[fdandrews]

I have read the article on Martin King's website which compares open baffle, U-frame and H-frame speakers. He talks about setting the depth of the U or H frame to push the first quarter wavelength resonance above the crossover frequency. Unfortunately for me, I do not speak math. My current bass speakers are MCM 21" woofers in U-frames that are 24"x24"x16" deep. These are mated with Fostex 166en speakers, which in their own 9" deep U-frames roll off at about 200hz at 12db/oct without any additional crossover. To get a decent combined FR, I am rolling the bass speakers off at 60hz, 12db/oct.  Based upon what I have read, I suspect that the bass U-frames are too deep. Can anyone explain to me how to calculate the correct depth for the U-frames? 

[JohnH12]

You can use ABC Dipole (in excel) to calculate the response from your U frame.  The help file will give you some decent background.

http://www.musicanddesign.com/A_B_C_Dipole.html

John H

[MJK]

I have read the article on Martin King's website which compares open baffle, U-frame and H-frame speakers. He talks about setting the depth of the U or H frame to push the first quarter wavelength resonance above the crossover frequency. Unfortunately for me, I do not speak math. My current bass speakers are MCM 21" woofers in U-frames that are 24"x24"x16" deep. These are mated with Fostex 166en speakers, which in their own 9" deep U-frames roll off at about 200hz at 12db/oct without any additional crossover. To get a decent combined FR, I am rolling the bass speakers off at 60hz, 12db/oct.  Based upon what I have read, I suspect that the bass U-frames are too deep. Can anyone explain to me how to calculate the correct depth for the U-frames?

For a 24 inch x 24 inch cross-section U or H-Frame I would try between 5.5 and 6.5 inches deep per side. Assuming a 3/4 inch thick baffle this gives a total depth of between 11.75 and 13.75 inches for a H frame and 6.25 or 7.25 for an U frame. I don't know anything about your driver so can't comment on how well this might work. The method for calculating the depth is shown on page 3 of the following link :

http://www.quarter-wave.com/OBs/U_and_H_Frames.pdf

r_effective is the radius of the circle with the same area as your square cross-sectional area of the H or U frame (in your case 576 in^2), r_effective is an end correction.

Why would you put the Fostex FE166en in anything but a plain open baffle? A U frame for the full range driver makes no sense.

Acoustic (not electrical) crossover should be around 200 to 300 Hz. You need to have a filter applied to both drivers to get the best result.

[fdandrews]

Thank you for the reply. The boxes were originally built before I understood enough about open baffles, sort of as a test to see if I liked how they (open baffles) sounded in my room. I realize that I made a mistake in putting the Fostex in a U-frame. At the time, I thought the U-frame might compensate somewhat for the low Qts driver bass rolloff. I will be correcting that at the same time that I fix the bass driver U-frame.

[OZZIOZZI]

questions for Martin if he is watching this thread. More a reality check than question. And I am maths challenged like fdandrews.

I have read MJK article about H and U frames and *think* I inderstand the differences between the two. I would like to use one or two 12" bass drivers per side to support a fullrange driver in the slimmest OB I can build for Aesthetic reasons. Problem I see with H frame is, if I place the FR driver above the woofer in vertical alignment (preferable?)  I don't think H frame with FR baffle above looks as good as a woofer placed in a U frame so that both drivers are in vertical alignment on the front baffle. For experimental purposes either H or U is fine, but for a more "finished" look I prefer the U frame. That way it looks like a panel speaker from the front and I can cover both drivers with a common grille cloth. And a cover to cut dust and cobwebs on the back--a problem in my location. What depth can I get away with if my U frame is 13"x13" or 14" x 14" inside dimensions and I XO at  below 200Hz.

If I place one woofer above the other on each side do I need to place a divider between the two and why?

Any advice appreciated. Below is the style of speaker I would like to build, but narrower and deeper. I also prefer birch ply with clear finish for sides so it sits "lighter" in the room. BTW I listen at moderate levels, mainly to acoustic jazz, blues, folk and country and earthquake bass is not necessary.
Ozziozzi

[MJK]

For a single U frame with an internal cross-section of 13" x 13" and a 2nd order crossover at 200 Hz, I would use a length of 9 inches.

For a pair of U frames with an effective internal cross-section of 13" x 26" and a 2nd order crossover at 200 Hz, I would use a length of 7 inches. I would also use a dividing panel between the drivers to act as a structural reinforcement of the long dimension of the cross-section and to more effectively "load" each driver in its own 13" x 13" channel.

[OZZIOZZI]

Martin, thank you for your advice. Now I can start drawings and construction.

Regards

Ian

[OZZIOZZI]

Martin or anyone else who cares to chime in. I have read other posts talking about modifying the Qtd(s) of a driver with series resistance and how damping factor of amp affects drivers etc. I'm not sure if I understand the discussion completely, but I got the idea that series resistance does have *some* effect on the driver's behavior. If I use two bass drivers and connect them either in series or parallel in one U frame above the other how does the bass extension or SPL vary from the pair. In each case. I have the option of passive XO or active plus biamping and 4ohm load is no problem for my SS amp(s). I realise I will have to change passive XO to account for 4ohm or 16 ohm load at the same frequency, but which connection of drivers would be preferable and why?

Ozziozzi
So many options!!!!

[MJK]

I have read other posts talking about modifying the Qtd(s) of a driver with series resistance and how damping factor of amp affects drivers etc. I'm not sure if I understand the discussion completely, but I got the idea that series resistance does have *some* effect on the driver's behavior.

Adding series resistance will increase the effective Qts of the speaker, it won't change the basic driver's T/S parameters (notice speaker refers to a system and driver is the woofer itself). The SPL response will be impacted and if looked at as a system, resistor and driver, it will appear to be a higher Qts speaker. It accomplishes this by reducing the SPL/2.83 V/m value, the resistor forms a voltage divider with the driver so the driver will only see a portion of the input signal. A resistor will increase the effective Qts but bring down the specified SPL/2.83 volt/m so there is no free lunch.

If I use two bass drivers and connect them either in series or parallel in one U frame above the other how does the bass extension or SPL vary from the pair. In each case. I have the option of passive XO or active plus biamping and 4ohm load is no problem for my SS amp(s). I realise I will have to change passive XO to account for 4ohm or 16 ohm load at the same frequency, but which connection of drivers would be preferable and why?

The bass extension will not change with two bass drivers. The shape of the SPL frequency response will be the same only the magnitude will change.

For two bass drivers wired in parallel the cone area doubles, the impedance is halved which doubles the current for the same applied 2.83 volts (1 watt into 8 ohms), the current to each driver is the same (double current is split for the parallel connection) as the single driver so the cone displacements are the same, with the end result being an increase in the combined SPL of 6 dB.

For two bass drivers wired in series the cone area doubles, the impedance doubles which halves the current  for the same 2.83 volts, for half the current each driver has half the displacement of a single driver, the end result being that the combined SPL is unchanged (but the driver displacement is reduced by a factor of two).

See the following document for additional information/explanation and a sample problem.

http://www.quarter-wave.com/General/Two_Drivers.pdf

If your single woofer has a high enough SPL/2.83 volt/m in a dipole set-up to work with the full range or mid range driver value then using two in series will buy you more maximum output if your woofer is displacement limited. If your single woofer's SPL/2.83 volt/m in a dipole set-up is below the full range or mid range driver then using two in parallel will increase the output by 6 dB and may bring you a better SPL match. For dipole speakers you really want the specified SPL/2.83 volt/m value of your woofer(s) to be 6 to 10 dB greater than the full range or mid range driver value to make up for the dipole roll-off below about 100-200 Hz.

Simulation ahead of time is your friend for understanding and designing a good dipole speaker system.

[OZZIOZZI]

Martin said
"Adding series resistance will increase the effective Qts of the speaker, it won't change the basic driver's T/S parameters (notice speaker refers to a system and driver is the woofer itself). The SPL response will be impacted and if looked at as a system, resistor and driver, it will appear to be a higher Qts speaker. It accomplishes this by reducing the SPL/2.83 V/m value, the resistor forms a voltage divider with the driver so the driver will only see a portion of the input signal. A resistor will increase the effective Qts but bring down the specified SPL/2.83 volt/m so there is no free lunch."

I realise that the Qts of each individual driver does not change since it is the combination of electrical and mechanical properties of that individual driver (correct?). I suppose my question should have been phrased  " can the Rvc resistance of each of the two drivers in series be considered as adding external resistance to the other driver and *that* raises the effective Qts of the *system* of two drivers in an open baffle".

Not many free lunches unless you go to supermarket product launches.

Re: introduction to your article about parallel vs series connection of drivers:
I was once told by my supervisor before I had to present a seminar, "only tell them half of what you know--that way you will have something to say during question time".

Ozziozzi

[G Georgopoulos]

Martin said
"Adding series resistance will increase the effective Qts of the speaker, it won't change the basic driver's T/S parameters (notice speaker refers to a system and driver is the woofer itself). The SPL response will be impacted and if looked at as a system, resistor and driver, it will appear to be a higher Qts speaker. It accomplishes this by reducing the SPL/2.83 V/m value, the resistor forms a voltage divider with the driver so the driver will only see a portion of the input signal. A resistor will increase the effective Qts but bring down the specified SPL/2.83 volt/m so there is no free lunch."

I realise that the Qts of each individual driver does not change since it is the combination of electrical and mechanical properties of that individual driver (correct?). I suppose my question should have been phrased  " can the Rvc resistance of each of the two drivers in series be considered as adding external resistance to the other driver and *that* raises the effective Qts of the *system* of two drivers in an open baffle".

Not many free lunches unless you go to supermarket product launches.

Re: introduction to your article about parallel vs series connection of drivers:
I was once told by my supervisor before I had to present a seminar, "only tell them half of what you know--that way you will have something to say during question time".

Ozziozzi

Hi Ozziozzi

Driver t/s parameters dont apply to ob,mho 
As for a resistor increasing Qts, are you serious??

[MJK]

I realise that the Qts of each individual driver does not change since it is the combination of electrical and mechanical properties of that individual driver (correct?). I suppose my question should have been phrased  " can the Rvc resistance of each of the two drivers in series be considered as adding external resistance to the other driver and *that* raises the effective Qts of the *system* of two drivers in an open baffle".

Wiring two drivers in series does create a voltage divider, half the voltage is seen by each drive over the entire frequency range. So the output is halved at all frequencies. The shape of the SPL frequency response will be unchanged other than it will be at a lower level over the entire frequency range.

Putting a resistor in series will produce a voltage divider with the speaker. At low frequencies around fs the impedance of the driver is so high that the series resistor has no significant impact on the SPL curve. Once you get above fs where the impedance curve settles and is close to the voice coil resistance then the voltage division starts and the SPL curve is attenuated. Just looking at the modified SPL curve it looks like a higher Qts driver but with a reduced output compared to the original driver's SPL curve.

[G Georgopoulos]

since vc is reactive two series connected speakers Qts is doubled at the terminals of the total connection, at the individual speaker terminals is as per speaker, with two parallel speakers the total Qts is always the individual speaker Qts.

the value of Qts is proportional to the energy stored, divided by the energy dissipated, and is defined at resonance (Fs).

like as i said driver t/s parameters like Qts are designed for use in box speakers...mho

[OZZIOZZI]

....Just looking at the modified SPL curve it looks like a higher Qts driver but with a reduced output compared to the original driver's SPL curve.

Disregarding the output levels, which can be easily adjusted, do two drivers in series "look" like one equivalent driver with higher Qts than either individual driver? I don't mean to be picky, just trying to get my head around this concept, based on,"if it walks like a duck and quacks like a duck..." This is all based on the general assumption on many fora that high Qts drivers are almost essential for OB applications unless you take the approach of eq boost and lots of analog or digital compensation and low Q drivers with lots of Xmax.

Re. GGs comment

As for a resistor increasing Qts. Are you serious?

Don't shoot the messenger.
I am just trying to understand whether it is better to connect bass drivers in series or parallel and WHY. High Qts woofers for OBs seems to be assumed as the best choice. It seems intuitively to me that putting two drivers in series must REDUCE the amplifier damping on both in the same way as putting a resistor in series with them or reducing the size of their magnets. I assume that lower damping would make the response curves of the two drivers "look" like one higher Qts driver. Obviously the drivers don't change, just the combination or equvalent driver system created by this arrangement.

 As Martin said, " it (the system) **looks** like a higher Qts driver" So is it an advantage to wire drivers in series for OB bass?

I have seen examples of both arrangements, but the designers seem more concerned with the load presented to the amp or avoiding larger inductors rather than improving the "behaviour" of the equivalent driver. Martin's response seems the closest I have found to someone saying that there might be an advantage due to an apparent raising of Qts for the series combination. Or have I mis-understood?

Ozziozzi

[MJK]

From Table 1 in the article I referenced in the above post, the equivalent Qts does not change with parallel or series connection of the woofers.

[OZZIOZZI]

Martin, with the greatest respect, what do you mean by "looks" like a higher Qts driver? Rather than "acts like" or "behaves like" a higher Qts driver? Is this just a mathematical anomaly that has no basis in reality?

Ozziozzi
AKA Bamboozled from Brisbane

[MJK]

The shape of the SPL curve looks like what you would expect from a higher Qts driver compared to the original. Don't over think the series resistance behavior. Adding series resistance makes the woofer appear to be a higher Qts driver at the expense of SPL.

Run a simulation of your potential woofer and then add series resistance and observe the changes in the SPL frequency response.

[OZZIOZZI]

Martin, thanks so much for that explanation. Now it all makes sense.

Regards

Ozziozzi