Extrapolating Dave’s baffle step compensation approach?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dave outlines six methods of dealing with baffle step loss: http://www.t-linespeakers.org/tech/bafflestep/intro-bds.html

Baffle step compensation in a typical crossover varies from 3 to 5.5 dB, gaining flat response at the loss of sensitivity. Especially with lower powered amps (tubes, class D) it’s desirable to maximise sensitivity . .

Dave's favorite approach (above) adds a further mid-woofer on the back (bipole).

If BSC was left out of a 2 way's crossover, and instead of adding a mid-woofer, a sub were added with an active crossover (eg a Behringer DCX 2496) having shelf filter abilities, with tweaking of the shelf filter and phase, would you agree that the subs could provide the compensation for the baffle step loss; so the sensitivity could be raised.

Eg with Jay’s MTM Usher 8945P / Peerless HDS http://www.geocities.com/woove99/Spkrbldg/Usher_MTM/Usher_MTM.htm , sensitivity could be

Usher averages 87 + 6 (two drivers) = 93 dB
Peerless averages 93 dB

A trade off is that 200 Hz is into the range of eg voices. But with
low powered amps to gain 6 dB, that’s not necessarily a major deficit . .
A benefit is that as the bass load of the mid-woofer is greatly reduced, they should preform better . .
A sub would also go deeper.

What do you think?
 

GM

Member
Joined 2003
otto88 said:

would you agree that the subs could provide the compensation for the baffle step loss...........

A trade off is that 200 Hz is into the range of eg voices.

Works for me as long as the 'sub' driver has a wide enough BW to be nominally flat through the XO BW with no strong break up modes above it. I do the same thing except with a 500 Hz/2nd order XO.

Not much of a trade-off if you look at how poor our hearing acuity is down low........... I mean look at a phone's BW, its low cut-off is 250 Hz.........

GM
 
max Hz from a sub

Most want to XO subs at 40 - 80 Hz, that seems a waste.

Most good subs are very clean way beyond that, eg an engineer at the former Peerless (now Tympani) said that the XLS 12 driver with a 4th order XO could be crossed at up to 300 Hz

So i'm right that sensitivity is gained for the mid-woofers . . and there are no other drawbacks?
 
Hi,

The Watt / Puppy did this yonks ago albeit with stereo subs.
(and proper high pass filtering of the mid unit to boot)

see http://www.deadwaxcafe.com/vzone/david/david.htm

FWIW it is also one of the many ways of implementing a 3-way
speaker. a 3 way can be designed with BSC in the bass, in the
mid, or shared between the two in any ratio. No wonder then
that people get 3 ways so badly wrong, except for sub + sats.

:)/sreten.
 
I think I must be a little confused. I mean bass is too directional by 300-500hz, thats lower midrange territory. You wouldn't want a single subwoofer centrally placed acting as baffle step. If it was included into the main speakers, then yeah, but at that point, we are talking about a fairly common 3-way approach. I'm thinking I must be misunderstanding what you are asking, so please correct me here.
 
There's a lot to be said AGAINST using subs for 300Hz. In addition to localization issues, there's lobing, poor transient response and breakup suppression concerns, and other stuff.

If you want the bass driver to run that high, a good solution is to pick a passive driver with the extra 3dB of sensitivity, and use it as a woofer proper, with an XO at the bafflestep frequency (assuming passive XO).

But using subs.... just asking for trouble. If you like muddy mids, it's a fine solution. Subs are best crossed LOW LOW LOW, to keep them out of the more sensitive ranges of our hearing, and from interfering with any but the deepest of instrumental fundamentals.
 
pjpoes
I didn’t say stereo subs, because I thought any (sensible) use of subs over about 80 Hz would be stereo, my oversight.

I only mentioned 300 Hz because a Peerless engineer said that their XLS’s would go that high. I’m assuming for the time being that baffle step would occur around 200 Hz +/- 50 Hz.


Sreten
A tad off topic, but in the Watt / Puppy, what did you mean by *proper high pass filtering of the mid?

badman
> localization issues, there's lobing, poor transient response and breakup suppression concerns, and other stuff
Are you assuming a mono sub: if not, can you elaborate?

The drivers I want to use - because I have them, and they achieve a higher efficiency tube friendly result are:

Tweeter: Peerless HDS 810921 (used in eg the Elsinore and Jay’s systems with Ushers) 93 dB
Mid: a B&C (cant recall model, Fs I know is 117 Hz) ~ 95 dB
“bass/ sub”: Eminence Kappa Pro ~ 95 dB

If used up to say 250 Hz, how would the Eminence muddy the sound?


One thing that’s since occurred to me is that reasonable inductors at say 250 Hz would cost a bit, tho’ I have no idea how much.
 
I actually would consider a 3.5 way or a 3 way. The setup may be too efficient in the bass if designed as a typical three way. If you want higher impedance in the bass, you could wire the woofers in Series, then design as a typical 3 way. But really, I think a 3.5 way would be a potential option.

I don't agree that there will be any problems such as lobing. This design will actually work quite well I think, and you could run those drivers higher if you wanted. I would have no problem running those woofers to 500hz. To be honest, I think they don't begin having lobing problems until around 1khz or maybe just below that.

If you need any help with some overall topologies let me know. I can post some schematics. Without the drivers themselves I couldn't give you exact values for the parts, but it would get you started. If you do have measurements, and just need help with the crossover, then you can send me the files and I can see what I come up with.
 
Yes my thinking evolved from a 2way with mid-woofer, but it’s a 3 way with the “bottom” active.
So a few 3 ways use the bass driver with a shelf filter for BSC?

pjpoes
Possibly I might end up running those 12” Eminences to 500 Hz, they should handle that easily enough; it’ll be by trial. .

Can anyone throw a rough cost on reasonable inductors at say 250/ 300/ 400 Hz??

Thanks
 
otto88 said:

Sreten
A tad off topic, but in the Watt / Puppy, what did
you mean by *proper high pass filtering of the mid?

Hi,

The Watt has no BSC. Adding the Puppy compensates this lack of BSC.
But this is not simply adding stereo subs to a 2-way, the Puppy also
includes a 2nd order high pass filter for the Watts ro reduce low bass
excursion. It is effectively a 3-way with overlapping filter functions.

:)/sreten.
 
Trial is a good idea, I've been told that 500hz is a bad place to have a crossover. I haven't tried it myself yet, but intend to do some experimenting and listening tests when I start work on my bigger projects. I've mostly played with 2-ways and 2.5 ways and not had a reason to run a crossover at 500hz. It was recommended that I try for 300 or 350hz as a better location.

How wide is the baffle these will be mounted to, that will have an impact on crossover point choice. As will the location of the drivers in reference to each other. however, I don't believe it could be possible that the bottom 18 would be so far from the midbass as to be a problem. If you chose a crossover point of 500hz, the center of the bottom woofer should not be more than 27 inches from the center of your mid bass/midrange driver handling above 500hz. However if you go to 350hz, that would allow it to be as far as 37-38 inches away, which might be more reasonable for your situation. Given that I would expect the baffle to be at least 20 inches, and probably more than that, you won't need baffle step compensation until the 250hz or below range.
 
otto88 said:
On further thought, I think I’ll go 3.5 way, with the two way passive part being http://www.geocities.com/woove99/Spkrbldg/8945A_versions.htm

The 0.5 of the “sub” doing the BSC, the bass below 100 Hz removed from the mid woofers, and removing the two way’s BSC gaining sensitivity.

Who can tell me how to remove the BSC from the crossover circuit shown on that page?

Thank you

What you describe is not a 3.5 way design. To do this, you will need to high pass the Usher. This will be a completely different design. To achieve what you're thinking of now, you need accurate modeling.
 
Well he says he wants to remove the bass from the Usher drivers at 100hz. I would actually do so at around 300hz myself. Again, this is going to depend heavily on the width of the baffle you end up using, but if you place the crossover at the baffle step point, then the lower woofers will naturally (assuming level is placed higher) account for the baffle step loss, and it won't be an issue.

If you were to build the mentioned speaker design and highpass it at 100hz, you don't want to remove the baffle step. Assuming the drivers are mounted on the same specified width baffle, it will roll off around 350hz or so, as that would be the baffle step point, and you would still need this compensated for.

You use a 3.5 way when you want to cross over the midrange higher than the baffle step point. If you were using the drivers like we had discussed earlier and crossed them at 300-500hz, and the baffle width was say 20", then you would need to have baffle step compensation at ~200hz, so one fix is to have a .5 woofer that offers this increase below that point. A second approach would be to have a larger (oversized) inductor in the first position of the lowpass crossover, which creates the needed slope change at 200hz. I think it would have to be something like a 5mh inductor or higher, assuming 8 ohms (thats a guess off the top of my head, I would have to model it or calculate it to know if thats even right).

Your new thought would be better as a 3 way system, imo. Also, i think it would need a pretty good reworking of the crossover to ensure a flat response. I have measurements for the drivers used in that, or at least the woofers on a similar width baffle, so part wouldn't be a problem, but without having measurements of your woofers, I couldn't help you much.
 
Jay,

I may misunderstand some terms, but as I see it:
Your design is a 2 way.
If an active single bass driver per side is crossed in at say 100 Hz, it becomes a 3 way.
If a second bass driver per side is added with low pass at the BS, I referred to that as a 3.5 way.

Maybe that's not 3.5 way, but that was my intention.
(Is a 3.5 way where the second bass driver per side is allowed to run full range?)

By accurate modelling, you are saying that with a different (larger) baffle, the modelling of the baffle's loss would need to be accurate?

pjpoes
The reasons I have to remove the bass from the Usher drivers around 100 Hz (maybe a max of 150 Hz) are:
- To 90% of the vocals in the passive two way, with only one crossover.
- Reinforced by the aim that the passive two way will have a tube amp, while the bass will have a SS amp. Its meant to be preferable to cover 100 Hz – 10 kHz with one amp (or actives of the same type).

(fyi I was somewhat torn between Joe Rasmussen's Elsinore www.customanalogue.com/elsinore/elsinore_index.htm and "adding to" Jay's design with bass drivers.
But I have some Ushers, the Elsinore is just a bit too tall, it uses 4 midwoofers a side and I think 2 midwoofers and two bass drivers a side is more sensible, if the amps can handle 4 ohms; and using Jay's two ways ensures at least one of the two crossovers should be very good!).

pjpoes
I liked your idea of placing the crossover at the baffle step zone.
The front will likely be 900 mm (36 inches) high (no higher due to SAF). At the bottom, 325 mm (13 inches) wide, to accommodate the bass drivers. At the top about 225 mm (9 inches) wide. So an average width of 275 mm, 11 inches.
From John Murphy's formula, that average width suggests the F3frequency is about 420 Hz. I think John's formula is indicative, and the BS be measured.
I don't know that I want to cross to the bass that high, though I’d consider it.

But . . now I think about it more, the speaker will be within 80 mm (3 inches) of the wall (SAF), closer to 2 Pi, so BS may be minimal.
In which case it would be back to Jay's 2ways, with active bass drivers added ( 100 - 150 Hz). .
So who can say how to remove the BSC from Jay's crossover circuit shown
http://www.geocities.com/woove99/Sp...5A_versions.htm

Thank you
 
otto88 said:

I may misunderstand some terms, but as I see it:
Your design is a 2 way.
If an active single bass driver per side is crossed in at say 100 Hz, it becomes a 3 way.
If a second bass driver per side is added with low pass at the BS, I referred to that as a 3.5 way.

Now, I see. This is a 2.5-way plus an active sub. A 3.5 way design would use two identical (or very similar) midbass drivers, one working as a .5 woofer.

Note that you can't add a .5 woofer to the MTM design unless you use another amp for that. The MTM's impedance is already below 4 ohms. You need a 3-way crossover to achieve what you want. BTW, even removing BSC from my MTM design is not a simple task, because BSC is not a simple add-on feature but incorporated into the design. You need to remodel it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.