Re: Cooling plugs
Absolutely. Rod Elliot did a study of voice coil heating at low power levels. He was looking at it from the other end of the spectrum than I was, but the consequences are the same.
Agreed. Convection cooling for the coil, pole piece wicking for the core. I like what you and Nick are doing. Seems like we've traveled the same path.
What's amazing to me is how fast the speaker core turns into an EZ bake oven.
Someone was propogating the myth a few years back that horn loading makes the drivers indestructable from a thermal perspective. The idea was that efficiency was raised so much that all the input energy was being transformed to kinetic and nothing remained as heat.
What I saw was that nearly all the failures were thermal. Sure, the excursion was reduced and the efficiency increased. But efficiency wasn't 100%, so some energy still turned into heat. A lot of it, to tell the truth. And with excursion reduced, the cooling vents were stalled. These guys were cooking the coils on a regular basis, always telling themselves that it couldn't be thermal. Truth was, it was thermal and it was obvious. Burned smell, black coil, unwound from the former.
Lynn Olson said:
Absolutely. Rod Elliot did a study of voice coil heating at low power levels. He was looking at it from the other end of the spectrum than I was, but the consequences are the same.
John_E_Janowitz said:
Agreed. Convection cooling for the coil, pole piece wicking for the core. I like what you and Nick are doing. Seems like we've traveled the same path.
nickmckinney said:
What's amazing to me is how fast the speaker core turns into an EZ bake oven.
Someone was propogating the myth a few years back that horn loading makes the drivers indestructable from a thermal perspective. The idea was that efficiency was raised so much that all the input energy was being transformed to kinetic and nothing remained as heat.
What I saw was that nearly all the failures were thermal. Sure, the excursion was reduced and the efficiency increased. But efficiency wasn't 100%, so some energy still turned into heat. A lot of it, to tell the truth. And with excursion reduced, the cooling vents were stalled. These guys were cooking the coils on a regular basis, always telling themselves that it couldn't be thermal. Truth was, it was thermal and it was obvious. Burned smell, black coil, unwound from the former.
gedlee said:
Funny how obvious this issue is for some of us, but when you try to explain voicecoil heating to most people (including some of the 'diy luminaries'), they dismiss it as a non-issue. Basically, they actually think that it's not a problem with their 85dB stand mount 2 ways.
I think the improved heat characteristics of most high eff speakers are a significant contributor to the notion that higher efficiency has better dynamics. While this is not always true, nor the only factor, it's a good reason to shoot for higher efficiency designs.
Plus they're bigger. And we all know that bigger is better.
Re: Re: Cooling plugs
Wouldn't be better to be no bigger than the magnet diam and have fins like in amps or cpus? So not to obstruct the back wave, in case it is not for a sub woofer?
nickmckinney said:
Thats the biggest "radiator" I have seen on a pole yet
If not in a sealed cabinet I like it a lot. I like where the thread is going, heat issues are a serious problem in loudspeakers since they are so pathetically inefficient.
Wouldn't be better to be no bigger than the magnet diam and have fins like in amps or cpus? So not to obstruct the back wave, in case it is not for a sub woofer?
Re: Re: Cooling plugs
The plate is mounted on the side of a basshorn subwoofer, doubling as the driver access panel. Two woofers are used in push-pull drive.
If used in sunlight, a cover snaps over the cooling plates so they aren't heated by the sun. I originally thought I would need airflow across the plate to cool it, but it evidently has enough area that heat dissipates into ambient air and the materials around it. So you can stack them together with no space in between.
The size of the plate was dictated by the access hole size. It had to be be large enough to fit the woofer. It is large enough that fins aren't necessary, although in a different application it could be made smaller and fins cut into the plate to increase surface area.
nickmckinney said:
The plate is mounted on the side of a basshorn subwoofer, doubling as the driver access panel. Two woofers are used in push-pull drive.
If used in sunlight, a cover snaps over the cooling plates so they aren't heated by the sun. I originally thought I would need airflow across the plate to cool it, but it evidently has enough area that heat dissipates into ambient air and the materials around it. So you can stack them together with no space in between.
The size of the plate was dictated by the access hole size. It had to be be large enough to fit the woofer. It is large enough that fins aren't necessary, although in a different application it could be made smaller and fins cut into the plate to increase surface area.
I am way out of my depth here but I wonder when I see talk of ceramic adhesives etc whether it might be possible to spray a mold of the inside of the core or former with a ceramic, wind the coil, spray with another coat of ceramic so that the former is ceramic with the coil imbedded in it.
jamikl
jamikl
Sensitivity and Efficiency
I am surprised to hear you say this, Earl. dB/W/M figures describe sensitivity, not efficiency. Without also knowing directivity, you can't translate SPL/W/M into conversion efficiency. But I think you are assuming omnidirectional radiation in this statement.
Even so, assuming omnidirectional radiation, 96dB/W/M is about 2.5% efficient which means most power is converted to heat. If the speaker is more directional, like a horn, then 96dB/W/M describes an even less efficient power conversion, since the sound is concentrated in a smaller field, and so not as much total power is radiated. If efficiency is fixed and the radiation angle is reduced, sound energy is more focused and sound pressure level within the coverage angle increases. So all that said, 96dB/W/M is 2.5% efficiency at best, maybe less depending on radiation pattern.
A speaker that generates 92dB/W/M in an omnidirectional pattern is 1% efficient. So 96dB/W/M is 2.5x as efficient. Any time you double the efficiency, you can expect 3dB increase and better thermal performance. But still, in either case, we're talking about single digit conversion efficiency, so most power is still converted to heat. One would hope for a good cooling system in the speaker, else the temperature local to the voice coil will rise. Voice coil venting may be useful at low frequencies where the cone can act as a pump and move air through the gap, across the coil. But at midrange frequencies, cone excursion is small and the size, shape and geometry of the vents isn't optimal for airflow at those frequencies. The cooling vents are stalled above bass frequencies, so the only way to get heat out of the motor is with other measures, like the heat wicking mechanism that Nick and I have been talking about.
This isn't limited to prosound applications, as Lynn said. Even if you're just putting a few dozen watts in your speaker, that power is converted into heat. At 2.5% efficiency, most of the power is converted into heat. Think of the heat of a 25 watt soldering iron contained inside a thermos bottle. That's pretty much what you have in a loudspeaker, with the voice coil on the pole piece surrounded by the magnet. It gets pretty hot in there.
gedlee said:
I am surprised to hear you say this, Earl. dB/W/M figures describe sensitivity, not efficiency. Without also knowing directivity, you can't translate SPL/W/M into conversion efficiency. But I think you are assuming omnidirectional radiation in this statement.
Even so, assuming omnidirectional radiation, 96dB/W/M is about 2.5% efficient which means most power is converted to heat. If the speaker is more directional, like a horn, then 96dB/W/M describes an even less efficient power conversion, since the sound is concentrated in a smaller field, and so not as much total power is radiated. If efficiency is fixed and the radiation angle is reduced, sound energy is more focused and sound pressure level within the coverage angle increases. So all that said, 96dB/W/M is 2.5% efficiency at best, maybe less depending on radiation pattern.
A speaker that generates 92dB/W/M in an omnidirectional pattern is 1% efficient. So 96dB/W/M is 2.5x as efficient. Any time you double the efficiency, you can expect 3dB increase and better thermal performance. But still, in either case, we're talking about single digit conversion efficiency, so most power is still converted to heat. One would hope for a good cooling system in the speaker, else the temperature local to the voice coil will rise. Voice coil venting may be useful at low frequencies where the cone can act as a pump and move air through the gap, across the coil. But at midrange frequencies, cone excursion is small and the size, shape and geometry of the vents isn't optimal for airflow at those frequencies. The cooling vents are stalled above bass frequencies, so the only way to get heat out of the motor is with other measures, like the heat wicking mechanism that Nick and I have been talking about.
This isn't limited to prosound applications, as Lynn said. Even if you're just putting a few dozen watts in your speaker, that power is converted into heat. At 2.5% efficiency, most of the power is converted into heat. Think of the heat of a 25 watt soldering iron contained inside a thermos bottle. That's pretty much what you have in a loudspeaker, with the voice coil on the pole piece surrounded by the magnet. It gets pretty hot in there.
Wayne - I know the difference between sensitivity and efficiency. I should have said "sensitivity drops below ..."
But it is the sensitivity that matters, not efficiency, regardless of directivity, because we are interested in a specific SPL at the listening point (ignoring the reverbereant field for the moment). So 6 dB more sensitivity results in 1/2 the heat generated for an equivalent SPL - thats 100% improvement. AT best a cooling system might increase the heat flow by 30-50% and that is high. So my point stands that efficiency/sensitivity is far more effective at thermal managment than any cooling techniques. One need only resort to cooling techniques when the efficiency has maxed out AND you still haven't reached the desired SPL. This is NOT going to happen in a home situation. Any speaker with sensitivity above about 96 dB/watt/meter is going to be fine in a home. Anything less than about 92 dB is going to have trouble with higher SPLs and anything less than say 88 dB is going to have thermal problems at almost any listening level.
But it is the sensitivity that matters, not efficiency, regardless of directivity, because we are interested in a specific SPL at the listening point (ignoring the reverbereant field for the moment). So 6 dB more sensitivity results in 1/2 the heat generated for an equivalent SPL - thats 100% improvement. AT best a cooling system might increase the heat flow by 30-50% and that is high. So my point stands that efficiency/sensitivity is far more effective at thermal managment than any cooling techniques. One need only resort to cooling techniques when the efficiency has maxed out AND you still haven't reached the desired SPL. This is NOT going to happen in a home situation. Any speaker with sensitivity above about 96 dB/watt/meter is going to be fine in a home. Anything less than about 92 dB is going to have trouble with higher SPLs and anything less than say 88 dB is going to have thermal problems at almost any listening level.
"Even so, assuming omnidirectional radiation, 96dB/W/M is about 2.5% efficient which means most power is converted to heat."
Hello Wayne
I think the real issue is about amount power going into VC in the first place. Thats what the higher sensitivity really tell us depending of course on how that was measured. Averaged over the bandwidth of the drivers vs. measure at a peak.
If you work the numbers from 92db vs 98db that's a 4 fold power increase into 92db driver. Assuming thats at 1 watt you get 4 into the other driver. If you go up 10db in level that becomes 10 vs 40 watts. 40 watt's is a lot of power and 10 watts seems a much more reasonable number for the driver to heatsink.
As you go up in level is where you see significant power level diferences. That's where I think the advantage lies.
Rob
Hello Wayne
I think the real issue is about amount power going into VC in the first place. Thats what the higher sensitivity really tell us depending of course on how that was measured. Averaged over the bandwidth of the drivers vs. measure at a peak.
If you work the numbers from 92db vs 98db that's a 4 fold power increase into 92db driver. Assuming thats at 1 watt you get 4 into the other driver. If you go up 10db in level that becomes 10 vs 40 watts. 40 watt's is a lot of power and 10 watts seems a much more reasonable number for the driver to heatsink.
As you go up in level is where you see significant power level diferences. That's where I think the advantage lies.
Rob
gedlee said:
I agree with you to a large degree, it's better to increase efficiency than to improve heat management. But the two don't have to be mutually exclusive and I think it is important to look at the real numbers in any case. On one hand we can agree that high efficiency speakers convert more electrical energy to kinetic energy and ultimately into acoustic energy. On the other hand, I think it is important to remember that even a speaker that generates 96dB/W/M omnidirectionally is only 2.5% efficient.
When we're sending only 10 watts to a prosound woofer designed for 600 watts, heat isn't an issue. But once you start increasing power to 50 watts, 100 watts, 200 watts, these thermal issues start coming into play. Our speakers are certainly capable of those levels and they perform very well. But there is room for improvement, and one place to do it is with motor core heat management.
Robh3606 said:
That's very true. It's really a matter of scale. What matters is the driver be used in a range where it has good thermal performance. I prefer high efficiency speakers over low efficiency speakers and do not make anything with on-axis sensitivity under 95dB/2.83v/M. Still, even a high efficiency speaker converts most input power to heat, and if the listening environment is large, a lot of power is required. In those situations, both high efficiency and heat management are important. Even in home hifi use, I think there is a benefit in having both. It never hurts, that's for sure.
Wayne Parham said:
In my listening room I could cause physical pain and never clip my 100 watt amp. The room would have to be huge for there to be a thermal issue with Hi-eff speakers. Since all "fixes" cost money, fixing a problem that doesn't exist is not good engineering. Even if "It never hurts".
gedlee said:
My four Pi loudspeakers are very similar to your Summa's, so I am quite aware of the SPL they'll generate. 100 watts isn't a lot for them, but it is enough to start introducing thermal and magnetic issues. If it weren't for gap venting and other thermal management already built-in to the drivers we use, they would be in serious trouble at 100 watts. As it is, they are pretty good for another 3dB or so above that. After that, they're getting pretty hot.
You may not care how speakers perform above 100 watts, and that's fine. But I do.
A bit about Lowthers
I was off at VSAC, and thus am a bit late getting into your discussion about Lowthers.
The aluminum voice coils have several sharp bends in them. The bends have resulted in work hardening of the coils. This is a potential point of failure. Warming up the voice coil before playing it will result in these areas becoming less brittle, and therefore less likely to fail. I have seen a few failures of this type - but they are rare.
Lowther does have a Hi-Ferric coating on the voice coils, which consist of baked on iron particles. It does help transient response. It is applied as standard, but can be left off by request. My personal preference is to leave it off of the Alnico drivers, and have it applied to the others. This is due to the warmer, richer nature of the Alnico sound, which does not come through as well with Hi-Ferric applied. On the other drivers it helps the transient response without any ill effect.
The voice coil on all Lowthers (except the 4 ohm versions) are wound on both the inside and the outside of the voice coil.
A little bit about VSAC. I used Tone Tubby 12 Alnico drivers with PM6A's on a 18" x 40" open baffle. VSAC information You can see a bit here. Thanks for the information about those Tone Tubby's Lynn, they really did sound great. Anyway, lots of positive comments about their sound. They hung out to about 60 hz or so, and my Hartley 18" in the corner picked up around 50 hz.
I was off at VSAC, and thus am a bit late getting into your discussion about Lowthers.
The aluminum voice coils have several sharp bends in them. The bends have resulted in work hardening of the coils. This is a potential point of failure. Warming up the voice coil before playing it will result in these areas becoming less brittle, and therefore less likely to fail. I have seen a few failures of this type - but they are rare.
Lowther does have a Hi-Ferric coating on the voice coils, which consist of baked on iron particles. It does help transient response. It is applied as standard, but can be left off by request. My personal preference is to leave it off of the Alnico drivers, and have it applied to the others. This is due to the warmer, richer nature of the Alnico sound, which does not come through as well with Hi-Ferric applied. On the other drivers it helps the transient response without any ill effect.
The voice coil on all Lowthers (except the 4 ohm versions) are wound on both the inside and the outside of the voice coil.
A little bit about VSAC. I used Tone Tubby 12 Alnico drivers with PM6A's on a 18" x 40" open baffle. VSAC information You can see a bit here. Thanks for the information about those Tone Tubby's Lynn, they really did sound great. Anyway, lots of positive comments about their sound. They hung out to about 60 hz or so, and my Hartley 18" in the corner picked up around 50 hz.
Re: Sensitivity and Efficiency
Wayne, lets not mix up two very different effects of heat in a speaker – which easily happens for those from the pro department.
- quality of sound
- reliability
Second one - wont be a problem to most home gear not heavily abused even if its low sensitive. At least I'm not that much interested into improving that unless the glue dribbles on my carpet.
First one – yes that's the one I am really interested in. The solution to improve "distortion" due to the heat up of the VC wire and to shorten temp decay DRASTICALLY in real world can't be provided by improved cooling techiques (at least not passive = without refrigerator) regardless of how big your cooling fins or heat pipes are nor by pushing sensitivity of a speaker.
The ONLY ways to get rid of the big sonic impacts of heat in a speaker - somehow sloppy called power compression - is either using something similar behaving like a current amp or using a VC suitable wire with close to zero tempco.
Figures of power compression are good to have (though seldom seen) but don't tell the whole story.
Thermal transients have to be better specified than by a random noise measurement. The rise and decay and to what degree the cooling mechanism is program dependant should be reflected.
Greetings
Michael
Wayne Parham said:
Wayne, lets not mix up two very different effects of heat in a speaker – which easily happens for those from the pro department.
- quality of sound
- reliability
Second one - wont be a problem to most home gear not heavily abused even if its low sensitive. At least I'm not that much interested into improving that unless the glue dribbles on my carpet.
First one – yes that's the one I am really interested in. The solution to improve "distortion" due to the heat up of the VC wire and to shorten temp decay DRASTICALLY in real world can't be provided by improved cooling techiques (at least not passive = without refrigerator) regardless of how big your cooling fins or heat pipes are nor by pushing sensitivity of a speaker.
The ONLY ways to get rid of the big sonic impacts of heat in a speaker - somehow sloppy called power compression - is either using something similar behaving like a current amp or using a VC suitable wire with close to zero tempco.
Figures of power compression are good to have (though seldom seen) but don't tell the whole story.
Thermal transients have to be better specified than by a random noise measurement. The rise and decay and to what degree the cooling mechanism is program dependant should be reflected.
Greetings
Michael