Nope. They are different, in the practical universe. Unless by huge OB you mean 10 meters wide or more. If the rear radiation can get to you via reflection from the walls, then OB and IB are different and need different crossovers. I guess the bigger the baffle, the smaller the difference.
But yeah - agree, some adjustment on the tweeter may be all you need. The open back of the mid will contribute to the tonal balance, the closed back of the tweeter will not. You may have to run the tweeter a few dB hotter.
Martin, thanks for the link to your project with the various iterations of “Full Range” Drivers. There was added support and insight for the concept that the overall speaker response curve is the interaction of the component parts. The amplifier, woofer, Open Baffle and the low pass filter typical do not have a textbook flat response curve. Same for the FR driver with its amplifier and high pass filter. A surprise to me, the LP and HP filters sum better if they do not share the same pole frequency; not your typical crossover. The non-zero output impedance of a tube amplifier may cause havoc in an already complicated model.
My empirical approach has been to us a Beringer DCX 2496 digital crossover and Rane Ma-6 6 channel amplifier. Next up I will try using different frequencies for the LP & HP filters and less equalization.
DT
My empirical approach has been to us a Beringer DCX 2496 digital crossover and Rane Ma-6 6 channel amplifier. Next up I will try using different frequencies for the LP & HP filters and less equalization.
DT
Yep! This is known as overlapped and underlapped crossovers. Very often needed for a flat response with typical drivers. Heavily underlapped low pass filters are often needed for OB woofers. Most builders don't understand this part.
Is that because of the dipole peak?
Why not just EQ out the peak and then get a normal response to work with?
Hi,
The dipole bump is one reason, 500Hz electrical for 400Hz acoustic,
the other is to utilise the extra bass sensitivity in combination with
the baffle loss to compensate for baffle loss, 200Hz electrical for
400Hz acoustic, this gives around 6dB compensation of baffle loss.
Further bass compensation is via the high Q of the bass drivers.
Like I said, understanding the offset c/o is critical to the schema.
rgds, sreten.
Last edited:
Right, the dipole peak plays a role but the main thing you are fighting is the rising response of the woofer on an open baffle. Because the bass wraps around the baffle and cancels out, you are left with a rising response. You are going to have to trade away that rising response in the crossover to get the woofer near flat. That can mean as much as 10dB correction to get the top and bottom to match. Many OB designers fail to understand this and end up with bassless, gutless wonders.
As sreten notes, and MJK has pointed out for years, a high Qts woofer can help flatten the response because of its bass bump. There are trade-offs, but that's one way to a better tonal balance.
As sreten notes, and MJK has pointed out for years, a high Qts woofer can help flatten the response because of its bass bump. There are trade-offs, but that's one way to a better tonal balance.
Hi,
i can't disagree with the above, but its the combination of an offset c/o
and a high Qts driver that does the job, a high Qts driver e.g. the GRS
unit only gives about 4dB, the c/o and driver sensitivity offset 6dB.
That gives you a 10dB bass boost baseline, the GW a further 3dB @ 30Hz.
FWIW nearly all FR electrostatics purposefully have a bass Q of 2 to 3.
The classic quad ESL57 is a case in point, very limited bass power
handling due to the limited excursion and high Q bass alignment.
This has been known for a very long time.
It is all about understanding whats going on and your choices.
rgds, sreten.
i can't disagree with the above, but its the combination of an offset c/o
and a high Qts driver that does the job, a high Qts driver e.g. the GRS
unit only gives about 4dB, the c/o and driver sensitivity offset 6dB.
That gives you a 10dB bass boost baseline, the GW a further 3dB @ 30Hz.
FWIW nearly all FR electrostatics purposefully have a bass Q of 2 to 3.
The classic quad ESL57 is a case in point, very limited bass power
handling due to the limited excursion and high Q bass alignment.
This has been known for a very long time.
It is all about understanding whats going on and your choices.
rgds, sreten.
Last edited:
One other variable creeps into the crossover component sizing. The woofer(s) have a huge impedance peak that will interact with the crossover filter so that the crossover will not roll off the response as low as if it was working against a pure resistance, you gain some dBs in the low bass range. Notice that there is no Zobel or trap across the woofer(s). The filter components should be iterated (I do it by simulation but it could be done in the actual build if enough big inductors and caps are available) to get the combined SPL frequency response desired to mesh with the next driver up in the OB system.
Martin
Martin
No speaker in a real box or on an OB has a flat response curve. If your software is calculating a flat clean SPL response over a wide frequency range then it does not include enough inputs and variables to yield an accurate result.
Martin
Hi,
Another reason I prefer the GRS drivers over the GW drivers for passive,
but I'm not arguing too much, I just think the GRS's are a better bet.
As stated in the original post, ideally you would at least model/sim it.
The bass to mid c/o is by far IMO the real critical issue and not easy.
The passive line level bass/mid EQ does help in flexibility, but like I
said in the original post, I'm not claiming what I suggest is optimum.
Its far better than a half baked plan, but not necessarily very toasty.
rgds, sreten.
The Jamo's seem to only use a 1st order high pass on the mid ....
With the inevitable peaking at driver Fs that inevitably entails ....
YMMV but what I'm suggesting in this respect is actually better ....
Last edited:
Martin, to name a few variables, baffle size, woofer resonance frequency and total Q. These variables contribute to the nature of the low end bass performance of the woofer. Typically for a IB box and sometimes for an OB people think of a Q:= 0.71 as a design goal, some where between over and under damped. As the Q of the woofer increases so does the impedance of the woofer at the resonance peak. Oddly a current source amplifier will increase power output at woofer resonance where a voltage source amplifier will output less power at the woofer resonance. P= (I^2) * R current source, P=(E^2)/R voltage source. Perhaps this explains some of the preference for SS amplifiers over Tubes for OB speakers.
DT
DT
I think either driver will work. At the frequencies being applied to the drivers they will both behave as pistons without any cone resonances so build quality is probably not an issue. The only difference might be an extra dB or two at the very low end created by the higher Qts of the GW. At the very bottom end, the room and placement will determine if this can be considered a real advantage or not. Out of phase reflections from the rear wall might suppress the low bass from either driver.
Martin
The floor reflection is another huge variable. It will reinforce the deep bass but cause ripple as the floor bounce combines with the direct sound from a driver at higher frequencies. Including the floor is not commonly found in a lot of the speaker enclosure design software.
Martin
>>> No speaker in a real box or on an OB has a flat response curve. If your software is calculating a flat clean SPL response over a wide frequency range then it does not include enough inputs and variables to yield an accurate result.
So true. I remember when Stereophile began producing 'in room' frequency charts to go along with their standard anechoic charts. They always look very different. Sometimes expensive speakers with flat responses no longer had flat responses in a real room.
While i realize the importance of a balanced frequency response it raises the question of what other factors contribute to one speaker sounding better than another. Simply placing superior (and expensive) components into a box seemed a decent starting point in producing a successful project, but as DIYers we realized quickly this is not always the case.
It's incredible to me how we are discussing some of the cheapest 15" drivers around (GRS and Goldwood) and using them in what would be considered an excellent sounding pair of speakers. These drivers certainly are purpose built to play low frequencies. Devising a plan to work with and maximize their strengths is incredibly seductive to the DIY community where money is not always no object.
So true. I remember when Stereophile began producing 'in room' frequency charts to go along with their standard anechoic charts. They always look very different. Sometimes expensive speakers with flat responses no longer had flat responses in a real room.
While i realize the importance of a balanced frequency response it raises the question of what other factors contribute to one speaker sounding better than another. Simply placing superior (and expensive) components into a box seemed a decent starting point in producing a successful project, but as DIYers we realized quickly this is not always the case.
It's incredible to me how we are discussing some of the cheapest 15" drivers around (GRS and Goldwood) and using them in what would be considered an excellent sounding pair of speakers. These drivers certainly are purpose built to play low frequencies. Devising a plan to work with and maximize their strengths is incredibly seductive to the DIY community where money is not always no object.
Actually in these types of OB systems, high Qts 15" drivers crossed at 200 Hz, most of the ripple due to floor bounce is generated from the mid range or full range driver. The two bass drivers do not exhibit much floor bounce even when stacked vertically.
Martin
You should do the math (find the 1/2 wl cancellation point), but typically it's between 250-350 Hz based on the geometry of the average listening position relative to the speakers. Oftentimes this frequency is part of the midrange rather than bass in a 3-way.
This neverending interest in 'pisswarm' beer, the cheeper the better !
Luckily with young ears Adolf Corkscrew have acted upon what he heard and subsituted Eminance Alpha15s for Beta15s, so also Nelson Pass did.
I myself changed for Acoustic Elegance IB15s instead of the Alphas just for the joy of hearing pure music and vocals. The GRS with it's spec won't be better than the Alphas.
Why not chase for quality in music insted of hunting $$$ and muddy tones.
Nothing said about Zaph's speakers.
/Erling
Luckily with young ears Adolf Corkscrew have acted upon what he heard and subsituted Eminance Alpha15s for Beta15s, so also Nelson Pass did.
I myself changed for Acoustic Elegance IB15s instead of the Alphas just for the joy of hearing pure music and vocals. The GRS with it's spec won't be better than the Alphas.
Why not chase for quality in music insted of hunting $$$ and muddy tones.
Nothing said about Zaph's speakers.
/Erling
Is the 200hz crossover point a figure of the design alignment or could a 15" driver that performs well up to say 600hz be used accordingly? There's the B&G Neo10 that IMO would be an outstanding driver to try out in this scheme but i wouldn't want to press it any lower than 500hz.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.