NEW (i think) hybrid

Ah, well you're not the first one.

+ 1 on Sketchup.

To give an idea on Hornresp, the next pic is of Cubo Kick 12 earlier prototypes, where the main difference is the compression ratio (green lower CR, blue higher CR, red 12" reflex of ~ equal size). The peak shifts are larger, sometimes wider and sometimes more narrower then predicted.

Also 4,0 pi is generally to pessimistic on low frequency response, 2,0 pi too optimistic (as placement is somewhere inbetween).

Best regards

Cubo

edit- This is the result of a small change but not even a large change simulates quite this drastic (to my knowledge).

how exactly do i calculate the CR (sorry if thats a dumb question, it probably shows how ambitious i am being in trying to design this). im amazed to see it makes such a huge difference in the lowewer range. is that due to a higher mechanical impedence on the cone limiting displacement at a given power/frequency? its a very steep roll off (cut off more like) in your graph above, i dont understand why and im very curious to have some idea as to whats actualy happening there.

you mentioned an octagonal phase plug on you design. is the back molded to the shape of the cone, or is it also octagonal? im starting to think that the shape and placement of the phaseplug will be of much less importance at these frequencies than i originaly thought.

my idea of putting the phase plug in the chamber was based in the thought that having a more uniform pathlength between the driver and horn throaght would help with the higher frequencies, and also keeping the chamber the same volume, but with a smaller and more uniform cross section of air as the narrowing chamber meets the smallish S1.

does it matter if the chamber expands/contracts from the ATC to the S1? im not sure if i have asked that question correctly. am i right in assuming that the correct expansion rate of the horn would be unaffected by the expansion rate (compression rate realy) in the front chamber, as long as the pathlengs are more or less similar, and make the transition in a fluid ´no sharp edges` way? im now thinking i would want to keep the acoustic resistance constant through the lenght of the chamber as the overall cross sectional area of air shrinks to the area of S1. i can see how an octagonal phase plug matching the octagonal walls of the front chamber and S1 would make this easier to aproximate.

I am also doubting how much a phase plug of this size can actualy do to correct pathlengths with these quite long wavelengths. im obvioulsy lacking fundimental knowledge here and this design has and will be a massive learning curve so sorry if some of my ideas and questions are a bit dumb, or unclear.

the phaseplug thing just looks right to me (i say looks right, not looks cool .... just so my lack of knowledge dont come across as a lack of inteligence). its only logic and assumptions, but i just cant see the shorter waves getting through the chamber and into the smaller throaght coherently unless everything is done to make that path as phisicaly/acousticaly smooth as possible.

Oh by the way. i just noticed a large electiacal impedence spike right where i have the high power frequency responce/displacement spike (just above port tuning frequency, about 40ish hz below the start of the horn roll off, so i imagine the problems may fix each other there.

anyway i guess what i just wrote is a bit of an incoherent rant, perhaps i need some pathlength correction between my brain and mouth/fingers. as i said the phase plug in the chamber is something that is realy beyond me at the moment. if you could point me in the way of some good reading about this kinda think that would be great

cheers
dave

thepersonunknown wrote: i just posted a longish reply and id didnt load, so ill just give a quick summery.

If you use Firefox you can use one of the add-ons for it called Lazarus, this automatically saves any text that you type and is then available for any loss of original text,

addons.mozilla.org/en-US/firefox/addon/lazarus-form-recovery/

Tony

@ Tony

yea i have that on my laptop, but it just eats up too much RAM on my PC (only 1 gig :blush: , its a real old tractor, but i love it to death). unfortunatly i have to keep it down to just ad block and the smooth mouse wheel scrolling thing which is great for reading. i also have one for ripping mp3s of any source (hand on heart that i only listen to them on my pc and i pod though)

thanks for the tip though, its a good one, and +1 on that if anyone else reading has this problem and dont mind losing a few mb of ram off their system

cheers
dave

Edit: i only have this problem here when i hit the quick reply button. i now copy everything before posting and the rant above, for example, was saved in this way.

could be a possible bug? or am i the only one

how exactly do i calculate the CR

CR is Sd divided by the horn throat. So if Sd is 880 and S1 is 440 than the CR 2:1 (or just 2) .

is that due to a higher mechanical impedance on the cone limiting displacement at a given power/frequency?

Never looked at it that way but it makes sense

it’s a very steep roll off

? It’s about 6 dB/ octave.

you mentioned an octagonal phase plug on you design. is the back molded to the shape of the cone, or is it also octagonal?

I think it was octagonal but flat. The main purpose was to raise the CR and adjust the first part of the horn path accordingly.

my idea of putting the phase plug in the chamber was based in the thought that having a more uniform pathlength between the driver and horn through would help with the higher frequencies, and also keeping the chamber the same volume, but with a smaller and more uniform cross section of air as the narrowing chamber meets the smallish S1

Usually (one of reasons) a phase plug is used, is to minimize the volume of the front chamber that is difficult to manage by woodwork only. If managed, this can partially aid to the horn path, which to me sounds what you’re doing (unless I’m not seeing the right picture).

For instance the narrowing chamber on the Xtro double 15” section, does act in my view, mostly as horn path and only for a small portion as chamber.

I am also doubting how much a phase plug of this size can actually do to correct path lengths with these quite long wavelengths

The next graph is from Cubo Double Kick (started a few days ago) and should expain much.

The red line is without a phase device, the green line is with phase device (which does raise the CR a bit as well) and it lowers the path length differences by as much as 30 cm. The blue line is 2 x 12” reflex of similar size.

Cubo Double Kick.bmp
Note: This one isn't reflex assisted yet.

Best regards

Cubo

thanks for they reply. realy i appreciate it a lot, and im sure someone in future will stumble over this thread and be thankful all over again, cos there isnt so much info about this.

ok ill try to be as clear as possible and explain myself well, because unless i have misunderstood you one of the assumptions im baseing this design on if fundimentaly wrong.

I see (and sim this as a BPH), where the shape of the front chamber isnt of such importance, as it dont count towards the flare rate of the horn. after what i have been simming as the throat (S1, octagonal), there is a short section of horn expanding slightly to a rectangular S2, then a longer section expanding quickly to the rectangular S3 (mouth).

however what i have taken from your post (which correlats with an experiment which i have read about carried out by centauri over on the other side) is that what i am seeing as the front chamber is realy the first section of the horn, and the chamber is nothing more than the air between the plug and the cone with the throat being the ring between the widest part of the phase plug and the chamber walls. do you see what i meen?

So its not realy a BPH ar all, its just a straight horn with a very narrow first section filled out to a larger, hopefully more phase coherent diameter by the phase plug. if ive understood you correctly with this then i realy need to go back to the drawing board and make some heavy changes to the way im simming.

so i actualy need to put great importance on the expansion rate within the chamber to keep the correct flare rate through the length of the horn (as the ´chamber`is actualy the first section of the horn, and not a bandpass chamber at all), not to mention sim the horn as a 3 segment horn, and ditch almost all of the front chamber volume.

this could be good news. i think having that little extra lenght to the horn will easily compensate for what goes with the fron chamber. i can also see this as being more coherent now.

Ok i think ive got my hear around that, im going to do a quick sim or two.

i guess a 6db drop because of changed CR isnt so much, but it supprised me, as i thought it would be much less. once again, a lack of knowledge. a lot of things that have come up in this design im facing for the first time. i meen everyone knows about CR and the like, but actualy working out exactly what effect they have and the controling it is another thing. the help, i repeat, is priceless.

the graph directly above makes it very clear the effect of the phase plug. very clear indeed. what do you think is causing the dip there around 250hz. i imagine its something you would rather not have, also you horn plays much lower than what ive been doing, perhaps thats just hornresp, but with a reflex port tuned so as to have flatish responce between tuning frequency and horn rolloff, i would expect you plan to get at least 20hz lower than mine.
personally i would trade off some lows to try and lift the 250hz dip as much as possible. i have no idea if thats a trade off you can chose to do in your design. i dont want to pry, and feel free to not give me info on this, but what would you layout look like. similar to mine? as i said, feel no obligation to answer that question at all. also what do you think of my driver choice? it seemed unconventional to me, but the criteria i was working with pointed me to it so...

anyway thanks again. im going to run a couple more sims and ill be in touch

cheers
dave

and the chamber is nothing more than the air between the plug and the cone with the throat being the ring between the widest part of the phase plug and the chamber walls

That describes the way (I think) most people see it.

So it’s not really a BPH ar all, it’s just a straight horn with a very narrow first section filled out to a larger, hopefully more phase coherent diameter by the phase plug

That’s pretty much it. Though it has some BPH effect . Making the throat chamber smaller (which in combination with the horn acts as a low pass filter) will extend the high frequency response of the horn. Whereas lowering the throat chamber in volume has a negative effect on the low frequency response (as does the higher CR). That should answer another question as well.

so i actualy need to put great importance on the expansion rate within the chamber to keep the correct flare rate through the length of the horn (as the ´chamber`is actualy the first section of the horn, and not a bandpass chamber at all), not to mention sim the horn as a 3 segment horn, and ditch almost all of the front chamber volume

How’bout calling it the first horn section from now on

what do you think is causing the dip there around 250hz. i imagine it’s something you would rather not have, also you horn plays much lower than what I’ve been doing, perhaps that’s just hornresp, but with a reflex port tuned so as to have flattish response between tuning frequency and horn roll off, I would expect you plan to get at least 20hz lower than mine

That dip is due to the specific design (it isn’t the octagonal one). It would be nice if the dip wasn’t there but I’m willing to disregard 250 Hz and up, if the low end excels.

I’m betting on a total of 5 horses (plus numerous deviations on those 5), of which 2 are octagonal. The latter having the best 200 – 350 Hz response.

PD 12SB30. i know it sounds like a strange choice given it should be a sub driver[./quote]Calling it a sub driver isn’t my cup of tea (regardless of PD), tho in the right design it will do.
Anyway, Le might be a bit on the high side but basically it’s a good low-mid driver (at least in my book).

Best regards

Cubo

Might as well add, that if you can approximate the three segment horn as an one-segment horn (hyperbolic-exponential flare rate), Hornresp allows you to calculate the extended response due to it's directivty. Without it, it's half the story.

Best regards

Cubo

Hay.
just a quick update, as i dont have much time now.

i was simming this weekend following what we talked about here. (3 sgement horn, rather than 2 segment BPH)

it looks good, very good. the responce/sensitivity is very similar, but the sims look flat to over 400hz, which is way higher than i thought would be possible when i first started with this.

the biggest improvements i am seeing though, are with the group delay, which sits very flat, apart from the spike at port tuning frequency, phase responce, which is almoes flat between 100hz and 350hz (and i meen flat, like a ruler!!!) and the impulse responce.

the impulse responce is realy different to the one i was simming as a BPH. the difference realy is night and day. ill get some pics of all this up tomorrow or the next day, but after the inicial pulse, there is almost no swing to the other phase, and the line just wiggles a bit after that. i have noticed that there is a little negative spike just before the positive/0 milisecond one. i have noticed this seems inevitable with ported horns, although im not sure y (i dont think thats a very important point though, and havent given it much thought. i imagine its some way related to the phase shift below tuning frequency or something of the like)

I still have this huge spike in output when simming at high power levels, matched by the driver being driven over xmas, but once again i notice a huge spike in electonic resistance at the same frequency, so i think thoes problems should solve each other. ill be sure to post up some data on what im talking about though, cos it seems quite strange to me and im not sure what causes it. i understand that xmas problems will start below the horn roll off due to driver being unloaded, but it seems strange that a port makes the problem worse. i can understand how the larger xmas displacing more air causes a sensitivity spike, but i dont realy know how that ties in with the high electiacl impedence. the reason is probably realy simple and right under my nose, but ive not worked it out yet...

I have been playing with ideas for the CR. i realy didnt know it would be of such importance in the lower range, but i am not taking it very seroulsy. you observations have further confirmed to me the necesity of going octagonal for as much of the horn as posible, however much this may complicate the build. im even thinking about adopting you idea for an octgonal phase plug, as it will be easier to aproximate the expansion in the first section of the horn.

anyway ill leave it at that for now, but ill post some pics/observations, and perhaps a question or two over the next couple of days.

thanks again

cheers
dave

PS: yea i think you idea of just clling it the first section of the horn is a good one. the bad thing is that in that part of my post you quoted, i was actualy trying to be clear.. :oops: