Dipole M frames -_ a look at their benefits

[Rudolf]

M frames are dipole constructions half way between H frame and W frame. Usually they look like this:



First important reason to go for a M frame is force cancellation: Since the angle between both drivers is 90°, half of the in/out movement of the drivers is cancelled in the up-down direction, and only the other half is expressed in forward/backward movement of the frame.

The other reason is the different ratio of frame size to lambda/4 resonance. Let´s look at the response of the shown M frame, measured at the crosspoint of the red lines:



The same response measured with another software, microphone and room:



As red line I have included the most applicable simulation of an H frame with the U & H Frame Woofer Worksheet:

http://www.pvconsultants.com/audio/boxmodel/uhframe.htm

This corresponds to a H frame depth of 21 cm (one chamber). Don´t worry about the big difference between 25-250 Hz. Measurement includes floor boost and was taken for the two-driver-frame, while the simulation is in free space and accounts for 1 driver only (and 1/2 of the opening area of the M frame).

Now let´s look how H frames of 28 cm depth (30 cm - 2 cm for baffle) and 14 cm depth would behave:



Red line is for the 28 cm chamber, purple for 21 cm (corresponding to the measured M frame) and blue for a 14 cm chamber.
To me this suggests that a M frame of a given depth could retain some of the bass efficiency of an H frame twice its depth by sacrificing the higher resonance frequency, a H frame of the same depth would provide.

Looks like a valuable compromise to me.

[johninCR]

Rudolf,

Thanks for the info. I've got a set of M-frames + extra U extension that I need to tweak into submission. You have me scratching my head trying to figure out how deeper extension that a same depth H is possible. The M is still a dipole, so where is the extra propagation delay for the rear wave? Sure some of the output comes from deeper in the rear cavities, but the same applies in front making the average D still the depth of the cab. What am I missing or what is the simulator missing?

[MJK]

The M is still a dipole, so where is the extra propagation delay for the rear wave?

If the measurements are near field, centered on the red "X", I think that the output from one side of the enclosure would dominate and the dipole influence would be supressed. If the simulation was run in the same manner then this would also be the case. That is my guess to explain what is shown in the plots. If I am correct, the correlation looks pretty good between simulation and test.

[Rudolf]

Originally posted by MJK
If the measurements are near field, centered on the red "X", I think that the output from one side of the enclosure would dominate and the dipole influence would be supressed. If the simulation was run in the same manner then this would also be the case.

Yes, the simulation was done for the front plane of the H baffle. The only difference is that the simulation was done for a quadratic opening with one driver.

@JohninCR:
Everything is strictly nearfield of course - with no influence of the rear wave. I wasn´t interested in the dipole behaviour in the first place, but in the H frame equivalent of that M frame.

@MJK:
Since I do not have comparable measurement data for low frequencies: Is it valid to assume that a M frame with a given opening area and resonance will have the same low frequency response as a H frame with the same data? When the M frame apparently needs only 2/3 of the equivalent H frame depth, this would be a real achievement.
As far as I understand it this would be the same question as: Has a tapered TML with the same Smouth and Fres the same output as a straight one?

[johninCR]

Rudolf,

OK, I think I got it. Just to summarize the M-frame advantages over an H-frame to be sure:

1. 50% mechanical vibration.

2. Smaller size due to the fold.

and most importantly
3. The fundamental resonance for an M-frame occurs at a higher frequency than an H-frame of the same depth. This allows us to build an M approx 50% deeper for the resonance to occur at the same point as an H, so deeper bass is possible for the same XO limit.

Did I summarize correctly?

Now let me throw a possible monkey wrench in there, and maybe why I'm having problems with mine (the larger opening is to the rear). What about the resonances on the other side with 2 openings at 45deg angles? Are they higher or lower than a corresponding H? Also, if they don't correspond with the same opposite phase resonance in the front, do they become more problematic, since there's no dipole cancellation of the resonances?

[MJK]

Is it valid to assume that a M frame with a given opening area and resonance will have the same low frequency response as a H frame with the same data?

Rudolf,

To be honest, I have not spent much time looking at H, U, M, Ripole, or any of the other interesting dipole styles of boxes to be able to answer your question. These are just something that I have not had the time, or given enough priority, to look at in any detail.

[johninCR]

Originally posted by MJK


Rudolf,

To be honest, I have not spent much time looking at H, U, M, Ripole, or any of the other interesting dipole styles of boxes to be able to answer your question. These are just something that I have not had the time, or given enough priority, to look at in any detail.

Martin,
Are your spreadsheets accurate for such short pipes at very low frequencies, or is behavior different so close to the driver that it becomes significant in relation to the short length? If they're accurate in that zone, then we can easily model the simple versions of these cavity shapes using your existing sheets.

[badman]

If so, a modest M frame could provide dipole bass without the need for EQ- use a short frame and a quad of 8"s to get enough motor to get the efficiency up. And by going with such a short frame, you push the resonance up pretty high too making for less coloration...

Sounds like a win-win, I think I'll try it with my next OB.

[MJK]

John

The worksheet algorithm should be accurate enough for short pipe enclosures. The only easily modeled geometry at this time is the U frame. In addition, I am in the process of updating all of the worksheets to include a new baffle diffraction calculation and probably a polar response plot which should make the worksheets even better for the U frame simulation. But at this time, I do not believe I have a worksheet available that will accurate model the H, M, or Ripole geometry.

[johninCR]

Originally posted by MJK
John

The worksheet algorithm should be accurate enough for short pipe enclosures. The only easily modeled geometry at this time is the U frame. In addition, I am in the process of updating all of the worksheets to include a new baffle diffraction calculation and probably a polar response plot which should make the worksheets even better for the U frame simulation. But at this time, I do not believe I have a worksheet available that will accurate model the H, M, or Ripole geometry.

Martin,
I was thinking of the TQWP sheet for modelling just one side of the tapered cavity shape like with these M-baffles.