George (on Topic Only)

Ok, so how does Vas compare to EBP/Qts? Although it may not tell you what type of enclosure it will work best in, it will get you a very good idea if it is a large box woofer or not. In fact it looks to me to be a little more accurate than the other two. What do you guys think?

 

Sorry geo, i made my last post before i even knew you responded to my question to Seth. My bad. I typed out the post and walked away from my computer before submitting it. I understand what you're saying. Im comfortable with it more now.

BTW, will the "geolemon" equation actually work? Lmao @ calling it the "geolemon"!

 

Sorry geo, i made my last post before i even knew you responded to my question to Seth. My bad. I typed out the post and walked away from my computer before submitting it. I understand what you're saying. Im comfortable with it more now.

BTW, will the "geolemon" equation actually work? Lmao @ calling it the "geolemon"! [/b][/quote]
The "geolomon" equation would work..but it'd have to look something like this:

Qts = ((Fs/EBP) x Qms) / ((Fs/EBP) + Qms)

And even if this were to look right, we'd be basing one form of interpreting alignment (Qts), using another (EBP)...I dunno how effective this would be....but obviously the factors making up the Qes (BL, Re, Suspension, and moving mass) become the most important when evaluating alignment and the use of sealed/ported enclosures..(of course, depending on whether or not you use EBP or Qts)..do I make sense...it's 2:30am?

Since Qes really has not been specifically talked about....can someone enlighten me as to how BL, Re, Suspension, and moving mass accurately produce the loss of the motor...I want to assume that BL and Cms are inversely related to Qes..and in that case, would a higher BL in turn result in a poorer alignment? (this kinda makes sense to me..looking for clarification)...but it's bedtime, and my assumptions are nearly 100% incorrect before hitting the sack..

Oh yeah...let's see if we can relate the next few posts to Qes...that way we all have an idea what it effects, and what affects it....then we can move on to another variable

This is fun.....I think a lot of us are enjoying this...keep it going!!

EDIT: Nevermind...answered my own question about Qes...how about someone give me a real quick definition of Fb...

 

Right, like you say.. sort of apples vs. oranges.

Vas will help you figure out how large of an enclosure the sub will need... smaller Vas = smaller enclosure.
Likewise, to that end, Qes and/or BL will help you figure that out... a stronger subwoofer (lower Qes value, or higher BL value) = smaller enclosure.

Scroll back up this thread for Dan's post on this, too. B)

On the other hand, neither EBP or Qts give any regard for how large an enclosure will need to be... simply indicators of what type of an enclosure you'll want to be using.

I think it is interesting to note that motor strength is not only a factor in determining enclosure size, but also enclosure type.
So.. a subwoofer with a beefy motor will not only tend to lean towards small enclosures, but also ported enclosures... unless other parameters compensate away from that.

That alone actually illustrates why "bigger" or "stronger" isnt' always better.

 

That Shiva sounds like a great example where Qts shows a strong driver, strong motor... which usually would drive Fs up higher.

My guess is that either by Mms or Vas (or both), they have driven Fs down to something quite low - low enough to actually offset the low Qes, to lower EBP more into the "sealed" range.

So yes, if you were only looking at Qes (or Qts, which isn't as Qms-influenced), without regard for Fs, you'd throw the Shiva in a ported box.
And it probably is a nice ported box driver also, though...
Generally when Qts shows one way, and EBP shows it the other way, it's one of those "middle of the road" cases.

 

Nis,

Buckle physically displaces the cone, via a stepper motor system. We attach to the cone/former, and sweep it through its range of motion. There's a force gauge attached to the unit, and we measure the force back. Since we know the position of the driver (we move it there), and we're measuring the force from the driver (with the force gauge), we know the stiffness of the suspension at every excursion point.

Aside: note that we can do a fast sweep so as to get a true Cms curve, or we can do a slow sweep and measure the relaxation of the suspension at each point and also calculate the Rms. Systems that measure the driver at a static point are inherently inaccurate, since the spider will immediately begin relaxing as soon as you stop moving; the Cms increases constantly, leading to an inaccurate measurement.

Then we do the same sweep again with a fixed DC current through the voice coil (typically 1A). Since we know the force of the suspension, we now measure the force of the suspension+motor, and subtracting out the force of the suspension yields the force of the motor. And because we measure at 1A, we directly read BL (force gauge reads in Newtons, we have 1A, BL is in units of N/A, thus the BL is directly what the force gauge reads).

Note: because we can control the current, we can measure not only the BL curve, but the flux compression of the driver as well. BL curves change with applied current; as power goes up, BL can go up on one half of the stroke, and down on the other half of the stroke, leading to further asymmetries. We can measure those effects - our test system can measure up to 30A of current, which would translate to ~900W into a 1 Ohm load.

Then we can sweep the driver again with a constant excitation sine wave, and with a basic impedance divider calculate the inductance over all positions. We can measure the inductance at any frequency desired, from 1 Hz up to 100 kHz, so that we can get a picture of the inductance versus position versus frequency.

Then we take all the measurements, and crunch them down to give us the result. The results are highly accurate and repeatable, and do not suffer from noise during testing (DUMAX), creep from static positioning (DUMAX and Klippel) or small sample sets at the extremes (Klippel).

We'll be talking with Dr. Geddes, Dr. Klippel, and Mr. Clark at the ALMA meeting about proper measurements of subwoofers, as well as correct modeling parameters (there's currently a body of thought that T/S does not apply to subs, since it is strictly a small-signal set of data; we believe it does, just that you cannot use a single T/S number for each parameter. Rather, you need to use a simple equation for each parameter and fill in the appropriate stimulus. We're going to advocate power as the primary stimulus/variable, but excursion would also work).

Dan Wiggins
Adire Audio

 

Other measurement systems:

DUMAX - My understanding is that the system is inherently sensitive to airborne pressure - noise or door closings/openings. Because there is a microphone in the system (whether it's a real mic or the driver is immaterial - there is an acoustic pickup device) and the stimulus is a pure tone, it can be disturbed and as such the data can be corrupted. Additionally, because the driver is held at one position, the Cms relaxation can affect your measurements, and the holding pressure must be constantly remeasured to back out the Cms value from the current SPL measurement level.

Klippel - A pseudorandom noise signal is fed to the driver. It is wide in bandwidth and amplitude. A laser monitors cone position, while current and applied voltage are measured into the driver. As a result the position of the driver is measured at known times, and the drive signal (current and voltage) is also known at those times, and you can statistically collect a lot of points relating position to signal. Then these samples are curve fit with trial BL/Cms/Le curves over power to find a suitable solution set. Upside is that it is very noise-tolerant (just like an MLS test signal), and is inherently resistant to Cms relaxation. Downside is that on high-stroke drivers you often times do not get enough samples at the extremes to get an accurate fit, and as such the fit does not converge at all.

Dan Wiggins
Adire Audio

 

Thanks for explaining the measurement systems Dan. I know a lot of people on here will probably print out your reply and study it for a few weeks.

 

Clarke,

Fb is the resonance of the box in Hertz.

 

This kind of peaks my insterest. I know that in home audio, they try to push amplifiers that are higher current. But in doing so, in a system utilizing a driver with a very parabolic BL curve, would the higher current actually increase distortion levels higher than an amplifier providing the same power, but more voltage? If so, would the differences be audible?

Also, would that also mean that wiring into lower impedances increases distortion as well? Is that why we don't see many higher frequency drivers using lower impedance coils (2 ohms)?

I remember when you visited, you mentioned that it is better to wire a dual voice coil subwoofer to each channel of a 2 channel amplifier instead of parallel bridged. I know that the power output is the same. But is the current higher in the bridged mode, causing higher distortion levels?

 

ill just add a bit more to stevens questions. i remember reading that tube amps werent good for low frequency drivers due to how they power them, but i cant remember if they were oriented toward mainly voltage, or current...though current is what im leaning toward.

i guess i didnt tie it in THAT well to stevens post, but he brings up some good points.

eric

 

Tubes are voltage oriented, they are transformer coupled to get some current out of em

On the topic of porting drivers based on specs... I haven't really liked a ported driver until I ported my Vifa 8s, they have a 0.25 Qts and they sound really great ported. Of course the EBP is really high on them too, so they scream ported all around.

 

Back in the "old days" of speaker ideology the Qts limits for ported boxes used to be divided at .40. Below that was suitable for ported, and above that was sealed only. With drivers that have normal parabolic curves and/or "normal" motors (i.e. underhung or overhung), I still use the .40 dividing line. But I don't just look at one spec to determine whether or not the driver should be ported or sealed, however.

 

I think this is often confused with how electronic componentry or the designed behavior of an amplifier behaves as you approach the limits of the componentry contained within that design.

Within the scope of any single amplifier (JL's RIPS amps being a potential big exception), as you wire the amp to a lower impedance load, the amp makes more power. This is because the amp is simply a voltage source, putting out a certain voltage out it's speaker outputs... the load you wire up determines how much current will flow through the speaker and the amplifier.
It's simply Ohms Law that states that the lower the impedance of the load, the more current will flow through the circuit. And given that the voltage is the same regardless, and power is current times voltage, the increasing current makes for increasing power.
But the more current that flows through the amp, the more stressed the components within become, as they can only handle so much current flowing through them.
Consider each of them in this regard to be like fuses... they'll melt once a certain amperage is reached. :blink:
This is why your amp has a minimum impedance rating... it's not really because of the impedance, per se.. it's because of how much current will flow as a result of the impedance hooked up.

At any rate, the more power the amp puts out, the more stressed it becomes..
All of it's specs begin to suffer... such as distortion.

That being said...
Yes, in the scope of a single amp, as you drop it's impedance, it's distortion will suffer.
However, if you designed an amp to be able to drop to very low impedances, and it's componentry contained was actually designed to drop that low and still stay within design parameters, then there's no reason why you couldn't build an amplifier that ran nice and clean way down low, 1 ohm, 1/2 ohm, whatever.
But amplifiers like that would potentially cost more money to make, also, to handle the current calmly.

 

True, but I wasn't talking about the distortion from the amplifiers. As Dan said, higher current amplifies the nonlinearities within the motor (BLI). So keeping the voltage the same, but increasing current would cause more distortion within the sub.

Now I remember that class A amplifiers are considered higher fidelity units because of the higher current they produce. This higher current of course limits their output power for the reasons geo lists above. But what Dan says seems to contradict this thinking. Of course I know which viewpoint I would agree with. After all, Dan doesn't advocate the sonic superiority of pure platinum wire, or using highly expensive ceramic holders to raise the speaker wire off the floor. <_< :lmfao:

 

Well, consider that BLi is the formula for the force at which a loudspeaker's motor will push and pull on the cone... which is also proportional to the output of the loudspeaker, and it's excursion, in a sealed or IB application (really with ported too, but you need to factor in the port's counter-forces along with it).
F = BLi (or, as on Adire's "woofer speed" tech paper, ma = BLi, and F = ma of course ).
So what I'm saying is, as you increase the amount of current through a given woofer, you'll increase the force of the motor, and therefore it's excursion...
And that does increase distortion potential...
But is that because you increased the current flowing, and therefore there was more distortion? Or is it because you increased the current, and therefore caused higher excursion, which increased the distortions? :unsure:

After all, you are stroking that voice coil for a longer distance, at higher excursions. Even in a ported app, or a smaller enclosure to compensate that back down to where it was originally, you have more current, more heat, and distortion possibilities tied to that possibly.

And for any speaker that doesn't have a variable impedance (and I don't know of any ), the only way to get more current to flow is to increase the amount of power going through it... or rather, to increase the voltage of the signal being fed to it, and through that same impedance (since it's not variable ) more current will flow... therefore, more current * more voltage = more power. B)
Just my thoughts on that.

 

Ok, then let's simplify this. Let's say you provide the exact same amount of power to the same woofer. The difference is that the voltage/current ratio is different. In other words, the first trial has more current, but less voltage than the second trial. What differences will we see? From what I read in Dan's post, distortion levels will be higher in the first trial even though the power is the same.