AE lists the TD18H+ as a bass woofer and that seems like a good description, it is a driver than can cover a lot of ground with it's low inductance and and high sensitivity. It is totally different to an Ultimax which has been designed as a true subwoofer it does not pretend that it can do anything else, the Fs is low the cone is heavy the sensitivity is low but it has more xmax and can take a lot of power which will be needed.
Both are good drivers but they are very different. The best or right one depends on what it is being asked to do. The Ultimax would be happier doing Dinosaur footsteps in movies at reference level but in a hifi system the TD18 seems a better choice.
Norman you read my mind
I was going to finish these boxes in charred wood and my buddy talked me out of it...I ordered my "stuff" I think it will look pretty bad ***...its still going to be black...Hopefully I can get this finished before my work starts back up...Spring break is the week after next and the kids are set on going somewhere (anywhere).
I received some schooling by Mabat, Ro808 and Tmuikku on Voicing and group delay.
From the material I studied about horns (JMLC) I picked up some strict standards for group delay....not to downplay that aspect but from what I've gathered now...its not the end of the world; this group delay bump in the response around tuning....So as I play with horns with intentions of running close to tuning or even lower than tuning....taking advantage of large diaphragm compression drivers in the home setting.... to achieve low crossovers with a smaller horn allowing wider HF dispersion....It seems that the resulting frequency response is first and foremost importance....I also seem to have under estimated the fineness of EQ...which really seems stranger to me but...**** happens right? i'm used to bending curves with my favorite eq in my DAW but apparently having experience looking at the type of FR commonly seen with a Horn and knowing what is possible makes the difference. I do not feel like my challenge in the unique voicing of a horn below cutoff was over estimated....its been a while but I recall not getting the real world measurements to play out as easily as the FR on the display of HR...though this may be my own fault if I chose say a high pass shelf when I meant to chose a low pass or etc etc, I'm definitely am going to have fun figuring it all out.
EQ's are not created equal...there are variables not accessible in the filter wizard of HR regarding certain filter types...the low or high pass shelf of my crown USP4 cards have things like linear phase brickwall hp/lp which will allow me to run a certain latency deeper than the 24db and 48db slopes and is not a static slope....getting shallower as you slide to its lowest ability for chosen latency....the hp/lp shelves have a 6db/oct version and one with variable slope/Q...but when the variable is set to 6...it doesn't match the 6db/oct the other shelf filter....so I say that not asking for advice or complaining...I say that to say....make sure you look into your eqing ability in the real world vs what you are modelling with...you may be surprised that you'll have to take a unique route vs just transferring over the metrics of the filters you've modelling in simulation.
I was going to finish these boxes in charred wood and my buddy talked me out of it...I ordered my "stuff" I think it will look pretty bad ***...its still going to be black...Hopefully I can get this finished before my work starts back up...Spring break is the week after next and the kids are set on going somewhere (anywhere).
I received some schooling by Mabat, Ro808 and Tmuikku on Voicing and group delay.
From the material I studied about horns (JMLC) I picked up some strict standards for group delay....not to downplay that aspect but from what I've gathered now...its not the end of the world; this group delay bump in the response around tuning....So as I play with horns with intentions of running close to tuning or even lower than tuning....taking advantage of large diaphragm compression drivers in the home setting.... to achieve low crossovers with a smaller horn allowing wider HF dispersion....It seems that the resulting frequency response is first and foremost importance....I also seem to have under estimated the fineness of EQ...which really seems stranger to me but...**** happens right? i'm used to bending curves with my favorite eq in my DAW but apparently having experience looking at the type of FR commonly seen with a Horn and knowing what is possible makes the difference. I do not feel like my challenge in the unique voicing of a horn below cutoff was over estimated....its been a while but I recall not getting the real world measurements to play out as easily as the FR on the display of HR...though this may be my own fault if I chose say a high pass shelf when I meant to chose a low pass or etc etc, I'm definitely am going to have fun figuring it all out.
EQ's are not created equal...there are variables not accessible in the filter wizard of HR regarding certain filter types...the low or high pass shelf of my crown USP4 cards have things like linear phase brickwall hp/lp which will allow me to run a certain latency deeper than the 24db and 48db slopes and is not a static slope....getting shallower as you slide to its lowest ability for chosen latency....the hp/lp shelves have a 6db/oct version and one with variable slope/Q...but when the variable is set to 6...it doesn't match the 6db/oct the other shelf filter....so I say that not asking for advice or complaining...I say that to say....make sure you look into your eqing ability in the real world vs what you are modelling with...you may be surprised that you'll have to take a unique route vs just transferring over the metrics of the filters you've modelling in simulation.
You may be right...but I'd rather here it explained in a more technical or scientific way...as I said before...if all things are within normal operating conditions....I can't see why would would be "happier"...and what does "happier" sound like? If both are running within optimal stresses....are you saying they are not going to sound identical or very close even though they have been FR and SPL matched?
Hi Camplo, i am not an expert on T/S parameters or Hornresp by any means, but here's a few things i look at when comparing sub drivers....that might help sort out the 18+ vs Ultimax comparison.
Starting with sealed.
In Hornresp, i put the drivers in a box volumes that gives each a 0.7 qtc, and look at their power curves, small signal.
For instance the 18+ centers and peaks around 250Hz, whereas the Ultimax centers and peaks at around 80Hz. (see plots below)
It's a sub, so i just look at the low end extension i'm interested in...let's say 20Hz.
I see the 18+ is down 25dB from its 250 Hz peak/center. And the Ultimax is down 6dB from its 90Hz peak/center.
So now I have a sense of what frequency range they were optimized for. Time to bump up input voltage until diaphragm displacement hits xmax, to see what SPL they can really make.
Came to 1300w at 8ohm, 102v for the 18+.
And 850w at 4ohm, 58v for Ultimax.
With these net usable response curves, first for the 18+, then for the Ultimax.
My guess is you're probably already doing all that.
But here's the real comparison imo....
Look at the lower bound you want, let's just stick with 20Hz.
It's about 105Hz for the 18+, and 109Hz for the Ultimax.
Ime/imo, Any further extra response, sensitivity, efficiency above those levels ...whatever you want to call it...
is immaterial in comparing these drivers if flat frequency response down to the corner is the goal.
It's just excess, unused capability.
So the 18+ 130dB peak capability above 105dB @20Hz adds nothing,
just at the Ultimax 115dB peak above 109dB @ 20Hz adds nothing.
IF flat response is the goal...which for me always is.
Let's say 30Hz works for low extension.
Now it's about 112dB for both.
Once above a 40Hz corner, the 18+ starts to rule.
From 100-700Hz she rocks in at about 127dB net usable!
Oh, and another biggie.. EQ at the bottom cannot be used to improve low end beyond those curves...all that happens is xmax is exceeded.
This is the Achilles heel of sealed subs using EQ like a Linkwitz transform to get low end extension.
Makes for perilous excursion, and also requires a doubling of amp power per each 3dB boost.
Come to think of it, in home audio without amplifier output meters or at a minimum clip indicators, how do you know:
a. when the amp is clipping, when you're short power
or b. when you hear distress, whether it's the amp clipping or over excursion
or c. maybe both Lol....
How can you sort out the distress without knowing amp output???
Anyway, hope all that helps. And i hope that you can see where some folks were coming from with their 18+ evaluations..
This way of comparison of course ignores box size differences, amps required, and other factors, but i do think it's a pragmatic beginning in comparing how to obtain real world SPL...flat that is
Starting with sealed.
In Hornresp, i put the drivers in a box volumes that gives each a 0.7 qtc, and look at their power curves, small signal.
For instance the 18+ centers and peaks around 250Hz, whereas the Ultimax centers and peaks at around 80Hz. (see plots below)
It's a sub, so i just look at the low end extension i'm interested in...let's say 20Hz.
I see the 18+ is down 25dB from its 250 Hz peak/center. And the Ultimax is down 6dB from its 90Hz peak/center.
So now I have a sense of what frequency range they were optimized for. Time to bump up input voltage until diaphragm displacement hits xmax, to see what SPL they can really make.
Came to 1300w at 8ohm, 102v for the 18+.
And 850w at 4ohm, 58v for Ultimax.
With these net usable response curves, first for the 18+, then for the Ultimax.
My guess is you're probably already doing all that.
But here's the real comparison imo....
Look at the lower bound you want, let's just stick with 20Hz.
It's about 105Hz for the 18+, and 109Hz for the Ultimax.
Ime/imo, Any further extra response, sensitivity, efficiency above those levels ...whatever you want to call it...
is immaterial in comparing these drivers if flat frequency response down to the corner is the goal.
It's just excess, unused capability.
So the 18+ 130dB peak capability above 105dB @20Hz adds nothing,
just at the Ultimax 115dB peak above 109dB @ 20Hz adds nothing.
IF flat response is the goal...which for me always is.
Let's say 30Hz works for low extension.
Now it's about 112dB for both.
Once above a 40Hz corner, the 18+ starts to rule.
From 100-700Hz she rocks in at about 127dB net usable!
Oh, and another biggie.. EQ at the bottom cannot be used to improve low end beyond those curves...all that happens is xmax is exceeded.
This is the Achilles heel of sealed subs using EQ like a Linkwitz transform to get low end extension.
Makes for perilous excursion, and also requires a doubling of amp power per each 3dB boost.
Come to think of it, in home audio without amplifier output meters or at a minimum clip indicators, how do you know:
a. when the amp is clipping, when you're short power
or b. when you hear distress, whether it's the amp clipping or over excursion
or c. maybe both Lol....
How can you sort out the distress without knowing amp output???
Anyway, hope all that helps. And i hope that you can see where some folks were coming from with their 18+ evaluations..
This way of comparison of course ignores box size differences, amps required, and other factors, but i do think it's a pragmatic beginning in comparing how to obtain real world SPL...flat that is
Humble Homemade HiFi discusses the merits of different 15" drivers in his design here.
http://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Calpamos.pdf
http://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Calpamos.pdf
Mark100 - that's all well and good if your room adds no LF gain. But almost any practical room will. Most USA-type construction rooms ad gain at about the same rate as bass is reduced in sealed systems, so not so much EQ is needed to flatten out the whole system (speaker AND room). In the end, most of the time all that matters is the total volume displacement of the driver (Sd * Xmax) all else can be corrected.
Mark100
I think making the box the same volume is more important to make a comparison....I can't remember what size box but these are idnetical sealed cabs sim'd for both drivers...
Nothing said so far....(that I can interpret) suggest that one of these drivers are going to sound significantly different at 20hz...If so which part...The 18h+ is about the same or more sensitive in every part of the FR....Just because the sensitivity continues to rise into the midrange, why is that a negative when it is as sensitive as the Umax everywhere else?
Ok I see your point on the showing those graphs now...combined with
I enjoyed this
We are talking about different ways to judge max potential but thats not it...
I am wondering why the 2 subs, level matched at 20hz....playing within xmax....would ever sound different...
I think making the box the same volume is more important to make a comparison....I can't remember what size box but these are idnetical sealed cabs sim'd for both drivers...
Nothing said so far....(that I can interpret) suggest that one of these drivers are going to sound significantly different at 20hz...If so which part...The 18h+ is about the same or more sensitive in every part of the FR....Just because the sensitivity continues to rise into the midrange, why is that a negative when it is as sensitive as the Umax everywhere else?
Ok I see your point on the showing those graphs now...combined with
That is exactly how I ended purchasing another set of 18"s to make pairs.
I enjoyed this
We are talking about different ways to judge max potential but thats not it...
Yeah I get it...especially if you look at transfer function (which horn reso doesn't show)...it probably confuses people...
I am wondering why the 2 subs, level matched at 20hz....playing within xmax....would ever sound different...
Last edited:
The room gain widely quoted is 12dB per octave starting from frequency where 1/2WL fits in longest room dimension, i think. Correct me if I'm wrong.
I do know whatever exactly it is, it's a theoretical best based on completely tight construction...so unless you're in a basement bunker (or auto maybe)...knock it down by at least half.
Yes, in the end it is all about volume displacement....that can actually be achieved with the driver in a cabinet across the entire intended freq range.
Cool. First off, i share your wonder why 2 subs playing within xmax at 20Hz would sound different.
I keep learning there...i think it's about large signal parameters being a different set of realities than what appears from small signal T/S simulations (pushed up linearly to high drive levels).
B&C's paper makes sense to me here..https://bennettprescott.com/downloads/TechTalk_Design_202002.pdf
Re box volume...i think you have to use different volumes, unless your goal is simply to build to a certain size. The 18+ example i posted used 49L, the Ultimax had a whopping 310L !
The box type, and size, totally rule imo.
For instance, I find it interesting how vented subs run into xmax limits above the low corner....easily demonstrating SPL at a given frequency is not completely about displacement.
If you want, i'll post an example of that....a vented build with a faital 18FH500...along with a crazy frequency dependent xmax limiter i dreamed up.
thats exactly why you are supposed to compare drivers in equal volume boxes! lol
I agree...in this case.. we have two different (ultimax18 vs 18h+drivers moving about the same air (18"s)....one is slightly more efficient and one has a higher xmax...one cone is lighter than the other....if both, using the same box...and being ran well under xmax....level matched at 20hz...You think you are going to tell them apart in a blind test? Which one is righter?
Last edited:
In the strictest criteria and noting that T/S theory peters out at [Fhm], BW is limited to:
upper: Fhm = 2*Fs/Qts'
lower: Flc = Fs*Qts'/2 [normally never used]
Qts': 2*Fs/Fhm
Qts': Qts + any added series resistance [Rs]: Calculate new Qts with Series Resistor
[Rs] = 0.5 ohm minimum for wiring, so may be higher if a super small gauge is used as a series resistor plus any added resistance from an XO/whatever.
Mass, inductance, diaphragm TL and breakup modes governs the rest.
So relatively speaking, the Ultimax is a true subwoofer [~16-63 Hz], Vs the 18+ being a true woofer [~30 -250 Hz]: Interactive Frequency Chart from Independent Recording Network
You know, perspective and goals are everything...
Mine, to help explain where i come from, is I could care less about comparing drivers on their own.
I have some particular design in mind and my total focus is on what driver/box combination best meets the design.
Imo, you simply can't compare the 18+ and the Ultimax in the same sized box. The two drivers are different animals needing different sized cages Lol
Well it was Martin King who taught that the way to compare drivers is with identical cabs???
Gm I see the numbers but I struggle to connect the dots...looking here
How do these formulas you shared relate to distortion??
Upper fhm 272
Flc is 3.3
Why do I get 3.3 and you get 30hz
The most confusing thing is how these play out in the real world...if playing 20hz...the quantity of excursion vs mechanical limitation surely holds priority....right?
Gm I see the numbers but I struggle to connect the dots...looking here
How do these formulas you shared relate to distortion??
Upper fhm 272
Flc is 3.3
Why do I get 3.3 and you get 30hz
The most confusing thing is how these play out in the real world...if playing 20hz...the quantity of excursion vs mechanical limitation surely holds priority....right?
I say 'peters out' to mean T/S theory for box alignments blends in with the driver's mid-band/can't design any other box alignment BW/whatever above this point [or below Fs if one wants to design a box alignment such as Bag End's? Infra-bass sub].