@all.
From the comments that are helpful for the topic here, I have taken the following points for myself:
_A tapped horn has no advantages over FLH, only disadvantages.
_No DIY taped kick horn has ever been built or published.
OK, then I'll probably start a self-experiment soon. I can't do anything more than fail, and you always need a bit of wood for a campfire....
But it is not easy to "force" the driver into this tuning. THs prefer to reproduce deep bass. You also have to approach a kick TH project a little differently. Lots of folds and sharp edges are not good for the fundamental midrange, which you get pretty close to with this tuning. That's why it gets a bit tricky.
The greatest strength of a tapped horn is the very dry and direct playing style. That is certainly beneficial for the kick 😉
This is my greatest hope!
away Post #23
I've considered this folding variant.
I am aware of this thread with your participation!
click me gently
It has 4 inner corners and is a bit borderline if you want to get to 200-230Hz.
But at least I can use all the skills that I've set out to do and that I've already used more or less in my other TH projects.
I won't do the lower slope as in the picture, but make it straight.
I'll also make the upper slope straight and have to live with a small gap.
I very often make concessions in my projects in favor of easy reproduction by deviating from the perfect horn contour. So far, this has never harmed the overall performance of my self-developed tapped horns.
We'll have to wait and see whether my plan works out as I hope. I really enjoy researching and experimenting and I'm willing to spend a little money on a few prototypes.
@
But as I've said elsewhere, I'm interested in a pure tapped kick horn! And not in a TH subwoofer that can play as high as possible.
Kind regards
From the comments that are helpful for the topic here, I have taken the following points for myself:
_A tapped horn has no advantages over FLH, only disadvantages.
_No DIY taped kick horn has ever been built or published.
OK, then I'll probably start a self-experiment soon. I can't do anything more than fail, and you always need a bit of wood for a campfire....
You have said what I was already afraid of. That I cannot reproduce the short wavelengths and the bandwidth of the fundamental tone in reality to the extent that Hornresp predicts.
But it is not easy to "force" the driver into this tuning. THs prefer to reproduce deep bass. You also have to approach a kick TH project a little differently. Lots of folds and sharp edges are not good for the fundamental midrange, which you get pretty close to with this tuning. That's why it gets a bit tricky.
The greatest strength of a tapped horn is the very dry and direct playing style. That is certainly beneficial for the kick 😉
This is my greatest hope!
I've already played this idea through in my head and rejected it. The housing will be very long and not very practical in PA practice. In addition, with this variant I have little scope for tuning games, which I always like to do. We recently had a discussion with @Brian Steele in this thread here:
away Post #23
I've considered this folding variant.
I am aware of this thread with your participation!
click me gently
It has 4 inner corners and is a bit borderline if you want to get to 200-230Hz.
But at least I can use all the skills that I've set out to do and that I've already used more or less in my other TH projects.
I won't do the lower slope as in the picture, but make it straight.
I'll also make the upper slope straight and have to live with a small gap.
I very often make concessions in my projects in favor of easy reproduction by deviating from the perfect horn contour. So far, this has never harmed the overall performance of my self-developed tapped horns.
We'll have to wait and see whether my plan works out as I hope. I really enjoy researching and experimenting and I'm willing to spend a little money on a few prototypes.
@
Brian Steele
Thanks to you too for the simulation.But as I've said elsewhere, I'm interested in a pure tapped kick horn! And not in a TH subwoofer that can play as high as possible.
Kind regards
Last edited:
I think the Hornresp predictions will be pretty close to the real cabinet if the horn contours are reasonably maintained.
Tricky for sure, to get better high end (less cancellation) the amount of "right hand" turns should be equal to the "left hand" turns to keep the inner and outer path lengths roughly equal.
Classic format of PA re-entrant horns or "W" bins.
Hope won't change the fact that the "dry and direct" low frequency sound of a TH turns into a mess of delay decay times in the upper bandwidth.
It's interesting to compare a TH and FLH of the same cabinet volume and similar bandwidth using the same driver:
The 15LB075UW4 does not have much excursion potential, but as you know, it's a decent inexpensive driver.
With a HP around 60Hz, neither the FLH or TH would exceed it's Xmax of ~4mm with 41volts or 30volts respectively while approaching 130dB in the upper "kick" range.
In a kick bin, one could argue the implications of the almost opposite excursion/impedance power distribution between the FLH and TH may favor one or the other, though overall the TH in the above example requires less voltage to reach excursion limits at the bottom end of the pass band.
Up to 100Hz, the group delay of both are similar, but in the range from 100Hz and up, the difference is obvious- delay times of over three cycles at some frequencies in your desired range up to 230Hz.
Anyway, the TH would be the opposite of "dry" up top, and as an old sound engineer with a preference for "pop" over "slop", easy for me to say which I'd prefer for "kick".
Art
Last edited:
A TH is basically a series-tuned 6th order bandpass alignment with bad out of band resonances. As with other 6th order bandpass designs, usable bandwidth is limited to between one and two octaves. Theoretically you could design one that goes from 80 to 200 Hz, or maybe 100 Hz to 300 Hz, but then I'd bet that when it's time to move from sim to build, you're going to run into some interesting challenges, like how to fit the driver into the mouth of the TH.
So you are of the opinion that the multiple, sometimes high group delay could be the actual “Achilles heel” of the TH kick?
What group delay would be acceptable or bearable?
And sorry if I'm digging deeper. Why can a TH sound dry at low frequencies and a higher tuned TH not? Both designs have the problem with the higher group delay, which repeats itself several times, right?
Last edited:
With a horn mouth size of about 2.5-3x Sd, I will have no difficulty accommodating the driver. At 8", the space problems begin, which require a maintenance hatch, but not at 12". There is enough space here to mount the driver.
I'll be surprised if there are any other difficulties!
I see where you're coming from with the description of a tapped horn (TH) as a "series-tuned 6th order bandpass with bad out-of-band resonances," but I think that view misses a key part of what makes a TH sound unique. It's not just about group delay peaks or response dips.
The outer tap in a TH introduces positive acoustic feedback into the system, and I believe this feedback shapes the sound character far more profoundly than what a simple Hornresp sim might suggest. While it's easy to think of the feedback as just smoothing out the response curve or filling in dips, the reality is deeper.
Positive feedback is a core principle in many life processes—it's what gives dynamic systems a sense of immediacy and vitality. In a TH, this feedback can create a kind of effortless, "alive" quality in the sound that transcends what you'd expect from a basic frequency or phase plot. It adds an organic, engaging character that isn't just about technical specs.
Hornresp and similar tools are great for mapping response and predicting issues like resonances, but they can’t fully capture the dynamic interactions created by positive feedback in the horn path. It’s one of those subtle effects that you hear more than you see on paper.
This is what my experience with tapped horns and similar designs has tought me over the years. This effect increases with more use of positive (acoustic) feedback, as seen in the ROAR designs. I believe you could design a very potent MB-TH, and you seem to be on the right track keeping bends to a minimum etc.
I have no idea what all of that means. My experience with simming and building is that if the sim shows it's bad, the build proves it. However, from my control systems engineering courses I did many years ago, "positive feedback" was a good way to introduce instability into a system and was therefore something generally to be avoided 🙂.
I asked ChatGPT to explain it better then I can (due to poor english and ME/CFS-brainfog):
Let’s dive deeper into how positive feedback (PFB) behaves in complex systems, especially in amplifiers, and how it influences transient and micro-dynamic reproduction.
1. Understanding Positive Feedback's Non-linearity
At its core, positive feedback feeds a portion of the output signal in phase with the input, reinforcing it. This reinforcement is non-linear by nature because as the input increases, the output is boosted not linearly but in a more exponential or compounding manner.
- In a linear system: A change in input results in a proportional change in output.
- In a system with positive feedback: A small change in input can cause a disproportionately larger change in output due to the self-reinforcing loop.
This non-linearity is typically seen as undesirable for accurate reproduction, but when applied subtly, it can enhance certain aspects of the signal — notably transients.
2. How Positive Feedback Enhances Transients
What is a transient?
A transient is a short, high-energy burst in an audio signal, like the snap of a snare drum or the pluck of a guitar string. These events often contain a sharp leading edge (the attack) followed by a rapid decay.
The Role of Positive Feedback:
- Emphasis on Leading Edges:
- Positive feedback tends to "latch onto" sharp changes in the signal, like the attack of a transient.
- Due to its self-reinforcing nature, the initial rise of the transient gets a slight boost, making it pop more in the mix.
- This can make percussive sounds feel snappier and more energetic, even if the underlying signal hasn’t changed significantly.
- Exponential Reinforcement:
- When a transient hits, the initial increase in voltage at the amplifier’s input starts to be reinforced by the positive feedback loop.
- This reinforcement is weakly exponential — it starts subtly but increases rapidly — giving the transient more impact without causing runaway oscillation.
- The system “reacts” faster and more energetically to these sharp inputs compared to a purely negative-feedback system.
- Counteracting Compression:
- Mechanical and environmental elements (like speaker suspensions or air resistance) naturally compress transients, smoothing out their leading edges.
- A controlled amount of positive feedback can counteract this compression by re-emphasizing those initial spikes, restoring some of the natural dynamics that were lost.
Visualizing the Effect:
Imagine a transient waveform — a sharp spike followed by decay:
- Without PFB: The leading edge might be slightly rounded due to system inertia and damping effects (negative feedback + mechanical factors).
- With subtle PFB: That leading edge is sharpened, perhaps even slightly exaggerated, bringing more clarity and punch to the sound.
3. Micro-dynamics and Positive Feedback
Micro-dynamics refers to the small-scale variations in volume and intensity that bring life and nuance to music — like the subtle plucking differences in a fingerpicked guitar.
- (Negative feedback tends to flatten micro-dynamics slightly as it works to linearize the signal, treating these small variations as "errors" to correct.) I added ( ) as this is something I don't agree with.
- Positive feedback,even in tiny amounts, is sensitive to these fluctuations. Because of its reinforcing nature:
- It amplifies micro-dynamic variations, giving the impression of greater detail and texture.
- Instruments sound more lively, and subtle expressive nuances are more noticeable.
4. Balancing Act — Stability vs. Enhancement
The key is delicate application. Positive feedback is a double-edged sword:
- Too much PFB:
- Can lead to instability, oscillation, and distortion.
- Might exaggerate transients to the point of harshness or introduce ringing.
- Subtle PFB (around 1–5% of the signal):
- Can enhance transients and micro-dynamics without destabilizing the system.
- Adds a sense of "speed" and "liveliness" to the sound, making the reproduction feel more engaging.
5. Practical Implementations in Audio
- Tube Amplifiers:
- Many tube amps inherently exhibit a mild form of positive feedback due to their design, contributing to the "lively" and "organic" sound often associated with them.
- Hybrid Designs:
- Some solid-state amplifiers introduce partial positive feedback in specific stages (like voltage gain stages) to achieve a blend of transparency (from negative feedback) and musicality (from positive feedback).
- Current Feedback Amplifiers (CFA):
- CFAs rely on a form of positive feedback in their architecture, resulting in higher slew rates and better transient reproduction than traditional voltage feedback amps.
6. Conclusion
Positive feedback, when applied subtly, acts as a kind of dynamic enhancer in audio systems:
- It exaggerates leading edges of transients in a weakly exponential way, bringing out attack and energy.
- It amplifies micro-dynamics, revealing delicate nuances in the music that negative feedback might flatten.
- It can counteract compressive effects from mechanical and environmental non-linearities, restoring some of the natural dynamics lost in the reproduction chain.
This balance between linearity (NFB) and liveliness (PFB) is at the heart of many high-end audio designs, striving to combine accuracy with musicality.
Horn response prediction on top, measurement on the bottom .
Is the ringing, group delay and resonace the result of the unique sound you’re suggesting ? Or the other side of the driver causing something unique? There is absolutely a pronounced ‘punch’ there, but the measurement shows a bunch of consequences as well?
Attachments
Yes. You are spot on. Positive feedback will tend to exaggerate dynamic behaviour, and this is seen as a slightly negative damping or ringing. Everything is a tradeoff. If you want to precompensate for the compressive effects of suspensions, imperfect ridgidity of enclosure walls etc, then you get a slight increse in ringing. Both are equally wrong and only your ears can tell you what is best for you or more correct, what you prefere. Most people seem to prefer a dynamically compressed and damped design, but I tend to appreciate the vivid life-like energy present in systems with more positive feedback (Triodes, ROAR etc), and that means I have to live with a little more overshoot and ringing.
It is definitely exciting and the punch/kick is incredible. Unfortunately, some types of music (or vocals) stir up that ringing and other don’t seem to…
I wish I had a front loaded horn to compare it to to witness/hear the difference
I had to suck some of that out with a stub.
I wish I had a front loaded horn to compare it to to witness/hear the difference
I had to suck some of that out with a stub.
Attachments
Yes, if you prefer a "Dry" kick versus a "Wet" one above ~125Hz.
The answer to what delay, or "smearing" of time is "acceptable or bearable" is subjective, and very dependent on the source material.
If you like a long-tailed "808" kick vs a dry "disco" kick sound, you might prefer the added upper frequency ringing of a tapped horn compared to a front loaded horn, bass reflex or sealed, that being the progression from most to least group delay (GD).
Blauert and Laws determined that around 3.2ms GD at 500 Hz was detectable, about 1.6 cycles.
At higher frequencies, more cycles are required to be detectable.
That would imply ~1.5 cycles to be under the GD detectable level, though the difference in arrival time between upper and lower frequencies may become apparent to some listeners with "machine gun" kick speeds.
Personally, I'd think up to 1.5 cycles of low mid "slop" would be acceptable for any genre of music.
In this Josh Ricci measurement of a 21" sealed sub, the red line is 1.5 cycles of delay, the blue line is one cycle (1000ms/frequency) of delay:
The 21" sealed sub is below one cycle of group delay at 120 Hz.
A "dry sounding" TH is usually crossed below the frequency where group delay exceeds 1.5 cycles.
The time of flight delay between the TH rear and front radiation makes that impossible at higher frequencies.
Frequency response can be equalized, but the group delay will remain.
Using the simulation format from post #22, here is the single driver TH compared to a quad driver sealed 161 liter cabinet:
The difference in group delay is obvious, at 220 Hz the TH has over 17ms GD, almost 7 cycles, the sealed cabinet has less than 1ms, less than 1/4 cycle.
No question the sealed cabinet will sound more "dry" than the TH.
Comparing the TH to a dual driver bass reflex in a 161L cabinet we see the BR is still under one cycle GD at Fb, and the same as the sealed at 220Hz.
When comparing the simulated or measured response of TH to FLH, BR or sealed, the overshoot and ringing is obvious.
Mixing sound in an outdoor environment, I found the difference to be audible even when equalized to the same target response.
That said, how anyone listening responds to the "extra sauce" (ringing after the signal has stopped) is purely subjective.
Art
I've tried running a TH higher than about 1.2*Fh. It was horrible. I don't know how anyone could like it. But I guess it takes all types ...
Yes, you are pointing to the Achilles' heel of an MB-TH. It will have a rather narrow usable bandwidth.
Here below is a real world GD measurement of one of my most extreme experiments with massiv broad band positive acoustic feedback.
Red curve is raw unfiltered and without EQ and blue curve is with xover and some EQ. From this thread and design:
This is a worst-case example where I did nothing to tame the ringing. I wanted to see (and hear) what massive broadband positive feedback would sound like and how it would behave. It does have a nice midbass punch and attack despite being a sub capable of reaching 23 Hz, and yes, it does have some pronounced ringing. I can hear it, but most of the time I don’t notice it. If I cross over the design above 90 Hz or listen to a pure bass guitar or something similar, then it becomes noticeable. If I were to build this design again, I would use a heavier-duty driver, add some damping in the pipe sections, and leave space to experiment with a Helmholtz resonator to tame the first peak above the passband. Group delay is seldom as bad in real life as Hornresp predicts, and positive feedback is a very powerful tool to increase attack, transient response, and create a 'firm, dry, and dynamic' character within the passband.
Here below is a real world GD measurement of one of my most extreme experiments with massiv broad band positive acoustic feedback.
Red curve is raw unfiltered and without EQ and blue curve is with xover and some EQ. From this thread and design:
This is a worst-case example where I did nothing to tame the ringing. I wanted to see (and hear) what massive broadband positive feedback would sound like and how it would behave. It does have a nice midbass punch and attack despite being a sub capable of reaching 23 Hz, and yes, it does have some pronounced ringing. I can hear it, but most of the time I don’t notice it. If I cross over the design above 90 Hz or listen to a pure bass guitar or something similar, then it becomes noticeable. If I were to build this design again, I would use a heavier-duty driver, add some damping in the pipe sections, and leave space to experiment with a Helmholtz resonator to tame the first peak above the passband. Group delay is seldom as bad in real life as Hornresp predicts, and positive feedback is a very powerful tool to increase attack, transient response, and create a 'firm, dry, and dynamic' character within the passband.
In audio mixing terms "dry" means a direct sound with no reflections, "wet" means a sound with added ambience.
Classic rockabilly slap-back delay ranges from ~40ms to 140ms, Elvis Presley's "Heartbreak Hotel" was 82ms.
To describe a sub with a 52ms slap decay at ~125Hz as having a "dry character" made me laugh!
Art
It is almost 20 dB down at 125 Hz, and the main loudspeakers in room tends to dominate at that frequency. But, yes, it was an experiment with lots of positive feedback, and yes I can hear this if playing a bass guitar through the sub without any other frequency content.
"positive feedback is a very powerful tool to increase attack, transient response, and create a 'firm, dry, and dynamic' character within the passband."
Same junk is in my parallel (blue GD ) or series (orange group delay) pipes and can be very annoying like a wet fart in paper underwear.
Especially audible at the second resonance in the long pipe. I wish I understood this a long time ago and wasted less wood on a few awkward paraflex
Especially audible at the second resonance in the long pipe. I wish I understood this a long time ago and wasted less wood on a few awkward paraflex
Attachments
Last edited:
Basically summing up things that have been said. I tried to design a tapped horn kick, with a throat chamber though. IIRC the chamber had to be really small, to the the point that it became a true tapped horn. The bandwidth was really small, like ~one octave small. The design was relatively easily simmed but hard to translate to practical dimensions. I don't remember if I came up with a trapezoidal cabinet design to make it work (I think so) but the trapezoidal cabinet shown is a good idea. It was definitely not following the kiss-principle though.
I finally gave up on the series tuned 6th order idea and went with parallel 6th order instead, which was a safe bet as it has been tried and tested all over. I also think that the low sub bass is often enhanced when the sound is smeared out a bit but as frequency rises, less and less is allowed.
I finally gave up on the series tuned 6th order idea and went with parallel 6th order instead, which was a safe bet as it has been tried and tested all over. I also think that the low sub bass is often enhanced when the sound is smeared out a bit but as frequency rises, less and less is allowed.