Hi,
No. I said you were misrepresenting the veracity of
often relatively complex crossovers you posted for
specific drivers, based on using other driver models.
Its not awful, but there is some guesswork. As soon
as the ambition for the x/o exceeds the guesswork
for tricky drivers, your batting on a very sticky wicket.
And its not like some of the drivers are unknown, its just
you can't create data files to import them into Boxsim.
rgds, sreten.
There IS a good reason to recess bass on the baffle if you wire the tweeter with negative polarity. And you sometimes might have to do that to get any sort of phase alignment on a flat baffle as I discovered with 8" bass.
Strangely, if you time align, you don't need the notches to line it all up. Troels is doing that sort of thing these days:
18W-8434G00
But I'm pleased with this project as proof of concept. I'll probably modify it a bit next based on experience. The bafflestep coil was too small, which brings voices too forward, and I'll rustle up a 2.5mH next I think. Wilmslow Audio only offered 1.8mH and 3.3mH which was awkward. The TW70 tweeter is also very efficient, and I need to rethink the attenuation. The 4uF capacitor in the 6kHz bass notch is select on test, but I think I got it right here. For slightly bigger 6" woofers, 4.7uF (5kHz) is probably right. The BBC dip is a good thing too.
I tried various bass and treble circuits, and this is the best sounding one for 6" IMO. But having put a lot of effort into the circuit, this speaker is really just plain too small, 20 litres would be more like it.
Below is where I think this speaker should be, based on the experience here.
Attachments
Hi,
Two wrongs don't make a right and the fundamental problem
is usually the tweeter being forward of the mid bass. If your
"solution" involves moving the tweeter forward or midbass
back then its really not a very clever solution at all IMO.
It is bound to increase vertical lobing assymetry.
Wiring the tweeter out of phase has nothing to do with
the issue and here is a fine example of it being needed :
Zaph|Audio
It is required for 2nd order L/R acoustic and here
there is no suggestion of increasing acoustic offset,
quite the opposite, it has low BM/T acoustic offsets.
Any design that requires you to recess the bassmid
is skating on very thin ice as a meaningful design.
rgds, sreten.
Two wrongs don't make a right and the fundamental problem
is usually the tweeter being forward of the mid bass. If your
"solution" involves moving the tweeter forward or midbass
back then its really not a very clever solution at all IMO.
It is bound to increase vertical lobing assymetry.
Wiring the tweeter out of phase has nothing to do with
the issue and here is a fine example of it being needed :
Zaph|Audio
It is required for 2nd order L/R acoustic and here
there is no suggestion of increasing acoustic offset,
quite the opposite, it has low BM/T acoustic offsets.
Any design that requires you to recess the bassmid
is skating on very thin ice as a meaningful design.
rgds, sreten.
Last edited:
sreten, let's not get into one of our famous spats here.
A centimetre of alignment can make 40 degrees difference to phase at the sort of 3.5kHz crossover you might use with an 8" bass: H1659-08 U22REX/P-SL
If that means you fit the driver to the back of the baffle, rather than the front, I'm willing to consider it. I've done designs where it helps.
So is Alan Shaw of Harbeth with the lovely M30.1 monitor: 6moons audio reviews: Harbeth M30.1
It may have been a BBC consideration in doing this sort of thing to protect the bass unit a bit too. But it works. It's also easy on carpentry!
I'm mostly reading Troels' master class on time alignment right now. Very good indeed:
http://www.troelsgravesen.dk/DiscoveryW18.htm
A centimetre of alignment can make 40 degrees difference to phase at the sort of 3.5kHz crossover you might use with an 8" bass: H1659-08 U22REX/P-SL
If that means you fit the driver to the back of the baffle, rather than the front, I'm willing to consider it. I've done designs where it helps.
So is Alan Shaw of Harbeth with the lovely M30.1 monitor: 6moons audio reviews: Harbeth M30.1
It may have been a BBC consideration in doing this sort of thing to protect the bass unit a bit too. But it works. It's also easy on carpentry!
I'm mostly reading Troels' master class on time alignment right now. Very good indeed:
http://www.troelsgravesen.dk/DiscoveryW18.htm
Hi,
Again two wrongs don't make a right.
And don't quote fashion as a design intention,
rear mounted drivers are BBC style, aesthetic.
I'd never use a 3.5KHz x/o point with an 8" nowadays.
(Due to off axis flaring and inconsistent power response.)
But if you did with a simple x/o I'd have to grudgingly
agree, though by definition vertical lobing will be poor
if 1cm really matters to the on axis response.
rgds, sreten.
Again two wrongs don't make a right.
And don't quote fashion as a design intention,
rear mounted drivers are BBC style, aesthetic.
I'd never use a 3.5KHz x/o point with an 8" nowadays.
(Due to off axis flaring and inconsistent power response.)
But if you did with a simple x/o I'd have to grudgingly
agree, though by definition vertical lobing will be poor
if 1cm really matters to the on axis response.
rgds, sreten.
In an ideal world, we'd always time align I suppose, but that is not what we do. We tend to use flat baffles. And I assure you that Harbeth speakers are well designed, not just whim to fashion.
Alan Shaw also likes the high 3.5kHz crossover as it goes. It makes tweeters sound better.
I won't bore everyone with an example where recessing the bass helps phase, but I have one. It's an asymmetric crossover, FWIW. But if phase is aligned at crossover, lobing is horizontal, simple as that.
I know YOU would never use 3.5kHz crossover with an 8" bass because you subscribe to Zaph and all that low crossover stuff, but I and Troels often DO, and it sounds sweet! Somehow, when a speaker is right, you just hear the music.
Alan Shaw also likes the high 3.5kHz crossover as it goes. It makes tweeters sound better.
I won't bore everyone with an example where recessing the bass helps phase, but I have one. It's an asymmetric crossover, FWIW. But if phase is aligned at crossover, lobing is horizontal, simple as that.
I know YOU would never use 3.5kHz crossover with an 8" bass because you subscribe to Zaph and all that low crossover stuff, but I and Troels often DO, and it sounds sweet! Somehow, when a speaker is right, you just hear the music.
I was just explaining why a recessed bass is not as daft as it seems. Whether there is a better solution is for you to decide. This is do it YOURSELF after all.
But my main interest here is in using those old retro cone tweeters. No-one likes them. Except me, it seems. Someone has got to do the work.
We had a similar sort of discussion about those strange D'Appolitos with TWO tweeters. Really, I think I was the only person who GOT IT!
http://www.diyaudio.com/forums/multi-way/248734-why-2-tweeters-3.html#post3772957
Interesting. Anyone for an MTTM design next?
The thing TO GET, surely, is that the standard MTM DOESN'T work!
Dear old Lynn Olson, who I love like a brother for his insights, is probably kicking himself... "Beyond the Ariel" was an Ariel with two tweeters.
Last edited:
After that wander off-topic, I thought I'd update you on progress on the Mordaunt Short MS15. Which looked like this when I picked it up:
I decided the LCR section was completely unnecessary, and easier replaced by a tank notch on the bass. I went conservative with 33R on the bass notch since I was guessing on 5kHz for a 6" bass. 15R would probably work better if it's accurate.
But with LOADS of stuffing, and the 2.5mH coil rather than the suboptimal 1.8mH original, it's now sounding very good indeed. Very pleased with this design which also uses the Visaton TW70 cone tweeter. The tiny 10L box sounds very good on bass too.
An externally hosted image should be here but it was not working when we last tested it.
I decided the LCR section was completely unnecessary, and easier replaced by a tank notch on the bass. I went conservative with 33R on the bass notch since I was guessing on 5kHz for a 6" bass. 15R would probably work better if it's accurate.
But with LOADS of stuffing, and the 2.5mH coil rather than the suboptimal 1.8mH original, it's now sounding very good indeed. Very pleased with this design which also uses the Visaton TW70 cone tweeter. The tiny 10L box sounds very good on bass too.
Attachments
Hi Steve just wanted to mention that my current bench speakers are MS 3.10, a 5 1/4 2way with the baby Audax tweeter in a Mission style FRP baffle. All diffractionized and stuff. The bass mid is a poly attempt at what Lynn O used in the Ariel, which can never be approached by anyone ever. Certainly not the SEAS P13RCY pair that I also have. That I sold new in the 80s for an autosound application and much later found in a second hand store with frame mods that only a freak like me would have bothered with just to get it to fit a Ford. I bought the little Audax 25 at a time for car installs, and when I ran out I went over to Radio Shack and paid an extra dollar for them. Three something as I recall; it's 30 years ago. Audax then created a larger version of the TW 54/60/6X9 which was bad, like the Infinity Polysphere was bad, unless the cap over the center was removed. Sometimes scaling up doesn't work.
I have a lot of respect for mylar tweeters, especially the balanced drive Audax derivatives. But the tweeters in the MS15 were just too cheap and cheerful for my taste.
The MS15 was an awful design IMO. You really can't run a 6" paper bass off a 0.4mH coil. Too much cone breakup at 5kHz. That's really what I was trying to correct. I was also trying to emulate a WLM Stella to some extent.
I now think the WLM Stella has a a series filter, but I have a notion to build a 5" polycone design next in best Mordaunt Short MS10i tradition. Polycones just have less cone breakup by nature because they rolloff faster. So I have some 4 ohm drivers lying around in a centre speaker that look promising. Mike Creek at Epos did a better filter on the Epos ELS-3.
The MS15 was an awful design IMO. You really can't run a 6" paper bass off a 0.4mH coil. Too much cone breakup at 5kHz. That's really what I was trying to correct. I was also trying to emulate a WLM Stella to some extent.
An externally hosted image should be here but it was not working when we last tested it.
I now think the WLM Stella has a a series filter, but I have a notion to build a 5" polycone design next in best Mordaunt Short MS10i tradition. Polycones just have less cone breakup by nature because they rolloff faster. So I have some 4 ohm drivers lying around in a centre speaker that look promising. Mike Creek at Epos did a better filter on the Epos ELS-3.
Attachments
As the owner of a pair of MS soundboxes, I can't comment on your choice of speakers, but I am most disturbed by the prominent display of your obviously treasured copy of a George Michael album - or is it empty and you just like the picture?
And WHAT is wrong with enjoying George Michael?As the owner of a pair of MS soundboxes, I can't comment on your choice of speakers, but I am most disturbed by the prominent display of your obviously treasured copy of a George Michael album - or is it empty and you just like the picture?
Like poor old Michael Jackson, a slightly troubled soul, and he was probably not being totally honest in tweaking thousands of girl's hearts with longing:
https://www.youtube.com/watch?v=izGwDsrQ1eQhttp://
But TBH, I just put the CD there for scale. I more usually use Mary Chapin Carpenter or Diana Krall these days.
An externally hosted image should be here but it was not working when we last tested it.
Now Diana Krall is my kind of singer.
https://www.youtube.com/watch?v=7sxK8ghb9PU
Sweet.
FWIW, I've abandoned that MS15 as being a thoroughly bad job. Cabinet too small at 10L. Broken and unfixable.
Well, I've surprised myself! Seem to have fixed these at last.
That big 2.5mH bass coil and 4.7uF just wasn't making an involving sound on the 6" bass.
http://www.diyaudio.com/forums/multi-way/221923-upgrading-mordaunt-short-ms15-bookshelves-6.html#post3980894
So I took it down to 1.5mH and 6.8uF plus the same tank notch of 33R and 0.33uF which now acts at 7kHz. Once you rest them on something, it adds a couple of dB to the modelled bass below 300Hz and it all sounds right. One of the ancient basses sounded rough when loud, so I rotated it 90 degrees to stop the voicecoil rubbing. I think I am totally happy with these now.
The Visaton TW70 Cone tweeter has a bit less top than the Monacor HT22, but was small enough to fit.
That big 2.5mH bass coil and 4.7uF just wasn't making an involving sound on the 6" bass.
http://www.diyaudio.com/forums/multi-way/221923-upgrading-mordaunt-short-ms15-bookshelves-6.html#post3980894
So I took it down to 1.5mH and 6.8uF plus the same tank notch of 33R and 0.33uF which now acts at 7kHz. Once you rest them on something, it adds a couple of dB to the modelled bass below 300Hz and it all sounds right. One of the ancient basses sounded rough when loud, so I rotated it 90 degrees to stop the voicecoil rubbing. I think I am totally happy with these now.
The Visaton TW70 Cone tweeter has a bit less top than the Monacor HT22, but was small enough to fit.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.