Push-push phase

[rick57]

My aim is to mimimise or cancel driver non-linearities. I've been searching on this, and understand that
push-push is more effective than push-pull.

Thanks to Dave's good drawing at www.t-linespeakers.org/FALL/push-push.html, push-push construction is clear.

Box modelling is as per a single woofer, apart from effective Vas doubling. Cone excusion halves.

with the 'normal':
- connected in parallel: impedance halves, sensitivity + 3 dB
- connected in series: impedance doubles, sensitivity - 3 dB

Am I right that the wiring is simply normal stereo electrical in-phase?

Thanks

 

[planet10]

My aim is to mimimise or cancel driver non-linearities.

Box modelling is as per a single woofer, apart from effective Vas doubling. Cone excusion halves.

with the 'normal':
- connected in parallel: impedance halves, sensitivity + 3 dB
- connected in series: impedance doubles, sensitivity - 3 dB

Am I right that the wiring is simply normal stereo electrical in-phase?

To get non-linearity cancelation you need push-pull where when one driver's cone is moving away from the motor, the other is moving towards the motor.

Referring to the attached diagram, normal push-pull is arranged as in #4.

Push-push is an arrangement where the drivers are oriented such that the newtonian motion of 1 driver is actively cancelled by another. The most effective way to achieve this is when the magnets are rigidly coupled as in #2.

#3 is an arrangement of push-pull that, if you couple the 2 drivers together mechanically with ready rod bolts or carefully placed braces, gets you most of the push-push advantage of #2. (one could mount the drivers such that they fire into a holey PVC pipe the same diameter as the driver)

#4 is the ultimate push-push push-pull. it also gives you a box that is 1/4 the size of 2, 3 & 4 (which all have the same box volume despite what you might gather from the cross-section). It also gives added isolation from waves coming back thru the cone.

In each case the drivers are wired so they all push on the boxes internal volume the same way.

2 drivers can be treated as a single driver with 2x the Vas (and 2x the Sd).

When wired in parallel, efficiency is increased by 3 dB and -- if driver by a true voltage source -- sensitivity goes up 6 dB (3 dB increase in efficiency + 3 dB due to the amp delivering 2x the power)

When wired in series, efficiency is increased by 3 dB and -- if driver by a true voltage source -- sensitivity stays the same (3 dB increase in efficiency - 3 dB due to the amp delivering half the power)

Sensitivity is a specification i'd like to see binned -- there are lots of amplifiers (tube amps of almost any ilk for instance) that do not satisfy the criteria of being a true voltage source.

dave

 

[rick57]

dave

(I'd moved to an option that might not have been the better one for the situation)

Which do you think is a more useful to cancel, non-linearity or the newtonian motion?

If the drivers aren't that good (current situation - some that i got for $20 each at a car boot sale, for a PC's sub) - probably better to cancel non-linearity, via push- pull?

While if the drivers are good, eg Peerless CSX (I also have), better to cancel the newtonian motion (push-push)?

And when you have 4 drivers (i have a set of 4 from a cancelled plan):
#4 (push-push push-pull) - what is the effective Vas and effects on dB & ohms?

Cheers

 

[planet10]

Which do you think is a more useful to cancel, non-linearity or the newtonian motion? ... the drivers aren't that good

A little ingenuity with the box bracing and #3 will give you all of the non-linearity cancellation, and most of the newtonian motion cancellation.

While if the drivers are good, eg Peerless CSX (I also have), better to cancel the newtonian motion (push-push)?

Yes. And gain the WAF. When i talked with Nick (Llambda) about these concepts he said that if using his drivers, there would be next to no gain doing push-pull.

And when you have 4 drivers (i have a set of 4 from a cancelled plan):
#1 (push-push push-pull) - what is the effective Vas and effects on dB & ohms?

I tend to like pairs of woofers so i would look at 2 push-push woofers, but when you have 4 drivers in isobarik configuration, except for the doubling of the volume inside the box that they take up, you treat them as a single driver. Vas, Qt, Fs is all the same. Efficiency goes up 3 dB, and assuming series/parallel we end up with a sensitivity increase of 3 dB.

dave

 

[rick57]

Thanks Dave,

when I read of this some time ago with the with hand built bass drivers and the bolt thru the pole piece http://www.t-linespeakers.org/FALL/push-push.html I was thrown off the benefits, but now the applications are clear

Cheers

 

[planet10]

but now the applications are clear