other considerations that would have to be dealt with.
What I am interested in is IF this is an audible factor.
Dome tweeters usually have a breakup mode in the range fo 16KHz~22KHz. This is what usually creates very unpleasant experience when listening at higher levels. Hard domes tend to be more harsh than soft domes due to the way residual energy dissipates.kstrain said:
Re: Re: Re: Re: Re: Re: thermal compression
I'm going to insert comments among edited quotes from Gedlee:
"When I was looking at this I asked a materials physicist about the heating and he gave me the answer that you question."
As a physicist too I hypothesise that the VC will heat up much less, even on short timescales, because it is in a narrow gap of air or oil against a heatsink, than it would in free air.
"The part that interests me is the instantaneous wire temperature changes that happen at the signal frequencies and depend solely on the mass of copper of the wire."
I think it needs to include the cooling effect as well as the mass of copper, even on ms timescales.
"IF - and this needs to be tested - the instantaneous Re changes audibly affect the sound then speakers with more copper - bigger voice coils - will have a significantly smaller change and their sound will be affected the least. "
If cooling is indeed important, the VC area may be as important as the mass. Normally they scale together, of course, so both hypotheses would predict the "improvement", just in differing degrees.
"As far as the audibility goes, the nonlinear distortion studies would imply that this instantaneous Re change would be inaudible. Thats why I say that it needs to be tested."
Yes, putting your distortion results together with my recent tweeter destroying experience leads me to expect that the thermal effect is not very significant compared to much simpler explanations for "dome tweeter sound" (but I won't argue with your conclusion: it should be tested).
"Obviously manufacturers seek to minimize the copper in the voice coil - its expensive. But this may not be the best thing to do for the best sound."
In dome tweeters there is about 0.1g to 0.2g of copper, so it is something apart from cost that is the driver.
Ken
ps. I'm going to break from this as I've said enough, and don't have the measurements to back it up.
I'm going to insert comments among edited quotes from Gedlee:
"When I was looking at this I asked a materials physicist about the heating and he gave me the answer that you question."
As a physicist too I hypothesise that the VC will heat up much less, even on short timescales, because it is in a narrow gap of air or oil against a heatsink, than it would in free air.
"The part that interests me is the instantaneous wire temperature changes that happen at the signal frequencies and depend solely on the mass of copper of the wire."
I think it needs to include the cooling effect as well as the mass of copper, even on ms timescales.
"IF - and this needs to be tested - the instantaneous Re changes audibly affect the sound then speakers with more copper - bigger voice coils - will have a significantly smaller change and their sound will be affected the least. "
If cooling is indeed important, the VC area may be as important as the mass. Normally they scale together, of course, so both hypotheses would predict the "improvement", just in differing degrees.
"As far as the audibility goes, the nonlinear distortion studies would imply that this instantaneous Re change would be inaudible. Thats why I say that it needs to be tested."
Yes, putting your distortion results together with my recent tweeter destroying experience leads me to expect that the thermal effect is not very significant compared to much simpler explanations for "dome tweeter sound" (but I won't argue with your conclusion: it should be tested).
"Obviously manufacturers seek to minimize the copper in the voice coil - its expensive. But this may not be the best thing to do for the best sound."
In dome tweeters there is about 0.1g to 0.2g of copper, so it is something apart from cost that is the driver.
Ken
ps. I'm going to break from this as I've said enough, and don't have the measurements to back it up.