I thought Jim had already determined an acceptable drive level for the speaker long ago when he posted max levels using sine waves.
I wrote today:
"Just keep an eye on the excursion, keep the time on short, don't go too far past Xmax and your speaker will be OK. If the excursion sims are correct (does not look like they are) the speaker should not be showing harmonics above 20 dB down from the fundamental at 46 volts."
This is the test plan we've used in previous year's Prosound Shootout events:
Two notes that aren't clear in that document:
1. We measure the impedance chart of each speaker, and then use minimum impedance to calculate the drive voltage. So the calculated power is only actually achieved at the placed in the impedance curve where impedance is lowest. Where the impedance rises, it is actually dissipating less power. But this was deemed to be the best approach. It is relatively easy to find Zmin (as apposed to exporting the impedance curve into a process that would find the area under the curve and provide an average impedance for us) and it leans toward the conservative side. In practice, the difference between decibel levels using Zmin and a value slightly higher than Zmin (as average would be) are insignificant (like 0.2dB).
2. We do not perform Step 7 - Heat Soak - because most people were uncomfortable with that step. It is something I personally am familiar with, and find valuable. But especially in light of the fact that we have a lot of measurements to do in a day, time is precious, and having a 15 minute heat soak period for each test cycle is probably time prohibitive.
Wayne, the only thing I could find on your website so far is that you might have forgotten to mention the magnetic field created by the voice coil does also heat up the magnet in the motor 😉, especially under high loads at high power compression figures. But don't border cause nobody seems to care, its all Xmax that rules for most...
Last edited:
When you don't have radar, like davy http://www.diyaudio.com/forums/subwoofers/153359-reading-driver-displacement-radar.html you gotta latch on to something to feel good about your subs lasting the night......
😉
Absolutely, magnetic eddy currents definitely heat the magnet and motor core. There's direct heating from those eddy currents, radiated heating from the voice coil, and convection heating from the air passing by. The magnet and motor core is like a thermos bottle that stores a LOT of heat, making the voice coil's local ambient temperature very high in powerful woofers. You could cook or boil water on them.
Worse yet, when the driver is used in cabinets that reduce excursion (like horns), the cooling vent is stalled, so forced air convection cooling doesn't work as well as it could. That makes cooling plugs very helpful in systems that are used at high power levels for extended periods.
Hi DJim,
This is not completely true, in Holland (and a lot of other European countries like Germany, Belgium, Austria, Swiss) the new laws indeed involve continuing measurements and logging. But the limits are determined in LAeq value of 15 or 30 minutes (values of around 102LAeq). This means two things:
- The values are A-weighted, so freqs <100Hz hardly count (only Germany has an extra limit on 135dBC(SLOW);
- Since the LAeq value is the equivalent dBA value which energetically averaged over 15 or 30 minutes, there is plenty of room for dynamics. This allows an engineer to play the music at 100dB(A) with peaks up to 110dB(A).
The interesting thing is that some countries have decided to measure these limits not at FOH (where the engineer is), at THE LOUDEST SPOT IN THE AUDIENCE, which means, off course, right in front of the speakers! Since this difference is approx. 15dB with FOH, it means one has to do the show at 85dB(A), which is no fun..(audience usually is around 90-95dB(A) )
This will form a nice challenge for venues where it is not possible to fly the speakers, or make sub arrays.
Kees
This is not completely true, in Holland (and a lot of other European countries like Germany, Belgium, Austria, Swiss) the new laws indeed involve continuing measurements and logging. But the limits are determined in LAeq value of 15 or 30 minutes (values of around 102LAeq). This means two things:
- The values are A-weighted, so freqs <100Hz hardly count (only Germany has an extra limit on 135dBC(SLOW);
- Since the LAeq value is the equivalent dBA value which energetically averaged over 15 or 30 minutes, there is plenty of room for dynamics. This allows an engineer to play the music at 100dB(A) with peaks up to 110dB(A).
The interesting thing is that some countries have decided to measure these limits not at FOH (where the engineer is), at THE LOUDEST SPOT IN THE AUDIENCE, which means, off course, right in front of the speakers! Since this difference is approx. 15dB with FOH, it means one has to do the show at 85dB(A), which is no fun..(audience usually is around 90-95dB(A) )
This will form a nice challenge for venues where it is not possible to fly the speakers, or make sub arrays.
Kees
Jbell, I meant in general here on DIY.
About your worries, trust your ears and eyes and you don’t need a radar. No driver will break mechanically over a couple of millimetres beyond a 9.6mm Xmax (and that is still way below Xmech). However, they can easily burn at even a couple of millimetres excursion (and that is way below Xmax) at the point where excursion is minimum in a tapped horn.
On Wayne’s website you can find almost everything in detail how that works. Earlier his method was described as unrealistic since sine-waves don’t compare to complex music signals. True but it would result in the same conclusion if he would run noise over longer periods of time. The reason is simple; heat builds up over long time and don’t disappear in 60seconds, even if the signal completely disappears that long. That means over a long period’s even music signals can drive a motor into critical temperature ranges.
Now, you can border yourself over a couple of mm at Xmax but that does not harm anything (except for crappie or unbalanced drivers) but realise that excursion is the best method for releasing heat from the motor! High excursion means lots of air movement in and out the motor system which is still the most effective way to cool down your motor (besides modifications)
But the reality is even you feed them up to their AES power rating and prediction says it will even pass Xmax, it won’t. That’s where power compression comes in and THD is showing its effects. If you push the driver beyond that point on continues base, the driver won’t break either but VC does burn. That’s why you need to keep the driver under its rated AES power.
Just a few rules you need to consider; the Fs of the driver must be below the Fb of the system and use specialised (extended) LF drivers since they are designed for the task.
For Wayne; I do appreciate all your efforts in your dedicated research and willingness to publish all findings openly but from a commercial point of view you sure know how to pick a lousy driver 😉.
About your worries, trust your ears and eyes and you don’t need a radar. No driver will break mechanically over a couple of millimetres beyond a 9.6mm Xmax (and that is still way below Xmech). However, they can easily burn at even a couple of millimetres excursion (and that is way below Xmax) at the point where excursion is minimum in a tapped horn.
On Wayne’s website you can find almost everything in detail how that works. Earlier his method was described as unrealistic since sine-waves don’t compare to complex music signals. True but it would result in the same conclusion if he would run noise over longer periods of time. The reason is simple; heat builds up over long time and don’t disappear in 60seconds, even if the signal completely disappears that long. That means over a long period’s even music signals can drive a motor into critical temperature ranges.
Now, you can border yourself over a couple of mm at Xmax but that does not harm anything (except for crappie or unbalanced drivers) but realise that excursion is the best method for releasing heat from the motor! High excursion means lots of air movement in and out the motor system which is still the most effective way to cool down your motor (besides modifications)
But the reality is even you feed them up to their AES power rating and prediction says it will even pass Xmax, it won’t. That’s where power compression comes in and THD is showing its effects. If you push the driver beyond that point on continues base, the driver won’t break either but VC does burn. That’s why you need to keep the driver under its rated AES power.
Just a few rules you need to consider; the Fs of the driver must be below the Fb of the system and use specialised (extended) LF drivers since they are designed for the task.
For Wayne; I do appreciate all your efforts in your dedicated research and willingness to publish all findings openly but from a commercial point of view you sure know how to pick a lousy driver 😉.
Thanks Kees for clearing up. I haven’t red the full details yet but I do know in Brussels they can’t seem to find agreement about a more general law…
I had a dream the other night that I built the same style of cabinet (tqwt) that Bjorno helped me with but it was 4 times the size and had a TC Sounds 5400 in it.
I've never really inspected the specs for the driver, but I suspect that would be a bad idea.
its just a cone moving backwards and forwards. my guess is, if there is enough power and design to make it happen in a controlled way,
more x max is always better.
more x max is always better.
You must remember that 12v X 100 amps =120v X 10 amps = 1200watts.
But 12v/100amps =.12ohms and 120v/10amps= 12ohms.
jer
Last edited:
and what compromises to the design have had to be made to achieve high Xmax.
Something has to give !
Will these other compromises affect the audio performance of the driver?
You and I agree, just jabbing you a bit with davy's radar. (which I still think is way cool....)
I changed my ss15 default high pass to 32hz, 24db/oct just for this exact reason. I had been using 48db/oct high passes until we went through (in another thread) the actual excursions with different slope/type of high passes. This allows some good SAFE excursion below 40hz in the ss15, and makes me feel a bit better because I can kick up some dnb and watch the cone and verify that the high pass is actually doing what I want it to do. Because the motor is hanging in the mouth pulling cool air in on each stroke, and the high pass is actually allowing excursion to happen -- I've not been able to feel any heat of any kind in the driver frame. In a FLH with 3015lf, I was able to elevate the rear chamber to +20 degree easily, and have obvious heat in the driver frame.
So do I
An advantage for cooling and a means for testing and monitoring the action of the driver are pluses over a buried driver.
That makes about as much sense as saying the longer the stroke of the piston in an engine, the better. Since a speaker is an electric motor, the "stroke" has to be designed in--same with engines.
In engines, if the stroke is getting too long to be efficient, they increase the diameter of the piston. Once that gets large enough to be inefficient--they add additional cylinders. Subwoofers are the same way--they are motors.
I'm very thankful my 2.4 liter engine has 4 cylinders--I shudder to think of the mess it would be to take a single piston out of my engine and keep increasing stroke until it made the same power. Besides the huge vibration, the height of the thing and the starter motor would be the size of a pony keg...
You can go with a larger diameter piston, shorten the stroke--go with two smaller pistons with shorter strokes or... go with four.
Get an 18 Sounds 21" sub that has 14mm of Xmax--6" diameter voice coil and two suspension coils. It is very efficient because the diameter of the thing is huge! It is all about balance. Better hurry and buy that sub, it runs a neodymium magnet and has not shot up in price yet--only $605.
With increased excursion comes increased diaphragm velocity.
Klippel data indicates that most drivers would be running at less optimum extreme portions of performance envelope.
It depends... two 12" drivers moving 12mm is generally better than one moving 24mm. (less power needed, less distortion, less acceleration / stress)
If you have the space, more drivers is better, more xmax is for space limited applications like Boom cars.
It would be interesting to see the SS15 tests done at 63 volts then, though that may be too much power below Fb.
The ported dual Lab 12 (7.76 cubic feet gross, 36 Fb) puts out about 120 dB at 40 Hz with about 3% distortion, at 49 volt input, one meter.
The thread below has distortion and output measurements Phil Lewandowski and I took over several months .
It compares many BFM and other small subs appropriate for small vehicles and venues.
JBL SRX718 - Basic Distortion Measurements:
PSW Sound Reinforcement Forums: LAB Lounge => JBL SRX718 - Basic Distortion Measurements
Page 6, message 564268 has a compilation of all the subs tested for distortion and output
To avoid any potential mechanical damage, power was reduced at in some cases by quite a bit below Fb or Fc.
Testing my own C horn was an eye-opener, I went to the Dual Lab 12 ported design after using the Chorns , loaded with EVX-150A or Lab 12s.
The Chorn at full power probably could put out similar output to the Dual Lab 12 at 40 Hz (120 dB at 49V) but it’s distortion would be 150% or so compared to 3%.
Reaction to distortion is subjective, using sine wave testing I really did not find up to 15% objectionable, but by 30%, I found it pretty bad.
Mixing rock and blues, I tend to like the 40-60 Hz bottom to be up, this meant with the Chorn (or most small bass horn cabinets) which rolls off steeply below 50 Hz, that the most power was being used down where the speaker makes the most distortion.
All though small horns in multiples flatten out in response, distortion does not change at a given drive level.
The LF distortion creates harmonics extending well above the sub range, for years while using the C-horns I had been pulling out 125-160 Hz from the main EQ due to hearing harmonic distortion (gaak) that ended up making that range “too hot” This in turn really had the opposite effect, only reducing the bass fundamentals in that range , since distortion can not be equalized out.
The Lab 2x12 ported sub has prodigious clean output for their size down low, but run out of steam in the 50-60 Hz range.
Hence the desire to build what turned into the Keystone Sub, which has the low end, and a lot more upper bass output in the same size as two of the dual Lab 12s.
http://www.diyaudio.com/forums/subwoofers/185588-keystone-sub-using-18-15-12-inch-speakers-2.html
Art Welter
Although it is true that “there is no replacement for displacement” with low frequency drivers, which also holds true in engines (a speaker is a motor, not an engine) the comparisons fall apart at extremes.
An engine can use a long stroke, large piston diameter, and slow RPM to produce power from a given amount of fuel, or a small piston, short stroke and high RPM to produce the same amount of power.
A speaker has no choice in “RPM”, it has to reproduce the frequency of the signal (music) presented to it.
This is far more complex than a piston moving up and down, to sound accurate the speaker has to both go in and out at the lowest fundamental frequencies while tracking many frequencies and their harmonics .
Big heavy cones may be good for the heavy lifting, but harder to get the detail from, smaller cones tend to be the opposite.
Too much moving mass (huge, heavy cones) requires to much motor to control it, while smaller cones are limited in displacement, a suspension that offers anything like the stroke of an engine would reduce the actual cone area to a fraction of the diameter, thus actually reducing useable displacement.
Although better cone materials and neodymium magnets have changed the game somewhat, a single huge cone will never have the detail available from a number of smaller drivers moving the same amount of air.
Art Welter
- Status
- Not open for further replies.