Basta
Hi,
Not sure if Bagby's justs adds a EQ filter or models cap+driver interaction.
I modelled the driver in 28L 1cuft and get an interesting response using
700 to 800uF in series (also added 0.55R series R) so two in parallel
in ~ 56L / 2 cuft should need 350 to 400uF.
It does flatten the bass peak and extend the bass response.
rgds, sreten.
Hi,
Not sure if Bagby's justs adds a EQ filter or models cap+driver interaction.
I modelled the driver in 28L 1cuft and get an interesting response using
700 to 800uF in series (also added 0.55R series R) so two in parallel
in ~ 56L / 2 cuft should need 350 to 400uF.
It does flatten the bass peak and extend the bass response.
rgds, sreten.
Sreten we seem to be getting similar enough results.
now I am starting to wonder where to put the XO between bass and midrange as I am having trouble with the program ( ever since I upgraded to W-7 and office 2007) I can use any reasonable frequency between 150 and 400 but these woofer probably sound best used below 200.
So I am wondering on the best baffle width now and thinking of making the baffle wider but also considering baffle step problems.
Most modern amps will handle a 4ohm load with ease and at a fairly high DCR of 7.6R I am wondering if I should run 3 woofers with one or two rear firing to counter baffle step.
Where would be the best frequency range to do so assuming either 300 or 450 wide baffles, at the F3 point or somewhere else?/ Higher or lower??
now I am starting to wonder where to put the XO between bass and midrange as I am having trouble with the program ( ever since I upgraded to W-7 and office 2007) I can use any reasonable frequency between 150 and 400 but these woofer probably sound best used below 200.
So I am wondering on the best baffle width now and thinking of making the baffle wider but also considering baffle step problems.
Most modern amps will handle a 4ohm load with ease and at a fairly high DCR of 7.6R I am wondering if I should run 3 woofers with one or two rear firing to counter baffle step.
Where would be the best frequency range to do so assuming either 300 or 450 wide baffles, at the F3 point or somewhere else?/ Higher or lower??
The very first thing I'd do is ask the wife what the largest sized boxes she will tolerate in her living room. Get some cardboard boxes of various sizes and show her. Then, once you have a rough maximum volume to work from, you can design accordingly.
Hi,
I believe the most difficult part of getting a 3 way to work properly is
the marriage of the bass units to to the mid/treble because there are
all sorts of issues and options that need to be balanced.
What sentivitity are the bass units ? specs / link for the midranges ?
What tweeter are you think of using ? specs /link ?
2 cuft internal is still a big box to some. I'm assuming a floor stander,
and you are going with two bass drivers sealed with a series cap.
Two possible options :
You make them deep and narrow for WAF, with dual side firing woofers
This implies "freespace" positioning away from the back wall. This will
have the best imaging (nice for classical) but meant issues dependent
on the relative specs of the bass and mids. the dual bass units won't
suffer bafflestep, but the midrange will, but it needs a low c/o point.
or :
You make them wide and shallow to be used up against a wall. My
guess with two front bass drivers, one close to the floor BSC, is
that BSC might not be an issue, or only a small amount needed.
Here you can c/o somewhat higher.
Regarding c/o point :
Nedd to know more about the quality of the drivers, the level
of ambition, budget, and philosophy regarding c/o simplicity.
For a build where budget is an issue the fist thing I'd investigate
is a cascaded 1st order series c/o because it is so elegant.
But it might be hopelessly inadequate - you have to model and see.
rgds, sreten.
Hi,
Bactracking it seems close to wall positioning is the option, so fairly wide
and shallow seems the best option. (Though NHT's classic models used
the exact opposite with low side firing woofers for this room position).
Depending on the ambition a two box arrangement ? Mid in the top
box (with the c/o most likely) and tweeter below, mounted at the
top of the bass box ? idea here is to keep the midrange away
from large speaker panels, just a thought .....
For simple c/o's the points move up, to get away from the mids
bass resonance, ferrofluid damped tweeters are good, there is
no resonance, here distortion issues raise the simple c/o point.
rgds, sreten.
Bactracking it seems close to wall positioning is the option, so fairly wide
and shallow seems the best option. (Though NHT's classic models used
the exact opposite with low side firing woofers for this room position).
Depending on the ambition a two box arrangement ? Mid in the top
box (with the c/o most likely) and tweeter below, mounted at the
top of the bass box ? idea here is to keep the midrange away
from large speaker panels, just a thought .....
For simple c/o's the points move up, to get away from the mids
bass resonance, ferrofluid damped tweeters are good, there is
no resonance, here distortion issues raise the simple c/o point.
rgds, sreten.
It's paradoxical but music professionals seem to be less concerned by music reproduction when at home, for this instance WAF says smaller is better.
having looked at the room even 1 woofer is going to be enough, the room is quite small even by the standards of the area we live in.
The mids are a cheap generic 5 inch polycone with a surprising big magnet and reasonable sensitivity and I have used them many times before and usually crossed second order electrical in enclosures of about 6 litres and high passed around 4k to a silk domed tweeter with an Fs of 1900Hz that is FF damped and designed for surface mounting.
For Aussies this is the discontinued Digitor 25mm tweeter, surprisingly good for a budget chain and of course no longer available as Dick Smith are no longer an electronics store.
4k was originally chosen to give me an octave above the tweeters Fs even tho it is a little high to cross from a 5 inch mid.
I can probably go down to 3500 with few problems given the low power of the amp that will be used and the fact that the volume will be lower given the small room
having looked at the room even 1 woofer is going to be enough, the room is quite small even by the standards of the area we live in.
The mids are a cheap generic 5 inch polycone with a surprising big magnet and reasonable sensitivity and I have used them many times before and usually crossed second order electrical in enclosures of about 6 litres and high passed around 4k to a silk domed tweeter with an Fs of 1900Hz that is FF damped and designed for surface mounting.
For Aussies this is the discontinued Digitor 25mm tweeter, surprisingly good for a budget chain and of course no longer available as Dick Smith are no longer an electronics store.
4k was originally chosen to give me an octave above the tweeters Fs even tho it is a little high to cross from a 5 inch mid.
I can probably go down to 3500 with few problems given the low power of the amp that will be used and the fact that the volume will be lower given the small room
Snapshot of the mids (2) in a large sealed box at 20watts and a nominal XO, experience and modelling say a second order electrical gives me a 4th order acoustic slope, also that the rising response of the woofer and the peak at the top end and the break-up after 2k mean I'll need a at a minimum a second order electrical to give me a second order acoustic slope although a strange first order may work.
These are not going to be "good" speakers, definitely not state of the art, LOL; but I want them to be many orders or merit better than the $10- computer monitors they are currently using for music.
My budget for building the XO is about $200- AUD and the drivers would cost about the same at retail if I stick to single drivers and I expect that the end result will be worth about 3 days carpenters labour.
He still owes me a half day for the lemon cheesecake of last month
These are not going to be "good" speakers, definitely not state of the art, LOL; but I want them to be many orders or merit better than the $10- computer monitors they are currently using for music.
My budget for building the XO is about $200- AUD and the drivers would cost about the same at retail if I stick to single drivers and I expect that the end result will be worth about 3 days carpenters labour.
He still owes me a half day for the lemon cheesecake of last month
Attachments
Last edited:
It really depends on the genre, but there's plenty of techno that digs into the single digits.
Here's some analysis of various tracks that I did a couple of years back:
BASS! How Low Can You Go? - DIYMA.com - Scientific Car Audio - Truth in Sound Quality
When I added a 15hz 350lb tapped horn to my subwoofer array, I immediately noticed additional 'weight' on tracks like 'One more time' by Daft Punk. This is house music more than techno, though.
My analysis indicates that techno tracks by Richie Hawtin have beats at three hertz.
Last week I went to a Skrillex show, and it was without a doubt the most bass I have ever heard in my whole life. There was easily ten hertz bass, maybe lower. There's an Amtrak train that goes by my house twice a day, and Skrillex was lower and louder. (Skrilex is mostly dubstep btw)
Yeah, there's lots of electronic stuff that goes down there. Sometimes I wonder if it's by accident or not, but the idea is that subsonics aren't just for organ pedals anymore...you're missing quite a bit in a lot of electronic music if you're not down at 20 cycles.
Hi,
So your going one bass driver in about 1 cuft with a series cap ?
If so mount it low and c/o to the mid to avoid the "floor dip".
The tweeter seems pretty good for a simple c/o, contrary to most
opinions, you want a highish Fs and no impedance peak for a simple
treble network.
The c/o will be simple or not depending on the details of the mid unit.
L-pads are assumed for differences in driver sensitivities.
Ideally what you want for 1st order series mid to treble is a point
where both drivers have the same -3dB downpoint. If you c/o 1st
order at that point you will be very near 2nd order L/R acoustic
and the drivers need to be relatively out of phase, one inverted.
But without carefully chosen drivers, this is quite unlikely.
Seems the two box arrangement would work, something like :
No rear port of course and reduced cabinet depth to get ~ 28L bass.
Getting 1st order series to work bass to mid is all about modelling and
choosing the c/o point that seems to work the best, probably ~ 400Hz.
Might not work, and 2nd order parallel electrical networks are the next
option. Targets are bass /mid 2nd order L/R inverted polarity, and often
4th order L/R acoustic, same polarity, utilising the mid and tweeter rolloffs.
rgds, sreten.
So your going one bass driver in about 1 cuft with a series cap ?
If so mount it low and c/o to the mid to avoid the "floor dip".
The tweeter seems pretty good for a simple c/o, contrary to most
opinions, you want a highish Fs and no impedance peak for a simple
treble network.
The c/o will be simple or not depending on the details of the mid unit.
L-pads are assumed for differences in driver sensitivities.
Ideally what you want for 1st order series mid to treble is a point
where both drivers have the same -3dB downpoint. If you c/o 1st
order at that point you will be very near 2nd order L/R acoustic
and the drivers need to be relatively out of phase, one inverted.
But without carefully chosen drivers, this is quite unlikely.
Seems the two box arrangement would work, something like :
No rear port of course and reduced cabinet depth to get ~ 28L bass.
Getting 1st order series to work bass to mid is all about modelling and
choosing the c/o point that seems to work the best, probably ~ 400Hz.
Might not work, and 2nd order parallel electrical networks are the next
option. Targets are bass /mid 2nd order L/R inverted polarity, and often
4th order L/R acoustic, same polarity, utilising the mid and tweeter rolloffs.
rgds, sreten.
Last edited:
Yes I wouldn't want to go any higher than 400.
As for trying to reproduce very low bass, the budget will not allow it so the cap in series will perform a dual function, WAF will not allow the 2 * 44gallon drums in each corner needed to reproduce 20Hertz and lower even if the small room made that practical.
Series XO seems reasonably well suited to the project, I assume I use the normal Zoble conjugates I have always needed with these cheaper drivers??
I have built boxes similar to you picture and they have worked for me in the past, although they all used multiple 8inch woofers the building of that style is relatively simple.
You help is appreciated Sreten, thanx again
As for trying to reproduce very low bass, the budget will not allow it so the cap in series will perform a dual function, WAF will not allow the 2 * 44gallon drums in each corner needed to reproduce 20Hertz and lower even if the small room made that practical.
Series XO seems reasonably well suited to the project, I assume I use the normal Zoble conjugates I have always needed with these cheaper drivers??
I have built boxes similar to you picture and they have worked for me in the past, although they all used multiple 8inch woofers the building of that style is relatively simple.
You help is appreciated Sreten, thanx again
Hi,
Generally the point of series 1st order crossovers is to utilise driver inductance
rather than nobble it with zobels. The mid Fbox impedance peak is also far less
of an issue with series rather than parallel 1st order c/o networks.
Take the 1st order series bass to mid c/o. the bass unit has a cap across it
but is being fed current from the inductor across the mid. Its inductance
can be seen as an extra series c/o element to a 8 ohm resistive driver.
There is no point "bypassing" this element unless its not useful, if you
do the most effect is on the mid unit, series c/o's are weird beasts.
A first order parallel network capacitor to a mid will cause a peak at the mids
box resonance, but the opposite effect occurs with a first order series network.
So no zobels unless they are needed for 1st order series, they usually aren't.
Their effects can be quite counter intuitive in 1st order series networks.
But they are more inflexible than parallel, you can't offset electrical c/o
points for an acoustic result like parallel, they are difficult to get right.
But so is speaker design in general, choosing the right options is critical.
rgds, sreten.
http://sites.google.com/site/undefinition/diy
(see if nothing else, the excellent FAQs)
http://techtalk.parts-express.com/showthread.php?t=219617
http://www.rjbaudio.com/Audiofiles/FRDtools.html
http://web.archive.org/web/20090902124715/http://geocities.com/woove99/Spkrbldg/DesigningXO.htm
Generally the point of series 1st order crossovers is to utilise driver inductance
rather than nobble it with zobels. The mid Fbox impedance peak is also far less
of an issue with series rather than parallel 1st order c/o networks.
Take the 1st order series bass to mid c/o. the bass unit has a cap across it
but is being fed current from the inductor across the mid. Its inductance
can be seen as an extra series c/o element to a 8 ohm resistive driver.
There is no point "bypassing" this element unless its not useful, if you
do the most effect is on the mid unit, series c/o's are weird beasts.
A first order parallel network capacitor to a mid will cause a peak at the mids
box resonance, but the opposite effect occurs with a first order series network.
So no zobels unless they are needed for 1st order series, they usually aren't.
Their effects can be quite counter intuitive in 1st order series networks.
But they are more inflexible than parallel, you can't offset electrical c/o
points for an acoustic result like parallel, they are difficult to get right.
But so is speaker design in general, choosing the right options is critical.
rgds, sreten.
http://sites.google.com/site/undefinition/diy
(see if nothing else, the excellent FAQs)
http://techtalk.parts-express.com/showthread.php?t=219617
http://www.rjbaudio.com/Audiofiles/FRDtools.html
http://web.archive.org/web/20090902124715/http://geocities.com/woove99/Spkrbldg/DesigningXO.htm
Last edited:
Are you building a speaker that YOU would be happy with, or are you building a speaker that your CLIENT will be happy with?
The reason that I ask this is that it's easy to do the former, but in engineering, you always want to do the latter. Otherwise you end up with very unhappy clients.
Classical music and techno music goes LOW, and if you want to reproduce them properly, your speakers should too. It's a whole different set of requirements than the typical 'polite' audiophile two-way. Techno is uninvolving on a system with limited dynamics or bandwidth.
My speakers already have multiple 44 gallon drums in the corners, so no; I'm trying to build my mate the biggest speakers his wife will allow him which is relatively small in comparison.
I mentioned the large drums merely to emphasize the point that I can't aim to reproduce very low bass.
Anything is better than his current system tho, which I mentioned are 3watt computer speakers with a 3inch * 2 inch oval full range driver.
If only he hadn't already demolished his brick chimney we could have used that to house the 2 * 15 inch subs I have spare on the shelf .
In the future he could add a subwoofer for the last octave, I will aim for an F3 of 40hz at least then he will get the first harmonics which he doesn't at the moment
I mentioned the large drums merely to emphasize the point that I can't aim to reproduce very low bass.
Anything is better than his current system tho, which I mentioned are 3watt computer speakers with a 3inch * 2 inch oval full range driver.
If only he hadn't already demolished his brick chimney we could have used that to house the 2 * 15 inch subs I have spare on the shelf .
In the future he could add a subwoofer for the last octave, I will aim for an F3 of 40hz at least then he will get the first harmonics which he doesn't at the moment
Modelling suggests that an F6 of 40hz is easily achievable an F3 less so but there is a school of thought that the F10 is the figure to chase.
If I make the box size slightly larger it becomes a little louder for the given input and I model an acceptable F10 of 32hz if I change the first order high pass cap to a nominal value of 25hz.
Would I be happy with it? of course not but I spend a large dollar figure to get my F10 to 18Hz which my mates budget and WAF will not allow, given this I think the basic design will be acceptable to them, at least I hope this will be the case.
Close to the wall and wider shallower boxes, and as I have some baffle panels that are 450mm wide already to go I guess that is what I will be using
If I make the box size slightly larger it becomes a little louder for the given input and I model an acceptable F10 of 32hz if I change the first order high pass cap to a nominal value of 25hz.
Would I be happy with it? of course not but I spend a large dollar figure to get my F10 to 18Hz which my mates budget and WAF will not allow, given this I think the basic design will be acceptable to them, at least I hope this will be the case.
Close to the wall and wider shallower boxes, and as I have some baffle panels that are 450mm wide already to go I guess that is what I will be using
Hi,
True, but I've no real idea if 1st order series will work here.
But anyway IMO its more likely to work well than 1st order parallel.
rgds, sreten.
As I have never built a series XO it is way past time I tried one, also those mid and tweeter components needed something other than a first order XO to integrate well in the past, as well as Zobels etc:
Every new speaker build is an experiment anyway and this will be my first wide shallow box too
Every new speaker build is an experiment anyway and this will be my first wide shallow box too
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.