| paulspencer |
How feasible is it to design a horn sub with a HT driver? In particular with my dual AV12 high excursion 12" sub drivers. I can live with an F3 around 30 Hz as with room gain and a little eq I can still get down low for HT. I had a play with hornresponse some time ago with my drivers and it seemed that the excursion would likely not be exceeded, which means I could easily use some eq boost down low where the ear isn't so sensitive to distortion.
In a 30 Hz horn how does the output at 20 Hz compare with a sealed sub?
I'm thinking of placing the horn behind the couch, as it's the least obtrusive option and interacts with room modes the least.
Width 1.1m
Height 0.8m
Depth 0.4 - 0.6m
Parameters are:
SD 498 cm2
CMS 0.25 mm/N
Mms 228 g (don't know what MMD is)
Re 2.97 ohm
BL 15 Tm
Rms 6.848 kg/s
LeVC 2.95 mH
Thanks for your help guys :D |
|
|
| paulspencer |
Bandwidth is 20 - 80 Hz max, might cross as low as 60 Hz.
While I have managed to model my drivers in Hornresp before, I haven't been able to make sense of the dimensions it works out.
What are some of the critical design parameters in a horn sub?
It's not so much that I have bought these drivers to do a horn sub, but rather I have them and have wondered what a horn sub would be like for them. I haven't had the pleasure of hearing a horn sub yet. (Anyone in Melbourne have some horns?) |
|
|
| AndrewT |
HI,
I'm still new on this horn theory, so please excuse any misconceptions.
Your question should be; how will the response and spl vary in the octave above Fc - a small mouth and short horn will not perform at all evenly.
Below Fc the speaker reverts to box loading without the horn to give reactance. This appears to happen over a very short octave range.
Once the horn is not functioning you will find that driver Xmax will be exceeded for very small acoustic outputs, try putting numbers into hornresponse, iterate various voltages & extract maximum spl over a range of frequencies below FC using xmax at chosen frequency as your limit for each voltage. You will not be impressed.
By the way how did you manage to achieve an Fc of 30Hz in such a small box?
regards Andrew T. |
|
|
| paulspencer |
| quote: | | By the way how did you manage to achieve an Fc of 30Hz in such a small box? |
I didn't
The Lab sub has 2x12" drivers, it's a 27 Hz horn and only slightly larger than I have in mind.
| quote: | | Once the horn is not functioning you will find that driver Xmax will be exceeded for very small acoustic outputs, try putting numbers into hornresponse, iterate various voltages & extract maximum spl over a range of frequencies below FC using xmax at chosen frequency as your limit for each voltage. You will not be impressed. |
I currently have them in 70L sealed boxes and their output is quite impressive already. The excursion is 46mm P-P and I have a combined VD of 4.6L, barely less than the Tumult. |
|
|
| BAM |
| Lab Subwoofers only work their best when you have four of them. A horn sub that you're talking about would need to be built into the floor. Just stick with direct-radiating subs for home use, I say. The AV12's Fs is too low, also. |
|
|
| paulspencer |
I'm certainly not going to get another driver for this. This is just a thought experiment that I might build ...
My driver isn't that different to the Eminence Lab12 driver, which is designed specifically for the Lab sub. Qts, fs, BL are quite similar. The VAS of the Lab is a bit higher (not sure if that's significant).
I modelled my drivers in the Lab sub and the result didn't look good.
I modelled both of mine in a box that measures 1.1 x 0.8 x 0.6m and what was strange was that two model better in a box that size than one. The mouth area was 0.8 x 0.6m. With this size I managed to get a Fc at around 40 Hz or a bit lower using as much power as I can get without coming close to the drivers excursion limits. However, the rolloff is so sharp that even with another 8db of headroom below Fc, it doesn't do much to get any more depth.
How do you judge a horn in terms of simulated performance prior to building it? How do you know whether it is a good design or not?
If Adire can get a 30 Hz horn for the Tempest, then surely I can make one for my drivers to at least get to 35 Hz without it being too much bigger! |
|
|
| BAM |
| I, personally, am going to advise against this, because if you want a horn that won't fill up the room, you're going to lose a lot of the AV12's low end, and then you'll also have trouble getting flat response. I enjoy the low bass that my AV12 puts out. You should go ahead and build those sealed subs you were looking at in the beginning. If you horn load the AV12, its low FS means that below the horn's flare cutoff, the driver will still be trying to put out sound with no help from the horn, and the kind of EQ it would take to equalize that lowered level back up to the rest of the horn loaded output would be huge, and the driver probably couldn't take it. Horns can be fun, but they have their limitations and tradeoffs just like any other speaker enclosure. These are the things I've learned through the course of the design of my very own PA bass horn design, and I've learned them from reading material from very knowledgeable people (in particular, a well-known individual named John Sheerin who evidently lived down the street from me back in my hometown a few years ago) and soliciting their opinions on my design. |
|
|
| paulspencer |
You may well advise against it, but this is a "thought experiment" right now. I'm going to toy with the idea whether or not you think it's a good idea :D ... I still intend to build small sealed subs as I can most probably only live with them in the room.
Looking at it another way, I'm not losing low bass but gaining higher output / or lower distortion in the midbass, where distortion perception is more significant. I just modelled using hornresp with a larger rear chamber. The rolloff was not so dramatic and it was possible to get 110 db within xmax limits at 20 Hz. Above this the response rises up to 130 db and above with more power if 20 Hz signal isn't present. I could then use eq to get the response flat to suit preferences. Seen this way I don't lose extension so much as gain low distortion midbass. Down lower I have to rely on excursion, hence higher distortion, but it's then not so critical as this would be LFE. |
|
|
| BAM |
| What I'm saying is you'd probably overdrive the woofer if you EQ'd up the lower ranges to match what the horn is doing to the upper ranges. |
|
|
| mike.e |
I most definately would NOT use a driver below its horn cutoff frequency on horn. Typical measurements into the double digit THD percent.
IF your getting 104db/1watt at the frequency you desire,with sensible design,yes it will be good!
if you still want low end,youl need another sub for bottom octave-and it will never be able to match the horn output. |
|
|
| paulspencer |
| quote: | | I most definately would NOT use a driver below its horn cutoff frequency on horn. Typical measurements into the double digit THD percent. |
Well over 10% THD is very common for subs below 40 Hz even at moderate levels. The question is how audible is this distortion? THD is in fact meaningless from a perception point of view, as recent research has shown - refer www.gedlee.com
Another sub would be overkill. I have two subs which are capable of 116db at 20 Hz. Horn loading to 40 Hz allows me to either:
* get more output for music (crazy really)
* lower distortion above 40 Hz for music, where low distortion is actually more important
Below fc the subs will merely be operating as sealed subs.
One day I'd like to build some 20 Hz horns for a dedicated room (basement horn type of design). This would then be able to reproduce LFE with astounding fidelity and output and impressive dynamic range. |
|
|
| mike.e |
| quote: |
Below fc the subs will merely be operating as sealed subs.
|
Yes apart from the horn cutoff frequency,where, a low BL driver will have an enormous excursion peak-unloading, heaps of labhorn owners blew drivers when not operating their subsonic filtering correctlly with a few kw :xeye:
One things for sure- you dont want distortion-unless your becoming a valveguy with single drivers...
Try it if you want.Are you only using them for music? Youl have to apply an LT to pull up the bottom end below horn cutoff-but it shouldnt be too much if you build a labhorn size unit + 1metre extension
Yes i found this site today-after he posted it on AA ,against the forum rules i think.
Personally my opinion is that if i can have low THD- why not
:D |
|
|
| paulspencer |
Both for music and HT ... I'm getting ultracurve and I would probably use eq with negative gain to pull back the horn loaded part to match the sealed end of things. I will deal with this and room eq at the same time. Thing is, I'm happy with the LFE output, but for music sometimes I get the urge to crank and get it to chest pumping levels, which I can't do right now without amp clipping!
Have you downloaded the samples? They will challenge what you think about both IMD and THD. They are meaningless. Still, I'm almost certain that horns will reduce all types of distortion, including the non linearities that are perceived subjectively and can be measured with the gedlee metric. This is what is important, not THD.
That makes sense, I have heard of people damaging Lab horn drivers due to overexcursion. I had wondered how this was possible, when they were driving them beyond their thermal limits. But in modelling different options, I have noticed that certain designs can have the excursion increase in a sharp peak around fc. You might have an excursion of 5mm which then jumps up to 30mm at fc ... crazy really. In something like the Lab horn I'd want perspex to see how much the driver is moving if possible. ... and it's fun seeing that thing move.
I like to watch how far the cone moves on LFE in a boys with toys kind of way :D |
|
|
| paulspencer |
| I made an attempt to model the Tempest HOrn today and what surprised me was that it was hard to get anything close to the curve on their site. The F3 was very high and the rolloff was much sharper. The curve shown on Adire's site rolls off gradually below 40 Hz and appears that with room gain you could easily get F3 @ 30 Hz and not lose that much extension at all. Perhaps even get to 25 Hz with a little eq at the cost of a little output. |
|
|
| mike.e |
Well the curve on their site isnt exactly flat.
Try 3metres expo with 2:1 compression ratio and whatever Vrc is required -if theres a midbass hole ,increase volume,if theres a big midbass bump ,decrease Vrc. Change one thing at a time
Make mouth a nice 4,000cm^2 or so.
Give me an email if your keen |
|
|
|
