AVS Forum

AVS Forum (http://www.avsforum.com/forum/)
-   DIY Speakers and Subs (http://www.avsforum.com/forum/155-diy-speakers-subs/)
-   -   Help with sealed enclosure for LMS Ultra 5400 (http://www.avsforum.com/forum/155-diy-speakers-subs/1574938-help-sealed-enclosure-lms-ultra-5400-a.html)

GizzeGutten 06-21-2014 06:36 PM

Help with sealed enclosure for LMS Ultra 5400
 
2 Attachment(s)
Got two LMS Ultra 5400 and a FP14k clone from Sanway. Using WinISD to estimate but have some questions as I'm a noob.

TC Sounds recommends sealed enclosures of 2.1ft3 (Qtc=.744) and 4ft3 (Qtc=.591).

Planning to build one of these or something in between, and then do a HPF at ~11Hz.

Have plotted these two in WinISD, put signal at 2kW and looking at the excursion chart [after adjusting for (no) leaks (Ql=50) and (no) dampening (Qa=120)] we see this.
Attachment 130066

As can be seen, there is plenty of excursion left with the 2.1ft3 even when pumping 2kW into it, so question is, is it possible to EQ it even more without risking melting the coil?

With the 4ft3 there is not a lot of excursion left at 10 Hz, but as can be seen below, it has more SPL.

Attachment 130074

Which of these would have the tightest bass (will mainly be used for music), and will the difference be very noticable? Thinking about doing xover at around 60 Hz or below.

BassThatHz 06-22-2014 01:38 AM

A bigger box is almost always better.
All a smaller box gains you is less power efficiency and a higher -3db point that must then be EQ'ed with even more power.

There is no such thing as tightness.

Electromagnetic fields travel at the speed of light.
So the length of the speaker wire (and voice coil wire) has practically zero effect on the speed of the field being established or collapsing.
So what causes the cone to move, at all?

Velocity from a start of zero after time t is:
V = A * T^2
Thus how "fast" the cone moves (velocity) is proportional to the acceleration.

Newton's second law:
F = M * A
Thus A = F / M

The acceleration is proportional to the force of the electromagnetic-system.
i.e. magnet size/strength and coil size/strength. (All else being equal.)

So a beefier subwoofer would only be marginally-faster if it also didn't have a heavier cone / coil; otherwise the same speed (excluding the above reason).

The electromagnetic field traveling at the speed of light will collapse near-instantly when the music stops, so the only other force acting on the
cone would then be the air box spring and gravity (if it is an up or downfiring cone) which has nothing to do with how expensive the subwoofer is, and nothing within it's ability to control.

As you crank the power up, the electromagnetic field gets stronger, and the cone accelerates more, thus the velocity increases, the cone moves a further distance in time T.
But the amount of time T the signal is applied for remains constant, the frequency doesn't change as the power goes up, just the distance the cone travels increases.

Think about it like this:
I give you a 150lb metal box and ask you to shake it back and forth from one side of your body to the other. You won't be able to do it very fast.
Now I give you a 1lb cardboard box and ask you to do the same. You'll be able to shake it back and forth very quickly.

Your muscle power remained the same (same magnet size and coil size / strength), but the mass changed.
Conversely, if you added an inhuman amount of muscle to your arms you'd be able to "shake the heavier cone faster".
How far it can be moved in time T is proportional to your effort applied (i.e. watts), not how much muscle you have (muscle just helps you move it easier).
The physics makes sense, it's just that people don't understand it.

I guess I did put that grade 12 physics to use after all!!! ;) LOL

GizzeGutten 06-22-2014 02:12 AM

Thank you for a very good and beefy reply, appreciate that! :) As I'm a noob, I've got a follow up question :)

What if I told you to pull on your shorts and dive into the water, then throw you a box with air inside and ask you to pull it up and down from the water. Then you will obviously have a hard time getting it down, but due to the air wanting to go up, you'd have a fancy time pulling it up. Also, let's say you gain more muscle strength at the instant the box touches the water on its way down and lose it again as it is deeply in the water ready to go upwards,.

I thought the same idea also applied to smaller enclosures, where moving the cone inwards would require more power, but then the air pressure inside the cab. would go up, and so there would be a springiness making the force outward bigger than the electromechanical force, hence making it move faster (at least outwards).

Thinking about it, since the added electromechanical force required to pull the driver in will be less than the springy force outward (as some of it goes into heat/friction), you get a less efficient sub. However, one would think that it would not take more time to pull the driver in, because more force is added, so then the cone would move equally as fast in as in a bigger enclosure, but due to the added spring force, it would move faster outward, giving you a gain in speed, but loss of power?

Is there a logical fallacy on my part, when I assume that you get more power at the instant the driver is pulled inwards?

If there is no such thing as tight bass, then why are there so many subjective reports about smaller cabs being tighter?

mbyrnes 06-22-2014 02:46 AM

A sealed sub box should be sealed so no air gets in or out. A small box when the driver pushes out creates a vacuum inside the box. As air pressure always equalizes, the box will pull the driver back into place. I would think a larger sealed box has less of a vacuum pull so it is easier for the driver to move inward and outward.

I could be wrong but this makes sense to me. The smaller the box, the more vacuum the driver has to deal with, making it more inefficient (why more power is needed).

I am sure this is one of the driver specs, but as I don't know much about those.

GizzeGutten 06-22-2014 03:02 AM

Oh, thanks, and you're right. The driver will always want to go back to its resting position, so you need an equal amount of force to push and pull. Some of that power will be wasted through heat, making it less efficient.

Quite puzzling that a lot of people rave on about smaller subs are being tighter then.

Also read this old article which relates to what Bass said. It's been taken down but is available through the web archives, so posting a relevant excerpt here.

---------------
http://web.archive.org/web/200804160...ut-subwoofers/
Myths: Subwoofers are fast / slow

More appropriately labeled Damping or Ringing, these concepts are really reciprocals of one another have nothing to do with speed, tightness, “boomieness” or any other misused and inappropriate term for subwoofers. Subwoofers, or rather bass drivers, all move at the same frequency when instructed to via an input single. The difference is really about the Q alignment of the system. There are many famous Q alignments which produce various frequency responses, but beyond the complex mathematics is a fundamental principal of force and acceleration and the driver will respond to a sinusoidal wave at various accelerations depending on the moving mass and force that the voice coil and motor generate on the cone. Therefore any driver can be faster or slower depending simply on the voltage! It makes little sense to call any driver faster or slower.

Damping or Ringing is really what we’re after and the amount of either is really a function of system volume along with the electro-mechanical damping factor of the driver. For example, in a sealed box system, as the volume of the cabinet becomes small, the internal pressures increase when the driver pushes in and out. This pressure is a force which, not nearly as strong as the electromotive damping force, works in the opposite direction. Contrary to intuition, higher internal pressure (which we tend to associate with tightness or stiffness) decreases damping and promotes ringing at one particular frequency (Fc in the case of a sealed box). The pressure from the air inside the box works against the driver’s natural damping factor of 1/(Qts). When the pressure becomes large relative to the motor’s damping factor, the driver will ring more and cause a peak in SPL at the given resonate frequency (Fc). This tends to be somewhere around 40-60Hz in a given sealed box, but could be outside that range under abnormal circumstances. This peak is ill desired and is accountable to the proclaimed “boomy” sounding subwoofers which tend to lack clarity, good transit response and dynamics. However some people prefer some ringing because it provides a natural boost in a very audible frequency band. Likewise, in a larger box, the Q will decrease and the ringing and SPL around that frequency will too, but the low end will open up and you’ll have more deep bass. This tends to sound better and more controlled.

On the flip side, over dampened drivers tend to have poor low frequency response and require equalization to boost the low frequencies. They tend to work better in vented boxes where their larger motor force factor (BL^2/Re) is put to good use with a resonator which then makes the low end much more efficient with its increased displacement. Likewise, drivers with high Qts will work better in sealed boxes and should be exempt from being used in a ported system without careful consideration. When high Q drivers are used in a vented system they will ring at the tuning frequency of the box (Fb in this case) and the “boomy” problem is considerably worse.
---------------

So basically what he is saying, is that BassThatHz is right (not that I didn't believe it, but wanted to know why), that Q is sort of a multiplier of the ringing of the Fc, and that with sealed enclosures you'll always have some resonant frequency, but by making it bigger the Q will be lower, hence the ringing less obvious, and the perceived tightness as well. It's not that small sealed subs are tight, it's just that they have a boost in an audible frequency and this makes them seem tighter?

GizzeGutten 06-22-2014 03:22 AM

My God, had an Eureka moment
 
1 Attachment(s)
My God... Just had an idea of making both enclosures ported and tuned at 15-20 Hz, then use a gasket and a panel to seal it up if it doesn't pan out... Just need to figure out a volume that would work well for both closed and open then.

Good thing I've got this Excel spreadsheet running, just needs to be modified a bit, lol.
Attachment 130578

BassThatHz 06-22-2014 10:53 AM

Compared to the gjallerhorn box, only below 11hz is the sealed louder.
Above 11hz, basically 1 gjallerhorn = ~3 sealed.

http://www.data-bass.com/systems

GizzeGutten 06-23-2014 03:10 AM

3 Attachment(s)
Quote:

Originally Posted by BassThatHz (Post 25168650)
Compared to the gjallerhorn box, only below 11hz is the sealed louder.
Above 11hz, basically 1 gjallerhorn = ~3 sealed.

http://www.data-bass.com/systems

Hehe, but that won't be built. :D

Even though you have explained the physics and I'm not one to want to go below 11 Hz, I still have problems going ported, as I'm afraid it wont sound as good. Call it irrational, but it might be that I do like that resonant frequency in sealed enclosures.

The reasoning behind this project is pretty much like this:
1) For sealed I don't want exceed excursion until after 10 Hz pumping 2kW into it. This is because of the HPF at around 10Hz, so anything below that is not important (at least not now, but as it might be in the future the enclosure is at 4.31ft3, Qtc=.577 and it will not go above excursion until 5Hz.)

Max SPL @ 2kW
Attachment 131810

Excursion @ 2kW
Attachment 131818

2) For ported I will have to tune it and it'll mostly be best at and above the tuning frequency (as I've understood it). As it'll mainly be used for music the ported enclosure will be superior as most of the content in music lies above the tuning frequency.

Have yet to find a good design for the ported enclosure, but the basic idea is to have a rectangular port at the bottom and in case I don't like it cover it up with a panel & a gasket. This is the basic idea, but only a simple draft.

Attachment 131826

May I ask your reason for going sealed? Guess its because you love those single digits, haha.

GizzeGutten 06-23-2014 04:27 AM

1 Attachment(s)
A second draft of the concept. Working on the constraints of the port in relation to the proportional size, as I'd like it to have these proportions even though it can be bigger or smaller etc. Using the golden ratio makes for a more visually pleasing sub!! OCD baby ;)

Attachment 131881

Are there any caveats of doing it this way?

splotten 06-23-2014 04:46 AM

You should incorporate the high pass filter in the model. Gives you a better representation of cone excursion (and vent speeds in a ported box).

In winisd:

EQ/filter tab > Add > Highpass > Choose filter you want

GizzeGutten 06-23-2014 04:59 AM

Thank you for participating in this thread as well, splotten. I believe it was you that recommended HPF @ 11 Hz with 12dB/Oct in my other thread about integrating LMS with miniDSP and FP14k. May I ask why 12dB/Oct and not something steeper? When going with 12dB/Oct as opposed to 48dB/Oct the roll off starts earlier than the chosen frequency, so I lose SPL > 11 Hz too.

The reason why I've not used a HPF in the model, is because I might find that I want frequencies lower than the HPF @ 11 Hz later. Is this poor reasoning?

GizzeGutten 06-23-2014 05:24 AM

1 Attachment(s)
Thinking about double configurations as well, this type of deign can work well by stacking the units. :)

Attachment 131977

Then you can have both closed, both open or one of each and do xover different for each of them :) If the front panel covers for the ports were made in aluminum and had a gasket, you could thread them, then screw them in place with wing nuts from the rear, making them easy to put on and remove. :)

Almost want to buy two more drivers to get a double set of these, haha.

splotten 06-23-2014 05:44 AM

Quote:

Originally Posted by GizzeGutten (Post 25184457)
Thank you for participating in this thread as well, splotten. I believe it was you that recommended HPF @ 11 Hz with 12dB/Oct in my other thread about integrating LMS with miniDSP and FP14k. May I ask why 12dB/Oct and not something steeper? When going with 12dB/Oct as opposed to 48dB/Oct the roll off starts earlier than the chosen frequency, so I lose SPL > 11 Hz too.

The reason why I've not used a HPF in the model, is because I might find that I want frequencies lower than the HPF @ 11 Hz later. Is this poor reasoning?

I dont think I have recommended any special frequency or 12/Oct, but i have seen others use 10Hz. Incorporating the filter in the model is precisely so you can see what it does as opposed to guessing what it might do. If you think a steep filter looks better go with that or at least try out the different options when you have built the subs.

I think you should always try and model the different options regardless if you go with high pass filters or not. You can have as many models in winisd at a time as you like and compare directly. Big or small box, filters or no filters, vented or sealed, ect, ect.

EDIT: I actually have no idea why 11Hz seem popular, but I have noticed that a 11Hz 2 order HP will usually give you max excursion at 20Hz. Maybe the reason is simply that it looks good and sorted in the winisd graphs :-)

EDIT EDIT: I am not really recommending for or against a high pass filter. I'm just saying that if you want it you should include it in you model. Thats all...

GizzeGutten 06-23-2014 07:39 AM

6 Attachment(s)
splotten, thank you. :)

Been playing with HPF at 10 Hz with 2nd and 4th order filters now.

Excursion @ 2kW into 4.53 ft3 with Fsc=35.34Hz and Qtc=.568
Attachment 132057
Green, no HPF
Yellow, HPF = 10 Hz, 2nd order (can take up to 2.9kW)
Purple, HPF = 10 Hz, 4th order (can take up to 2.5kW)

If xover is done at 60 Hz, that would be like adding a 2nd or 4th order LPF.

Excursion @ 2kW with HPF and LPF.
Attachment 132073

Green, no LPF
Yellow, LPF = 60 Hz, 2nd order
Purple, LPF = 60 Hz, 4th order

The transfer function after these filters has been applied.
Attachment 132081

Not exactly flat, haha.

Questions
1) Is my goal when estimating and setting filters to get the transfer function flat between the xover point and around the HPF point?
2) Not sure how the 'System input power' under the 'Signal' tab in WinISD works: Is that the maximum I want to give it, even including filters that boost it, or is it just to see how the whole system behaves if you give it that? Asking because when I drive the current sub super hard without any filters I'm at around 800w. Should I set this at 1w, add filters I want to get the transfer function flat and then increase the input power to see how much I can feed it?

Adding Linkwitz-Transform
I followed this process: (a) set the input power to 1w. (b) set the LT-filter depicted below, (c) went to the excursion chart and (d) raised the input power until I were close to the linear excursion constraint.

This is the LT params.
Attachment 132193

This is the new transfer function
Attachment 132209

This is the excursion
Attachment 132201

Purple, can only take 270w before it goes above excursion constraint.
Yellow, can only take 320w
Green, no filters, still 2kW

Last questions
3) What does the 'System input power' tell me? Is this the maximum of what the system can handle with the filters, even when I have to use additional power to use the filters or what? Me no comprende!!!
4) How does the above look to an experienced set of eyes?

GizzeGutten 06-23-2014 09:26 AM

This is an excellent thread for people wondering about how to EQ and optimize models in WinISD. In particular the post below:

http://www.avsforum.com/forum/155-di...-question.html

MBetnz explains how to optimize a sealed enclosure to any given frequency with a flat response.

Quote:

Originally Posted by MBentz (Post 23605496)
This is where the MaxSPL plot will help you out....it basically tells you what the maximum SPL is for each frequency - which is usually limited by excursion at the lowest frequencies (unless your enclosure is crazy small).

Maybe it'd help to put it in step form:
  1. Pick the bandwidth you want from the system.
  2. Go to the MaxSPL chart and find the lowest point within that bandwidth. Draw a horizontal line (or just remember the SPL number).
  3. Go to the SPL chart and modify the input power until the highest part of your passband matches the lowest point in the MaxSPL chart.
  4. Modify your filter until your filter shape fits within the MaxSPL limits (you may need to adjust your input power since you'll be boosting at some frequencies)
  5. After this is completed, the Apparent Power will tell you how much amplifier you need, and the Excursion chart shows you how things approach the limit.


GizzeGutten 06-23-2014 11:41 AM

By following the above procedure this is the result of a 4.53 ft3 with Qtc=.577 and Fsc=35.91.
.
  1. Xover is @ 60 Hz
  2. LPF, 4th order at 60 Hz
  3. HPF, 4th order at 10 Hz
  4. Linkwitz-Transform, fp=20 and Qp=0.707
  5. System input power = 227W
.
Question:
The 'System input power' is at max @ 227W. I can't go above that, because the excursion will pass the constraint. In the thread I link to in the last post, something is mentioned about the 'System input power' being mistakenly labeled, and that WinISD converts this numbers into Volt rms and uses the Vrms to calculate and estimate.

Is this because the voltage applied is constant (if you run a sine wave), but as the impedance varies with frequency so will the current?

Is this what 'Amplifier apparent load power VA' shows you? The power at any given frequency given a constant voltage and varying current due to impedance changes?

If so, what is the constant Vrms applied to this estimation? Is it:

System input power = 277W
Impedance @ 20 Hz = 7.072
VA @ 20 Hz = 1003.270VA

1003.270VA = V/7.072 * V => V = sqrt{1003.270 * 7.072} = 84.23Vrms

Is this correct? The constant voltage to this system is 84.23Vrms?


Maximum SPL
http://i.imgur.com/ERaoyLv.png

SPL chart with filters
http://i.imgur.com/eKwUZq2.png

Excursion chart
http://i.imgur.com/rpJRSQ3.png

Amplifier apparent load power VA
http://i.imgur.com/vGgoakl.png

GizzeGutten 06-23-2014 11:51 AM

Haha, all other threads in the DIY forum is buzzing with activity, but my threads always end up dead in the water, lol.

What am I doing wrong? Am I coming across as a complete idiot far past the point of being helped or what? LOL.

splotten 06-23-2014 01:10 PM

Im not sure you are doing anything wrong. Your questions are just not that easy to answer. You are certainly not coming across as an idiot. Quite the contrary in fact.

I believe your calculation is correct. Also I believe that the reasoning behind doing it is good. The voltage is the only thing that doesn't vary with frequency.

Obviously 227W to reach Xmax with that driver is WRONG, so in essence it doesn't tell you anything. I believe winisd is like that because most dont know how to convert between power and voltage, so when you enter say 100w and the driver is 3.2 ohm winisd will internally convert to 17.9V, However your EQ is not reflected in the system power field so you have to go to the apparent power graph. That is why the 227W is so far off when you apply a LT.

The better way to do it is as you did by calculating the needed voltage from the impedance and apparent power graphs. Once you know that you can compare with the intended amplifiers maximum voltage to see if it has the needed voltage and it can point you in the direction of where to set the limiter.

EDIT: You should look up the difference between (true) Power and Apparent Power. They are related but they are not the same. The Apparent power graph shows the -apparent power- at any given frequency given a constant voltage and varying current due to impedance changes.

MBentz 06-23-2014 05:04 PM

What specifically are you trying to do with the answers to your questions?

The true insight into what is happening doesn't come about by knowing the steps of a calculation...so what does 84Vrms meant to you? Or what are you going to do differently if there is a different voltage happening at different frequencies?

What is your EQ doing to the voltage coming out of your amplifier? (Hint, it's increasing the voltage at some frequencies, but not at others...)

BassThatHz 06-24-2014 12:39 AM

What I do is build the biggest box I can get away with, and then set the gains of my amplifier such that: the max output voltage is at the excursion limited frequency with a 0dFS sinewave.
You perform this post EQ; while doing EQ keep the volume at a reasonable excursion to avoid problems.
This ensures that it is not possible to bottom out the system, while gaining maximum efficiency.
It should be noted that extreme EQ boosting might change the excursion limited frequency, this can be determined by running -6dFS sweeps with the amplifier turned all the way down to minimum and then slowly increased until clipping or excursion becomes dangerous.
That then is your correct gain structure, and maximum subwoofer output (maximum "HOTNESS" that is possible to set without destruction).

In this way, the box design becomes irrelevant.

That's not saying that modeling the peak excursion to the peak watts isn't a bad idea. But doing so is still no guarantee of protection from a bad gain structure with too many watts.
Also, I'm certain an LMS can be fried at 60hz well before xmech is reached, so there is the high frequencies to consider as well.
From what I'm been able to tell, if the cone is warm, then the coil is almost ready to fry. If it feels like room-temp metal, then you haven't likely reached max-thermal yet.

GizzeGutten 06-24-2014 03:39 AM

Edit: When the below was written I had misunderstood the concept of Vrms. I thought it was the average voltage over a ("long") period of time, not that x Vrms AC is the mean of one period and is equivalent to x volts DC at any particular moment in time. Now I understand that for any given value of Vrms, you also have a peak which will always be higher than the rms value. Given this, I understand how Vrms can change for discrete frequencies when using EQ to boost, because it is not an average voltage for the whole system.

However, I'd still like replies to the below if anyone care to give them.

Quote:

Originally Posted by MBentz (Post 25205689)
What specifically are you trying to do with the answers to your questions?

The true insight into what is happening doesn't come about by knowing the steps of a calculation...so what does 84Vrms meant to you? Or what are you going to do differently if there is a different voltage happening at different frequencies?

What is your EQ doing to the voltage coming out of your amplifier? (Hint, it's increasing the voltage at some frequencies, but not at others...)

Hi! The question was asked to understand two things, (1) how 'Signal input power' and impedance relates to VA (to better understand VA), and (2) to get a ball park figure of where to set the voltage clip limiter on FP14k to avoid over excursion. Should have mentioned this, as it was a big part of the last thread I had going. If the number had been >~120Vrms, I think the design could possibly damage the driver by frying the coil, so wanted to make sure that it was below the ~120Vrms'ish limit.

When I take three pair of data points (impedance and VA) and calcultate the voltage on an estimation which has active filters the V is different for each pair, so I was thinking about why last night. Now that you say EQ increase the voltage at some frequencies I understand that it's the filters :) So then I went ahead and made an identical model with no filters and calculated the voltage from three new pairs of values, and guess what, the V is different now too!!?

What am I calculating here? If it is the Vrms, why does it change with frequency even though no filters are added to the model? I thought Vrms was a way of calculating the mean, so it's weird that it changes with frequency, haha.

If you think I'm heading for a dead end with these questions, please say so. Have a bad habit of over complicating things, and especially things I don't understand. :D

Quote:

Originally Posted by splotten (Post 25199449)
EDIT: You should look up the difference between (true) Power and Apparent Power. They are related but they are not the same. The Apparent power graph shows the -apparent power- at any given frequency given a constant voltage and varying current due to impedance changes.

Only part of your quote, since I've responded implicitly to you above :) Did look this up here (in layman's terms), and if true power is the same as real power, then the only difference is the power factor which is a number between 0 and 1, hence always making apparent power bigger or equal to real power.

Yes, that's was what I assumed too, but when I calculated more values and saw that the answer was not the same for all frequencies even w/o filters, the assumption broke down and I don't understand why the voltage varies... :)

GizzeGutten 06-24-2014 03:50 AM

Quote:

Originally Posted by BassThatHz (Post 25214481)
What I do is build the biggest box I can get away with, and then set the gains of my amplifier such that: the max output voltage at the excursion limited frequency with a 0dFS sinewave.
You perform this post EQ; while doing EQ keep the volume at a reasonable excursion to avoid problems.
This ensures that it is not possible to bottom out the system, while gaining maximum efficiency.
It should be noted that extreme EQ boosting might change the excursion limited frequency, this can be determined by running -6dFS sweeps with the amplifier turned all the way down to minimum and then slowly increased until clipping or excursion becomes dangerous.
That then is your correct gain structure, and maximum subwoofer output (maximum "HOTNESS" that is possible to set without destruction).

In this way, the box design becomes irrelevant.

That's not saying that modeling the peak excursion to the peak watts isn't a bad idea. But doing so is still no guarantee of protection from a bad gain structure with too many watts.
Also, I'm certain an LMS can be fried at 60hz well before xmech is reached, so there is the high frequencies to consider as well.
From what I'm been able to tell, if the cone is warm, then the coil is almost ready to fry. If it feels like room-temp metal, then you haven't likely reached max-thermal yet.

This is really excellent and practical advice, just wonderful, thank you!:)

Do you run measurements with REW as you increase the SPL to see when the distortion starts going through the roof? At what point do you start being concerned?

Using miniDSP's UMIK-1, not sure how reliable that is for the low frequencies, so have to check it out.

splotten 06-24-2014 07:16 AM

Quote:

Originally Posted by GizzeGutten (Post 25215418)

When I take three pair of data points (impedance and VA) and calcultate the voltage on an estimation which has active filters the V is different for each pair, so I was thinking about why last night. Now that you say EQ increase the voltage at some frequencies I understand that it's the filters :) So then I went ahead and made an identical model with no filters and calculated the voltage from three new pairs of values, and guess what, the V is different now too!!?

What am I calculating here? If it is the Vrms, why does it change with frequency even though no filters are added to the model? I thought Vrms was a way of calculating the mean, so it's weird that it changes with frequency, haha.

If you think I'm heading for a dead end with these questions, please say so. Have a bad habit of over complicating things, and especially things I don't understand. :D



Only part of your quote, since I've responded implicitly to you above :) Did look this up here (in layman's terms), and if true power is the same as real power, then the only difference is the power factor which is a number between 0 and 1, hence always making apparent power bigger or equal to real power.

Yes, that's was what I assumed too, but when I calculated more values and saw that the answer was not the same for all frequencies even w/o filters, the assumption broke down and I don't understand why the voltage varies... :)

I have no idea why the numbers dont match. The formula is S = E^2 / Z and we know that the voltage E should be constant and independent of frequency when no EQ i applied. Then sqrt(Z * S) should give the same result no matter what frequency you look at. I can offer no explanation for the weird results but i believe your method is good.

GizzeGutten 06-24-2014 02:48 PM

Yes, it's weird given the definition of VA and the result following from Ohm's Law.

I restarted WinISD to calculate new values for a sealed enclosure w/o any filters, thinking that I might have mistakenly read the second data set off one of the prior estimations with filters, or that there were some bug in the app as I get error messages from the application from time to time.

This time the calculations were far closer together.

http://i.imgur.com/cy5RawK.png

Decimal separator in Europe is "," not "." . Furthermore, I can not get integer values for frequencies in WinISD when I move the mouse pointer to select different frequencies.

These are probably just variations due to rounding errors in the algorithm winISD runs off of. It seems plausible that they don't use a lot of decimals to make the app less straining on the CPU.

GizzeGutten 06-25-2014 11:21 AM

Feet for subwoofers
 
Have been thinking about what kind of feet to use. What do you guys think about Crowson's rubber feet? As they are intended to sit under furniture they should happily accommodate the weight of the subs. Also, they're only $10 a pop, far cheaper than the SVS' SoundPath Subwoofer Isolation System. What do you guys think? Will they fit the purpose?

Crowson's feet
http://crowsontech.com/_CrowsonTech/...e72aeebe71.jpg

SVS SoundPath

http://i.imgur.com/NvXowGx.png

GizzeGutten 06-25-2014 02:37 PM

Subwoofer design comming together
 
The enclosure design is beginning to emerge, and I seek your seal of approval. :D

After reading further, I've come to the realization that I'd like to go sealed, because of better transient response and group delay figures, as well as the ability to dig below into the single digits should that be a 'want' later down the line.

After BassThatHz wrote how he builds as big as possible then optimize gain structure etc post, I've not paid as much attention to the size as far as performance goes. However, I do have some physical constraints and set it up so that the ability of building two more identical subs are possible in the future. (Should never have come here, :D)

Specs (sorry about the weird numbers, but metric systems in Europe are different, so these have been converted).

  • Material: Dual layer 3/4" MDF and tripple baffle
  • 21.65 x 23.62 x 29.02
  • Net internal working volume: ~4.96 ft3 (might be a little less if I brace more than calculated)
  • Qtc = .552, Fsc = 34.33 Hz


Will using dual layer MDF be beneficial and necessary? If not, then the internal volume will be higher.


Concept draft

http://i.imgur.com/r36pL4f.png

Notice the perspective, making them look dominant, :D

Estimations using WinISD

http://i.imgur.com/OiZMkXi.png

One sub less room gain will do 111 dB @ 20 Hz. This might become

http://i.imgur.com/JiGYpzj.png

Group delay

http://i.imgur.com/KExXNo4.png

Excursion at maximum power handling of 2kW
http://i.imgur.com/UaEOPqQ.png

Maximum power handling w/o over excursion
http://i.imgur.com/b4kzFSp.png

Impedance
http://i.imgur.com/xyPCbKL.png

Amplifier apparent load power VA
http://i.imgur.com/jecwdjJ.png

Adding filters
  • HPF @ 2nd order 10 Hz
  • Linkwitz Transform: fp = 20 Hz, Qp = 0.707


SPL with max voltage to the system
http://i.imgur.com/fQqFkDh.png

http://mjanja.co.ke/wordpress/wp-con...at_success.jpghttp://i.imgur.com/JiGYpzj.png

Amplifier apparent load VA
http://i.imgur.com/sD9dBvT.png

So what do you guys think of this, is it a go?


jpmst3 06-25-2014 02:57 PM

I also have two sealed LMSs and a LOWARHORN in my HT.
If I had my way I would rather have another LOWARHORN or two. The efficiency is off the charts in comparison. Yes, you sacrifice space, but you need much less amp/boost to achieve the same effect.

But, then again, I have had just about anything and everything in there. I have finally reached the point where any changes would be simply be that, changes and most likely not improvements, bass-wise.

I also love my clone FP14K. It is one bad ass piece of gear!

http://www.avsforum.com/forum/155-di...orn-build.html

GizzeGutten 06-25-2014 03:58 PM

It would be fun to build something like the LOWARDHORN to hear it, but as this is my first proper build besides a test enclosure for the first LMS-U driver (waiting for the next one to arrive within a week or two), and it would not be enough time to get it done until the second driver arrives here as I need more experience with power tools & estimations etc (don't even know how to properly design a ported woofer), and when the King arrives, I'd like for everything to be ready and perfect! :D

My perspective: I've not tried a lot, but as so many of you savvy bass people are giving the Ultra such high praise, I thought, OK, should I peddle through a lot of subs, buying and selling them at a loss, or should I just go all in now with what I really want and pay the price up front and live happily ever after? Luckily, I paid the price and I'm very happy about that. It gives peace of mind knowing you've got something people who really know their bass hold in such high regard. If anything, I might get obsessive over getting those low notes at an ever higher SPL just because I've got an obsessive personality, but I hope to curb such inclinations before they take over completely, haha. I would probably not have gotten this second one if it was not for the fact that it was cheap relative to the first one, where I also needed miniDSP 4x10 HD and the FP14k. Seems like a waste to let all that capacity go to waste :D

As for the FP14k: Amazing! It not only works well, it looks cool too. My version does not have any custom silent fan order or mod, so I'm going to have it in the next room, but since it looks good, it's probably up for a fan mod later down the line.

GizzeGutten 06-26-2014 06:32 AM

Here's a couple of drawings of the enclosure. There will also be bracing the other side, but I don't care to draw it as this gives a good overview of the build anyways.

Do you guys think this is good enough?

http://i.imgur.com/GzOUY9s.png

http://i.imgur.com/9qHnKnH.png

http://i.imgur.com/bLaOL9B.png

GizzeGutten 06-27-2014 07:58 AM

Changed it around a bit

http://i.imgur.com/YeZg2wi.png

http://i.imgur.com/ymphexK.png


All times are GMT -7. The time now is 08:08 PM.

Powered by vBulletin® Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.