I understand this bit, that the first-order low-pass will give me BSC automatically. My question was: does this first-order low-pass filter have to be passive? Can't I make it active, by using one opamp, one R and one C, in a separate stage before or after the xo? Please note that when I say "active low-pass" I don't mean shelving filter... I mean a low-pass, as required for the 0.5-way config.sreten said:
Anyway, forget it. I think I've pretty much understood what you have been pointing out.
tcpip said:I understand this bit, that the first-order low-pass will give me BSC automatically. My question was: does this first-order low-pass filter have to be passive? Can't I make it active, by using one opamp, one R and one C, in a separate stage before or after the xo? Please note that when I say "active low-pass" I don't mean shelving filter... I mean a low-pass, as required for the 0.5-way config.
Anyway, forget it. I think I've pretty much understood what you have been pointing out.
To make it active you need another amplifier for those two drivers.
Hmmm.....
sorry I'd forgotten your using one amp per driver, I was
thinking of a standard 2 way 2 amplifier active speaker.
Of course with an amplifier per driver you should implement
this electronically. An op-amp is not required, a simple RC
network at the amplifier input will suffice for the 0.5 way
drivers amplifiers.
Sorry for misunderstanding your point.
sreten.
Well, I was beginning to think that we had taken a simple idea and made it much too complex, but thinking it over, the design really won't be much if any more complex than what you originally proposed.
You were always planning on low pass filters of some sort on the woofers. Initially to cross over to the tweeters, now some of the woofers will cross over lower instead,
OK, where are we?
Good Ideas:
A 2.5 way system, i.e. some of the woofers are rolled off low
It could be 2 woofers or it could be 3, right?
Obiviously if it is an MTM above, and a total of 4 woofers, than yes, it is two woofers rolled off low. If it is TM then 3 of the woofers could be rolled off low.
The remaining questions are:
Is there some reason that passive filters are preferable for the BSC?
Yes, you kill two birds with a 2.5 BSC filter, BUT if tcpip wants to
have the single woof/mid cover lower than the BS freq. then he
might have complications- We need to know the baffle step frequency of the cabinet to proceed. What is it, about a foot wide cabitnet? we have to calc that Baffle step!!
Gotta work for the moment though!!
You were always planning on low pass filters of some sort on the woofers. Initially to cross over to the tweeters, now some of the woofers will cross over lower instead,
OK, where are we?
Good Ideas:
A 2.5 way system, i.e. some of the woofers are rolled off low
It could be 2 woofers or it could be 3, right?
Obiviously if it is an MTM above, and a total of 4 woofers, than yes, it is two woofers rolled off low. If it is TM then 3 of the woofers could be rolled off low.
The remaining questions are:
Is there some reason that passive filters are preferable for the BSC?
Yes, you kill two birds with a 2.5 BSC filter, BUT if tcpip wants to
have the single woof/mid cover lower than the BS freq. then he
might have complications- We need to know the baffle step frequency of the cabinet to proceed. What is it, about a foot wide cabitnet? we have to calc that Baffle step!!
Gotta work for the moment though!!
Sorry Sretin, I was writing my previous as you posted, so I missed your response. That pretty much clears things up.
The one amp/one driver approach really makes a lot of thinks easier once you get all thoose amps and crossovers made
tcpip:
My calcs for the 200hz crossover point are 4560/220= about a 20" wide width. That's a lot wider than you mentioned originally(well, you called them towers I think )
You wanted the 200hz crossover to keep the crossover out of the midrange, but the crossover is going to be about as benign
as any crossover can be, seeing as it is between two identical drivers! Speakers are always a compromise, so I'm thinking that
a 12" wide baffle (380hz) would be fine.
So the next question is what the ideal slope of the crossover between the woofers and the (identical) midrange. Of course a first order slope would be appealing because of the phase advantages, and the entire midrange would be a smooth blend of the two drivers, Could someone tell me what the ideal slope is for baffle step? Another, probably more important issue is that if both bass drivers and mid driver are active over much of the midrange, do we go back to having lobing, etc? problems? I suspect that we would, so probably a higher than first order is in order......
Eager to hear opinions on this...
Mark
The one amp/one driver approach really makes a lot of thinks easier once you get all thoose amps and crossovers made
tcpip:
My calcs for the 200hz crossover point are 4560/220= about a 20" wide width. That's a lot wider than you mentioned originally(well, you called them towers I think
)You wanted the 200hz crossover to keep the crossover out of the midrange, but the crossover is going to be about as benign
as any crossover can be, seeing as it is between two identical drivers! Speakers are always a compromise, so I'm thinking that
a 12" wide baffle (380hz) would be fine.
So the next question is what the ideal slope of the crossover between the woofers and the (identical) midrange. Of course a first order slope would be appealing because of the phase advantages, and the entire midrange would be a smooth blend of the two drivers, Could someone tell me what the ideal slope is for baffle step? Another, probably more important issue is that if both bass drivers and mid driver are active over much of the midrange, do we go back to having lobing, etc? problems? I suspect that we would, so probably a higher than first order is in order......
Eager to hear opinions on this...
Mark
I'm really glad we have converged on this one... I no longer feel like a fool.sreten said:
Now please bear with another freshie question. This is about a passive line-level first-order low-pass. I was asking Gooroo Angshoo about exactly this a couple of days ago.
If I have to build this 6db/oct as passive low-pass, I'll have an R in the signal path, followed by a C which connects the signal to ground. And in my case, this will go between preamp and power amp. In this case, the R will be the Zout of the preamp, i.e. something like 100Ohms. In order to get a C for this, I'll need a very large cap. If I am on a budget, such large caps will only be electrolytic. Is this okay, or will I need to go in for an expensive MKT/MKP high-value cap? It was for this reason that I was thinking of active line-level low-pass instead of passive. If I go active, I can put an additional R and then have a proportionately smaller C, and then put an opamp as unity-gain buffer so that the power amp continues to see a Zsource which is low-impedance.
Any help on this?
He spells it "sreten", Variac. Get it right, or he'll send those Dementors after you.Variac said:
Yes, I think 200Hz will be a bit too much. So I think I'll just stick to putting the low-pass at the BSC frequency, and just live with the fact that this freq is in the middle of the Oh-So-Special 300-3K Magic Range.My calcs for the 200hz crossover point are 4560/220= about a 20" wide width. That's a lot wider than you mentioned originally(well, you called them towers I think
I think we should use something like a first-order or even less (e.g. 3dB/oct) for this slope. Going by whatever I've seen of in-room response plots, the slope due to baffle step is very shallow. I don't think we should go above first-order at all. And I think the interference with the tweeter is the lesser of the evils in this case. At least that's how I intend to try it and then see how it sounds.
Zounds! Zat's ze deziding vaktor.
tcpip said:
Rout of the preamp is not relevant (as long as its low) Rin of the
power amplifier is more important as it affects the frequency of
the filter slightly.
You add a series resistor of say 10Kohm and then shunt to ground.
You should do this at the amplifier input, 10K is not good for driving cables.
sreten.
Member
Joined 2003
My previous system was a 2.5 with line level BSC. Reading the circuit description will explain how...really just a .5 series xo on the drivers rather than .5 parallel xo.
2.5 with line-level BSC
2.5 with line-level BSC
Ha!! I just spelled it wrong to wake him up!- He obviously wasn't
paying attention
OK so if you cross over at 380 hz and the bass roll off slope is 3dB/octave then the woofers are only down about 8db at the tweeter crossover- seems like you would then need to have a tweeter crossover on them too!!
Let's see, with a 6db slope it would be maybe 16 db down. That sounds like it might be low enough not to be a problem.
Nowwith the mid driver, a 3db slope doesn't even keep up with the increased excursion as the notes get lower, but I gues it can handle bass so maybe is OK.
Clearly I prefer the 6db slope, but equally clearly, I don't know anything about the subject.
Does this make sense?
MArk
Don't get me wrong... I too will use a 6dB/oct slope, and largely because the circuit is simpler than a half-order filter, and because of the other reasons you pointed out. But I'd mentioned 3dB/oct earlier just to say that when I see existing graphs of speakers who have not corrected their BSC, the slope is very shallow, maybe even shallower than 6dB/oct, thus making it a bit unlikely that a higher-order low-pass would act as a good compensation. That's all.Variac said:Clearly I prefer the 6db slope, but equally clearly, I don't know anything about the subject.
Great stuff, Paul, thanks. Now that that pair is a year old, how does it sound? More specifically, did you have audible lobing or comb filtering problems? How far away was the listening position from the speakers?Paul W said:
One point which puzzled me was that the text referred to MMTMM, while the photos seemed to show MMTMMMM columns. (Or were they ports??? Oops!)
Now that you've walked this path of multi-driver midbass, would you prefer a smaller number of drivers for the best sound? What are your overall impressions about this entire approach?
just note the active or passive 1st order low pass filter only has a
6dB/octave slope in the roll-off region well above the c/o frequency.
In the transition region of the filter around
the c/o frequency the slopes are much lower.
For a 0.5 way low pass filter discussion of excursion is not relevant.
sreten.
6dB/octave slope in the roll-off region well above the c/o frequency.
In the transition region of the filter around
the c/o frequency the slopes are much lower.
For a 0.5 way low pass filter discussion of excursion is not relevant.
sreten.
tdpip is is going active, using inexpensive drivers and wants reasonably high SPLs:
thinking aloud,
if he goes for one of the .5 options available to him, the mid will suffer from IM from the bass signal, but @ -3dB due to the BS.
whereas if he goes 3-way, the mid will see no bass below the BS/300Hz/whatever, therefor no IM, but at the expense of an extra active way on his X-over.
i'm just wondering which is more favourable in this sort of situation.....
thinking aloud,
if he goes for one of the .5 options available to him, the mid will suffer from IM from the bass signal, but @ -3dB due to the BS.
whereas if he goes 3-way, the mid will see no bass below the BS/300Hz/whatever, therefor no IM, but at the expense of an extra active way on his X-over.
i'm just wondering which is more favourable in this sort of situation.....
Member
Joined 2003
tcpip,
My listening room is very live, listening distance is about 12 feet, and the bottom two holes are flared ports (which I sealed 99% of the time). I used the MMTMM config to increase vertical directivity & reduce floor and ceiling reflections. It helped, but not as much as the 4-way dipole replacements (see link). Imaging was super.
The .5 configuration worked fine...I had larger issues with the 2-way part where I had to push the tweeter lower than it really wanted to go, and the mids higher than they really wanted to go. (I'm very picky about that.) Played loud, they both strained a little even with a 4th order xo. I think the .5 approach is fine...having "been there, done that" I would go straight to a 3-way with more specialized woofers & mids...but cabinet and everything else gets more complicated. Darned trade-offs!
W-LM-UM-T-UM-LM-W
My listening room is very live, listening distance is about 12 feet, and the bottom two holes are flared ports (which I sealed 99% of the time). I used the MMTMM config to increase vertical directivity & reduce floor and ceiling reflections. It helped, but not as much as the 4-way dipole replacements (see link). Imaging was super.
The .5 configuration worked fine...I had larger issues with the 2-way part where I had to push the tweeter lower than it really wanted to go, and the mids higher than they really wanted to go. (I'm very picky about that.) Played loud, they both strained a little even with a 4th order xo. I think the .5 approach is fine...having "been there, done that" I would go straight to a 3-way with more specialized woofers & mids...but cabinet and everything else gets more complicated. Darned trade-offs!
W-LM-UM-T-UM-LM-W
I guess there's a third choice: MTMBB. Right? And BTW, if I go MTM as against TMM, do I need to change anything in the xo? A friend mentioned something to this effect off-line, but didn't elaborate. Does an MTM config need any special xo feature? I was thinking of going plain LR4 electrical.sreten said:
Dear Paul,
I'm still studying your test mules page. Thanks. And BTW, you seem to have a lovely living room.
Thanks very much. This helped.Paul W said:
Me fresh to this game, but I'd have thought that you'd have no problems integrating the M's and the T, because of the super-low Fs of the 9900 Revelors. Isn't it something like 500Hz?
I'm still studying your test mules page. Thanks. And BTW, you seem to have a lovely living room.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.