Is power compression a concern with tweeters?

[noah katz]

I'm upgrading my plans for new L/C/R's to keep up with my new subs in output and lowered power compression.

I'm thinking 4 ea of the Dayton RS180 drivers (which I gather equal the ScanSpeak 7" drivers) per speaker.

Should I double up on the tweeters? 28 mm aluminum dome, 28 mm voice coil, ferrofluid.

I guess it comes down to how much of the power in program material is there in the range from 1.5 - 2kHz on up.

Thanks

[Ron E]

Hi Noah,

I think it depends on the average SPL's you are going for. I know the Dynaudio D28's had none for transient signals up to 1000Watts, because they used to brag about it - and there was nothing super revolutionary about their motor AFAIK. As long as you have a decent tweeter rated for ~100W or more, I wouldn't worry unless you are going to run 100-110+dB continuous. If you want more than that, I would go horns.

[noah katz]

Hi Ron,

"I know the Dynaudio D28's had none for transient signals up to 1000Watts"

Do you know for how long a signal that is?

Dan Wiggins said not too long ago that 1000 W for 1 sec raises a Tumult's (15" uber sub driver) VC temp to 200 C, and a tweeter's VC will have a much shorter thermal time constant.

I guess an important question is, if program material is played at a level giving peaks of 110 dB, how high are the peaks in the tweeter's passband?

Thanks

[DSP_Geek]

Originally posted by Ron E
Hi Noah,

I think it depends on the average SPL's you are going for. I know the Dynaudio D28's had none for transient signals up to 1000Watts, because they used to brag about it - and there was nothing super revolutionary about their motor AFAIK.

The Dynaudio transient tests were for 10 milliseconds if I recall correctly, so 1000 watts dumped only about 10 joules of heat into the tweeter if the tests ran once per second, and less heat for longer repetition rates. The D-28 has an aluminum voice coil, and aluminum features a specific heat of 0.9 joules/(gram * degree C), which means that 0.9 joules raises the temperature of a gram of Al by 1 degree Celsius. Tweeter voice coils typically run around 0.1 gram, so 10 joules would raise the VC temperature by about 110 degree C. The temperature coefficient of Al is 0.0043/degree C, so 110 C means voice coil resistance has increased by 48%. The higher resistance would decrease current, thus sensitivity, by about 3 dB at the transient test mentioned above.

The graphs in the Dynaudio adverts did show power compression beginning to happen at 1000 W transient tests, so that accords with the above calculations.

Why use aluminum instead of copper in voice coils? The numbers show that voice coils of either Al or Cu will reach about the same temperature under power, but aluminum's lower thermal coefficient of resistance (0.0043 versus 0.0068 for copper) means heating effects will have a lesser effect on sensitivity. As a side benefit aluminum is lighter for a given resistance than copper, decreasing Mms and increasing sensitivity, most noticeably in tweeters where the coice coil is a higher proportion of the moving mass than for other drivers.


Cheers,
Francois.

[noah katz]

Very interesting, thanks Francois.

Is there a net efficiency gain with aluminum?

Doesn't it's higher resistance decrease it because more voltage is required to force the same current through?

And if thicker wire is used the radial gap needs to be bigger for the same number of turns, meaning lower efficiency from less B.

[DSP_Geek]

Originally posted by noah katz
Very interesting, thanks Francois.

Is there a net efficiency gain with aluminum?

Doesn't it's higher resistance decrease it because more voltage is required to force the same current through?

And if thicker wire is used the radial gap needs to be bigger for the same number of turns, meaning lower efficiency from less B.

Well, the voice coil will be 6 ohms whether it's Al or Cu, although you do have to make the aluminum wire a bit thicker: Al has a resistivity of 2.65 * 10^-8 ohms-meter, whereas Cu has 1.68 * 10^-8 ohms-meter, so aluminum needs (2.65/1.68) more area to have the same resistance as copper, or about 25% more diameter. Considering the voice coil is also covered with a finite thickness of varnish, and it's wound on a former, that ends up being closer to a 15% wider gap - not a big problem with powerful magnets available.

The greater resistivity of aluminum means you need about 55% more material than copper for the same resistance, BUT copper is denser than aluminum (8.96 g/cc versus 2.70 g/cc), so the copper voil coil ends up weighing about twice as much as the aluminum one. Not a really big deal for woofers where the weight of the voice coil is maybe 5 to 10% of Mms, quite important for tweeters where the VC can be 20 to 30% of Mms.


Cheers,
Francois.