Wolfie62 wrote: "Looks there is a hardwired inductor blocking output at 50 Hz. Is there?"
I speculate that the inductor is the woofer’s voice coil. In general the longer the voice coil (the more turns of wire in it), the higher the inductance. There are construction techniques which minimize voice coil inductance, but they are expensive, and may not be practical at some price points.
Duke
|
"Attempting to set crossover points, and delays for both subs at the same time is a lot of work. Far easier to do after you’ve gauged the overall response of a single sub."
Hi Erik,
My experience, and the experiences of many customers who are using four subwoofers intelligently distributed, has apparently been different from yours.
To the best of my knowledge most if not all of my customers have previous and often extensive experience with subwoofers (no newbies as far as I can recall), and instead of integration being far more difficult than when they were using a single sub, many report it to be far easier and faster. Not a single one has reported setup to be more difficult than when they were using fewer subs. And is seems reasonable to me that two subs (intelligently distributed) would fall in between one and four in difficulty of integration.
I don’t dispute your experience, but I do disagree with your generalized conclusion that "if you can’t integrate 1 sub, you can’t do 2", though a different setup paradigm may be called for.
Duke
|
In my opinion a good starting point would include seeing what the sub itself is doing without the room’s effects. Since this is a sealed-box sub, this will be fairly straightforward.
Measure the sub’s output with the microphone about 1/2" from the center of the cone. This is called "close-micing". The output from the cone will much louder than the room’s reflections at the microphone location, and will essentially overwhelm the room’s effects.
The difference in shape between that curve and your normal in-room curve is what the room is doing to the sub’s response.
Speaking of differences, the differences between your 50 Hz, 90 Hz, and 130 Hz curves look to me like they are "in the ballpark" for a 12 dB per octave crossover. The 50-Hz feature you are identifying as the "crossover point" is probably either a room interaction effect, a frequency response anomaly native to the subwoofer, or both. Having the close-miced curves in addition to your in-room curves will give you valuable insight into what is native to your subwoofer and what is caused by room interaction.
Pinwa, you mentioned that you are using two subs. Two subs can work together to give smoother in-room response than either one alone. How much freedom do you have to re-position your subs?
I disagree with Erik’s statement that "if you can’t integrate 1 sub, you can’t do2. " As the number of intelligently-distributed subwoofers goes up, not only does the in-room response become smoother, but the specific location of any one sub becomes less critical.
Duke
subwoofer designer and manufacturer
|
Erik asked: "Then let me ask this question another way. Looking at the OP’s original data, have you ever seen 2 subs start out that poorly in a room?"
I’ve never measured two subs in a room.
There may be significant room for improvement, but do not know how much of what we see in his curves is room interaction and how much is the subwoofers’ native response.
Pinwa asked, "my starting question, at its simplest, is really just whether or not subwoofer crossovers are supposed to behave the way the Starke crossover is behaving."
Your measurements look to me like they are in the ballpark for a variable-frequency second-order lowpass filter.
I can elaborate if you’d like.
Duke
|
"Please do [elaborate]..."
"First order" means "6 dB per octave". "Second order" means "12 dB per octave. "Third order" means "18 dB per octave". And so forth.
I made a mistake and mis-read the vertical scale. Upon closer inspection, the slopes I’m seeing are closer to 6 dB per octave than to 12 dB per octave.
For the moment let’s focus on the SPACING between the the lines on the 50 Hz setting, the 90 Hz setting, and the 130 Hz setting. The change in the spacing between the curves will tell us what the different crossover settings are doing.
Over the octave from 50 Hz to 100 Hz, the spacing between the 50 Hz and 90 Hz curves grows by about 4-5 dB. This is about what we’d expect from two 6 dB per octave ("first order") filters that are less than an octave apart (90 Hz is less than one octave above 50 Hz).
Over the octave from 50 Hz to 100 Hz, the spacing between the 50 Hz and 130 Hz curves grow by about 7-8 dB. This is about what we’d expect from two first order filters that are a little more than one octave apart (130 Hz is more than one octave above 50 Hz).
Okay, now let’s stop overlooking the ups and downs in your measurements. The rolloff certainly LOOKS LIKE it begins at 50 Hz regardless of the filter settings.
What we don’t know is, what the response of the subwoofer is WITHOUT the room. For instance, that EXTREMELY steep rolloff between 50 and 55 Hz looks to me like a room interaction effect.
But the general trend over the two octaves from 50 Hz to 200 Hz is a significant amount of rolloff, so I SUSPECT that’s in the subwoofer’s native response. Mentally subtracting out the crossover filters, it LOOKS to me like the native response of the subwoofer is rolling off at ballpark 12 dB per octave north of 50 Hz.
ONE possibility is that the subwoofer’s voice coil has very high inductance which results in significant peaking in the 40-50 Hz region, followed by a lot of rolloff above that region, due to inductance or cone mass or both. (A very high voice coil inductance can easily happen in a subwoofer driver, resulting in a response peak like what we MAY be seeing in the 40-50 Hz region, followed by an inherent rolloff like what we MAY be seeing north of 50 Hz, but without a close-mic’d response curve we cannot reliably say what is room effect and what is not.)
So I THINK your subwoofer’s crossover is working as designed. I THINK that the response curves you are getting are a combination of the native frequency response of the subwoofer, the effects of the crossover, and room interaction.
Duke
|
Cheeg wrote: "There is some great information here, but this thread REALLY has me wondering if I want to mess with subwoofers!" Well said. This thread would be an extremely discouraging introduction to subwoofers. There are some very successful schools of thought when it comes to subwoofers and integration with the mains, and that fact is not at all obvious from this thread. REL, Rhythmik, Vandersteen, Hsu, JL Audio, and others embrace various different approaches which have worked very well for many people. Yes these subs are more expensive than the ones in this thread but imo they are worth it. Erik has a blog post about his approach, and while I’ve been known to gripe about parts of it, I’m going to post the link because imo it offers a well thought-out roadmap to success: https://speakermakersjourney.blogspot.com/2020/04/how-to-not-buy-subwoofer.html
Erik and I may disagree on some of the finer points of subwoofers, but we both think they can be well worth messing with. Duke |
@kenjit wrote:
"They [Erik and Duke] are the worst of enemies and any suggestion otherwise is disingenuous."
This is a lie.
Stop stalking us both.
Duke
|
|
Hello Pinwa,
Imo the close-miced curve of the Starke provides a lot of useful information. Now you can tell what was room interaction and what was native to the subwoofer, and it is much easier to tell what the crossover is doing. Good job!!
I think the close-miced curve on the Klipsh is giving you an incomplete and therefore misleading picture. From that curve I’d assume it has a port or passive radiator tuned to about 26 Hz.
If so, getting a representative frequency response curve becomes vastly more complicated. You ALSO need to close-mic the port or passive radiator and SPLICE that curve with the woofer’s curve, adjusting for the relative RADIATING AREAS of the woofer cone and the port or passive radiator. You also have to take into account the relative phase rotation between the two and I don’t know how to do that math - I’d have to use a computer program.
In other words, close-micing ONLY the woofer of a vented box DOES NOT give a complete and accurate representation of what the system is doing.
I suggest simply assuming that the Klipsch is competently designed, rather than doing more measurements plus a ton of math.
Also, pay attention to what geared4life is telling you. If his analysis is correct (and I think it is), your system configuration is not allowing you to adjust the level of the subwoofers with the same volume control that you are using for your main speakers.
Duke
|
Pinwa, in the spirit of plowing with the horses you’ve got, I like Erik’s idea of plugging the port(s) on the mains and then overlapping the mains and subs. If the Moabs have multiple ports, you might leave one open. This might work as a plug: https://www.zoro.com/test-tite-t-cone-combination-cleanout-test-plug-4-86400/i/G3337302/
Its maximum diameter is less than 4 1/4" but you could expand it to the maximum and then wrap the perimeter with duct tape until it fits snug. To remove the installed plug, unscrew the orange piece to relieve the expansion pressure and it should come out easily. But try Erik’s T-shirt suggestion before ordering the plugs because a rolled-up T-shirt will work just fine. The Moabs will still be low-bass sources even after port pluggage, but not as loud. So now you can turn up the subs a bit louder. And you can use different crossover settings on each of the two subs to smoothe the blend. I’m under the impression that repositioning the subs is not practical, but I glanced at the specs and apparently they have phase controls. Use them, and they can have dissimilar settings and still work well. Duke |
