Port assisted subwoofer design for PA
Printed From: Speakerplans.com
Category: General
Forum Name: Advanced Discussion
Forum Description: Advanced discussion area for higher lifeforms
URL: https://forum.speakerplans.com/forum_posts.asp?TID=108660
Printed Date: 27 March 2026 at 4:54am
Software Version: Web Wiz Forums 12.08 - https://www.webwizforums.com
Topic: Port assisted subwoofer design for PA
Posted By: Crashpc
Subject: Port assisted subwoofer design for PA
Date Posted: 02 August 2024 at 9:01pm
|
I have touched that topic in few discussions, but never got enough good arguments, so here it is. Modern subwoofers have enough motor force and displacement, that the bassreflex port on a ported bin is often almost not needed. At least not fully in band. Now I use it just for cone displacement control, impedance shaping and cooling, otherwise it works for me almost out of band. As it is out of band, it can be small, it doesn't get the efficiency hit of small port, port noise is low, power compression is low, what's not to like? Speaking about 18" subwoofer in a 95l/3.55cu.ft box, tuned at roughly 28Hz, used between 34 and 85Hz. I first did it with 21"s. Seemingly it works great for its box volume. Only the price per woofer is somewhat higher. Is here someone willing to discuss and object to get to the bottom of things? Why noone is doing it ever? Many thanks. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Replies:
Posted By: Conanski
Date Posted: 03 August 2024 at 7:30am
|
What does the system response look like in this config? Isn't a 28hz port still contributing something to output at 34hz? If it's grossly undersized then aren't you are throwing away LF output potential. An extended shelf response can be useful when the full output of the subs in the traditional 40-100hz range isn't needed but I have never found myself in that situation, the subs are always pushed for all they are worth.
|
Posted By: Crashpc
Date Posted: 03 August 2024 at 8:14am
|
The response looks like rolling off towards the low end. Yes, the port is atill somewhat contributing at 34Hz, but way less than at the tuning, so it can be smaller, and there is no issue with efficiency and chuffing, and also power compression is avoided, as the driver is only used around the impedance peak, not in the impedance valley of the port. Seemingly LF potential is theown away, but only from the angle of view of the driver. Not from angle of view of the compact box design. One cannot pull such SPL from such box size anyways. To me it looks even after Hornresp sims, that this design wins over conventionally tuned solution. LF18X451 in 95l/3.55cu.ft bin tuned at 28Hz, properly HPF'd and EQ'd gives more output at 40Hz than when it is tuned at 38Hz. Exactly fue to the port loading (efficiency) and power compression at tuning frequency. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: vertx
Date Posted: 03 August 2024 at 9:38am
|
If we're talking modern drivers in DSP Assisted Reflex enclosures the port is everything, and when properly tuned and processed will allow you to use you're high power high xmax drivers to their full potential. Once a port chokes up you get no further gain or efficiency from it, only more distortion. If you have a 3600w peak power driver with 16-19mm linear xmax you could easily choke you're ports up at half that power and xmax at which point you've wasted money on big boy drivers in enclosures that are velocity limited.
------------- www.instagram.com/translate.sound |
Posted By: Crashpc
Date Posted: 03 August 2024 at 9:55am
|
Yes, but it is generic mumbo jumbo - in the design, frequency response and SPL are always a chosen compromise. More SPL comes at a price, at given box volume. The driver in my case will be used to its full potential by cone excursion. Some power will be left on the table, but power and heat dissipation is not equal to sound, so I am inclined to discart it as irrelevant. Port in this design is not to be choked up. Top velocity is set at about 14-15m/s. As mentioned, from the point of view of the driver and maximum SPL, it is a waste. But that's the price to pay for more power dense compact design. The high excursion driver here is tasked to move most air by itself in order to exactly not choke the port. In conventional design, the port WOULD get choked. Not here. The driver works more instead, and the port just past the edge of selected usable frequency response, somewhat assisting. Yep it is expensive to do it like that, but heck it is a lot SPL in reasonable frequency response in a very compact box that can be grabbed easily. //Edit: grammar.
------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: fudge22
Date Posted: 05 August 2024 at 10:46am
|
The following graphs show the predicted maximum output of
the drive unit you mentioned in a 95 litre cabinet. The green line is a bass reflex tuned to 40Hz, the red line
is tuned to 28Hz and the blue line is a closed box. In my opinion, no-one is doing it because they consider the extra of output from 35Hz upwards is more beneficial than sacrificing the overall output to extend the response.
|
Is here someone willing to discuss and object to get
to the bottom of things? Why noone is doing it ever?
Posted By: vertx
Date Posted: 05 August 2024 at 11:09am
|
There's a few interesting bits in this https://audioxpress.com/files/attachment/2721#:~:text=DSP%20Assisted%20Reflex%20lets%20you,%2C%20extension%2C%20and%20dynamic%20range.%20" rel="nofollow - https://audioxpress.com/files/attahment/2721#:~:text=DSP%20Assisted%20Reflex%20lets%20you,%2C%20extension%2C%20and%20dynamic%20range.
------------- www.instagram.com/translate.sound |
Posted By: Crashpc
Date Posted: 05 August 2024 at 1:54pm
Yep. People look at simulated graph with "dollar eyes look" thinking they instantly get this performance in real life. Well?! Port compression is ignored, speaker power compression is ignored. Once you account for that, the situation nearly reverses. It depends. For dynamic signal, live music, It doesn't have to be far off, but for synthetic music and bass heavy music, it is the case. Once we simulate with all kind of losses and power compression in place, we see very different story. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: Xoc1
Date Posted: 05 August 2024 at 9:21pm
Forte Audio Q18 Q15 Q12 Tony Wilkes (RIP)
|
Posted By: Crashpc
Date Posted: 06 August 2024 at 5:50am
|
I have heard that name. Dangit, it's even called the same thing! Many thanks assuring me that I'm onto something here. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: fudge22
Date Posted: 06 August 2024 at 8:06pm
|
Is this the same Q12 that is discussed in this old thread? https://forum.speakerplans.com/q12-12-sub_topic28555_page1.html?KW=BMS12" rel="nofollow - https://forum.speakerplans.com/q12-12-sub_topic28555_page1.html?KW=BMS12 If so, it seems to be a 40l box tuned to 38Hz, or there
abouts. That is just standard 4th order tuning, comparable to the
green plot in the graphs I posted above. If the other cabinets are similar
designs (presumably using 15” and 18” drive units) they are not tuned similarly
to that which the OP is referring. If the OP wants some sort of validation to brilliance of his
design, try searching for sixth order or B6 alignment cabinets. Yes, they work,
but on the whole, the cabinets are no smaller, they require additional eq boost
to create a flat response, and subsequent additional power requirement at low
frequencies to achieve that boost. As I mentioned above the maximum output is
significantly reduced compared to fourth order cabinets. Check out the book “Bullock on Boxes” for more background on
the original alignments used. |
Posted By: Xoc1
Date Posted: 06 August 2024 at 9:58pm
|
The Q18 was 100 litres an used an 'assisted alignment' The Q15 was described as being similar to a B6 allignment. Not many details left as Tony passed away 10 years ago.
|
Posted By: Crashpc
Date Posted: 06 August 2024 at 10:30pm
|
fudge32: Heh, no. I do not need validation to the briliance of my design. At least not directly. I am fishing for issues, and so far it seems I am mostly "debunking them". In this case, as someone else, elsewhere briliantly described it, it even looks like circular self negating arguing, but it is all coincidence. Especially with alignments. We can benchmark it. Take B&C speakers 18IPAL or better yet, 15DS115. The alignments suggest between 7 and 30l of volume. At no point I would suggest any of that as a preferred compact design. Those alignments only sometimes end up being similar to my suggestion, and I'd say it is accidental. I disagree with additional power during EQing the response out. The undertuned design works outside the impedance minimum. It is more efficient in the 40Hz region and eats less power, therefore it can get the boost without penalty in power input. Also the cone is moving there better, so the speaker gets the needed cooling, compared to the one working at port tuning frequency. There are always "dollar eyes glowing" when seeing higher sensitivity in the bass region in the sims, but it is negated by poor efficiency and heavy compression on the coil and on the port too. Easily 5dB down. That's where small signal sims fail you big time. Hornresp does have the capability to sim all this too somewhat. I encourage you to not skip that part. Now, there is this claim that one is not getting more SPL from such design. Well, let's settle on some case we can simulate and let's see how things work out. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: snowflake
Date Posted: 06 August 2024 at 11:01pm
it is surprising that Hornresp doesn't include an option to set a maximum port speed in the Maximum SPL tool. @ David McBean |
Posted By: Crashpc
Date Posted: 07 August 2024 at 7:10am
|
I don't want to nag him too much. It is very tedious and impractical to compare driver power plots. These should be in the menu where frequency response is situated. Anyways i fail to see how putting a driver with most motor force (most Newtons per Watt pushed at the cone) and displacement volume could end up in less SPL compared to lesser device in the same volume/situation. The math aint mathin here. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: fudge22
Date Posted: 07 August 2024 at 8:13pm
|
The impedance curve for the cabinet tuned to 28Hz is not that significantly different to one tuned to 40Hz. Max impedance peaks are just shifted down by about 4Hz. Unless you play a specifically chosen sine wave the dissipation/power requirement in each design will be roughly similar.
So, the design is limited by cone displacement rather than heat dissipation.
The curves that I posted were the max spl, as in large signal, not the transfer function. They represent the maximum output irrespective of any signal manipulation.
If you are happy with your design, that is great. I simply answered the question you asked, i.e. why noone is doing it ever? |
Posted By: VECTORDJ
Date Posted: 08 August 2024 at 2:22am
| Putting a 18 woofer in too small of box tuned deep needs lot of Power boost in the lows...This will eat up Your Amp Watts very quick....This idea is very old...EV. EAW, and others have used this for many Years...Bag End uses this with a small box sealed 18..Works great at med volumes...Runs out of steam at high volumes...There is No FREE Bass...There is No Magic...Any Good Speaker design Program allows You to Overlap a range of Box sizes and Tunings to See which is best for You.. |
Posted By: Crashpc
Date Posted: 08 August 2024 at 6:20am
That´s funny interpretation of things. I see that way one could be easily dismissive. I would. But that´s not the meat of the difference. All the world happens UNDER the peak of the lower tuned box. While the 40Hz box has lowest impedance at around 7,2Ohm and average impedance between 40 and 50Hz is lingering at 8Ohm, AND there is excursion minima - heavy thermal compression AND port losses, the 28Hz tuned box has minimum nearly 10,8Ohm and averaging at 18Ohm. That is more than double the average impedance, while the driver is moving and cooling, and port is not fully loaded. I think you missed that.
Yes, nothing wrong with this. Although the larger cone displacement driver would solve that (either 21" or the new 16" and 19" RCFs), this limit is chosen instead power limit, that robs you from quite many decibels due to the thermal compression and port choking in restricted box volume. The choice between these two limits seems to be now favourable for the excursion limits. Mind you, LF18X451 in my bench test spent only 120VA on the suspension to get to 15mm of excursion. Heat losses plus port losses will be in ballpark of 800Watts. That exactky underscores the differences in losses that are happening, and explaining why my design most probably will not trail the high tuned one at all.
These curves ignore power compression losses and port losses, which might be as good as 5dB easily.
Given the explanation and general ignorance, it leads me to the answer "because noone gives a damn". That´s why. I am now at peace with it. By the way, some older folks than me remember some designs that actually work that way. For various different reasons, as enough available displacement from dual design and lack of proper DSPing at that time, and such. Found my answer and my peace, and it seems my design is not wrong. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: Crashpc
Date Posted: 08 August 2024 at 6:29am
The simulations I made indicate about 2.5dB boost at 38Hz including power compression, which is covered power wise by heightened impedance, which means lower power input in the first place. It equals itself out almost to "zero net", so no heavy power boosts. Either same, or less power goes to the driver. Even stronger drivers need even less boost, because the driver is able to drive the port in a very small volume. 18DS115_4 can do such a thing with its powerful motor, but unfortunately it will NOT be happy working between 14-15mm of excursion, spending much power on its suspension. It takes specific driver to do it all. We are very getting there though with technology and new released drivers. At no point I am asking for free bass. It is just different approach, allowing me to do different things. Such as going down to 30Hz with more SPL than the competing conventional design if I restrict the input power. The box gets more universal, and there is less stress from burning the driver, as power is little harder and more expensive to control in contrary to the excursion.
------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: fudge22
Date Posted: 08 August 2024 at 8:14pm
|
So, your question was somewhat rhetorical. Ignorance and apathy were the answer all along.
By the way, I am one of those older (ignorant and apathetic) folks. |
Posted By: Crashpc
Date Posted: 08 August 2024 at 9:45pm
|
No. The (my) answer was a hint to the fact that your answer was not correct, because it ignores any kind of compression on the output side. It needed to be not so subtle, yet I still didn't get the point across. The remark was probably taken for personal attack for ageism, while the remark explains that some old folks used such design that I use. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: airbell
Date Posted: 09 August 2024 at 11:17am
|
What you didnt mention here is that a ported sub with same driver+same tuning but smaller enclosure can also sound very different. For example all my first prototyes for the 21DS115-4 had around 200 to 220l Net at 35Hz tuning. BCspeakers recommends 130L at 40Hz. What I realized later was that even if the subbass sounded very clear and LOUD and impressively "fat", the overall timing and so the actual "punch" of the sub was not so good. Probably okay in a 2 way bass setup used up to only 60Hz, but even then there are other 18" subs are more precise in that band. EQ doesnt help here much, even not (enough) the ipal mod as I mentioned some years ago. After a lot further testing my best results were with 132L net and 38Hz tuning, its way smaller, same output up from 40Hz but much more "punchy" and better timing, still enough low end but obviously less than with 220L net. But for me it is the best overall performance, not to forget its way more compact and easy to handle. Second thing is with a lower tuning for a given enclosure volume you can also make the sub sound more clear or precise in the upper bass because the (out of phase) output of the port is shifted away from the upper bass, so it comes more from the cone than from the port (it is still a mix!). But this works only to a certain point in my experience. I tried to tune the 21DS115 with 132L volume lower, as low as 35Hz and also 30Hz but, depending on music played, the subjective spl in the bass was less loud and also"too thin" at a certain point. You can try to compensate that with some eq but even the 21DS115 will run out of xmax and juice at some point then. Not to forget that for the lower tuning at the same net volume and same port diameter you need a longer port that could have also more turbulences. You can simulate that very well with hornresp, same vrc but longer lpt (lower tuning) max. power at 1W and you get the efficiency. With the lower tuning it is (in my example with a 18NTLW5000) less efficient above 50Hz but more under 50Hz. But if you simulate max. spl with realistic values, especially limited excursion, you will notice 3-4db less spl at 40-60Hz and you can actually hear this very clearly. Even with more port compression, and more average power, for most music with the same volume and even a long excursion driver the higher tuning will give you more max spl. Thats why a lot of commercial low to medium budget subs, for example also bose F1 sub, are tuned quite high and can be actually quite loud, but you miss all the lower notes and it tends to sound more one note than very detailed. The point is, too low tuning for PA subs has also its price. A lot of bass heavy music has most bass content at 40Hz, so even a 18NTLW5000 or 18IPAL will run earlier out of juice here if you tune it at 31Hz instead of 35Hz, and will (subjectively) sound around 2db less loud here, but then from 55Hz upwards the same loud again. I tested exactly that last year. But in generell I would prefer much more a under tuned compact sub with a good bandwith and extension than a huge over-volumed high tuned one note "only made for spl" subwoofer. Except maybe for carnival
 |
Posted By: Crashpc
Date Posted: 09 August 2024 at 2:06pm
|
You have a nice experience, similar to mine. I would agree. BUT, with the content I played, the coils start burning. Therefore the 3dB advantage claim is moot to me. There is no 3dB advantage when the coil is burning down. So maybe it is an individual thingie, that works differently for different people for these reasons. Recently saw some EAW subs conparisons and limiter behavior. When the persistant tone hits, the SPL goes as much as 10dB down. So I can't agree with claimed superiority in SPL. It doesn't work either. That design and behavior might be possible to ignore in some/many cases, not so in others. To the credit of the community, I was not told I am stupud and doing it wrong. I was presented with "why others do it differently". It is difficult to get on it in the sense of "also my case is diferent, that's why we are not agreeing". ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: snowflake
Date Posted: 09 August 2024 at 3:41pm
I designed a hybrid double 21" with the 21DS115-4 and settled on 261L rear volume and 39Hz tuning - almost exactly the same! |
Posted By: Crashpc
Date Posted: 09 August 2024 at 6:51pm
|
By the way, my compact ported 21" design with 21DS115 that I am leaving have about 130l too. This is not a coincidence I guess? ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: bob4
Date Posted: 21 August 2024 at 11:27am
The other day I saw a comment on fb, in a discussion about new celestion woofers. .....  |
Posted By: Crashpc
Date Posted: 21 August 2024 at 6:54pm
|
Yes? I am finding my ways. The experience and knowledge here is paid by wallet though. Today I had a call with a friend, to which I recommended B&C 18DS115. He put em into X1s, and had some negative remarks on that. I had to explain that these might need some EQing to sound more familiar and warm compared to the speakers with less strong motor. He told me he did an after-party on the gig, and blew almost everything he had, except 18DS115s. It looks to me that newer 18DS115s with longer voice coils compared to older 21"s for whatever reasons might be tougher. I was pleasantly surprised. Hard electronic music, low dynamics, going hard in X1s with steady bass lines, the 18DS115 handled abuse with grace. My 21DS115s in the ported box didn't. The conclusion, although not very strong is that it takes only slight change on the same model of the speaker for the outcome to be different. Now I wonder if I shouldn't go with 18DS115s then. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: Crashpc
Date Posted: 22 August 2024 at 6:01pm
|
So I redid the case I have. Sims in HR. Compact ported box that goes straight to 40Hz, -3dB@ 35Hz, and then random roloff. RCF LF18X451, 96l, once tuned at 39Hz, once tuned at 30Hz. In both cases the port takes the same additional box volume.  Top row is the 39Hz box, bottom one is 30Hz box Thermal compression implemented as best as possible to keep SPL same. It might be usage dependent, but as can be seen, the 30Hz version has more cone movement and cooling in the majority of the range, and gets less power in the bass and deep bass region for less thermal compression. Also port flow is lower, and I guess HR does not take into account port losses. Maybe this can be neglected as longer port might have same high losses with lesser flow. Still this does not look like the undertuned solution loses at all. Yes you can up the power input on the 40Hz bersion, but wait the moment when the DJ pulls up that steady 40Hz bass line. Then either the limiter kicks it down 6-10dB or the thing is in fire in no time. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |
Posted By: DMorison
Date Posted: 22 August 2024 at 7:43pm
|
Hi Crashpc, The forum shrinks images down to 1000px wide when we upload directly, this makes it really hard to see the axes on your graphs. Is there any chance you could split it into 2 images, or host offsite and link from your post so we can see full resolution please? Thanks, David.
|
Posted By: Sypa
Date Posted: 23 August 2024 at 12:16am
|
And what do you mean newer 18ds115 with longer coils? And compared to which older 21 ? |
Posted By: Crashpc
Date Posted: 23 August 2024 at 4:03pm
|
@Dmorison: here: https://ibb.co/xmmhf3h" rel="nofollow - https://ibb.co/xmmhf3h Sypa: Not enough experience, mixed opinions. The stuff suspension stands, the 18DS115 will burn quite some power on its suspension at and past Xvar, so it is not wise to run synthetic music with low dynamics to Xvar. On the other hand, compared to 21DS115, indeed, the 18DS115 has longer coil, so it will have good motor force still at these excursions, suffering somewhat less thermal compression. My colleague did an afterparty recently, with the rig made of various drivers, even in the bass portfolio, and only what survived was 18DS115. So yeah, sturdy awesome driver! I am looking for next gen drivers from RCF with Xmax of 18mm. Their suspension behavior is benevolent, so I would at least bet that some models wpuld regularly work with 20mm no problem. They are 16" and 19" though. There is always something on the horizon, and it is rarely worth the wait. ------------- Nikon and Canon people should not be married to each other. Why did you let this happen? |