Beyond the Ariel

[Caferacer]

And let me add that a 4x12 16 ohm cabinet sounds better than a 2x12 8ohm cabinet also. It doesn't matter how it's wired. More speakers sound better. Especially at concert volume.

[john k...]

Originally posted by Caferacer
"quote:
Originally posted by dobias

Series-parallel wiring appears to be an exercise in frustration since two, in parallel, will output as much as four in series-parallel?
Please help me understand where (I'm sure) I must be wrong.

dobias "



This is an area where theory and reality go seperate paths. I don't know if it's because of more driver area. Which increases the size of the baffle. Or less individial driver distortion. Or both. But I can tell you from personal experience. I play electric guitar. And a series/parallel wired cabinet with four 12" drivers just blows a single 12" cabinet out of the water with the same amplifier setting. It sounds louder, cleaner, warmer....... just better. Trust me.

There is no conflict between theory and reality here. The problem is confusion between sensitivity and maximum output capability. As a simple example consider only two drivers. Wired in series the sensitivity remains unaltered. That means that for a given voltage across the series connection the SPL output will be the same as a single driver. But the voltage required to push both drivers in the series connection to their maximum excursion will be twice that required for a single driver. Since the series system can handle twice the voltage it can deliver twice the maximum SPL as a single driver, +6dB, even though the voltage sensitivity remains the same.

With the parallel connection the voltage sensitivity increases by 6dB. But the voltage required to push both drivers to their maximum excursions will remain the same as for a single driver. Thus the maximum SPL output of the parallel pair will again be +6dB compare to a single driver.

If you consider the power delivered into the parallel and series connections at the max SPL limit you would find they are the same. That is, compared to a single driver pushed to the excursion limits the series pair requires twice the voltage but the same current. The parallel pair requires the same voltage but twice the current. Since power is IV we have 2IV in both cases compared to the single driver. Thus the power input to two drivers wired in series OR parallel, pushed to max excursion is +3dB compared to a single driver and the out SPL is +6dB in both cases for or a +3dB increase in EFFICIENCY.

This may seem confusing but remember not to confuse efficiency with sensitivity. Sensitivity is just a measure of some single output parameter (SPL) with respect to a single input parameter, typically voltage. Efficiency is a measure of how well a system converts input power (or energy) into output power (or energy) in a different form.

[mige0]

Hi

IF you want to keep Qts under control any series OR series parallel wiring is contra productive.

For highish Qms drivers Qts changes roughly like:

Qts(new) = Qts(original)*(1+R(series resistor)/R(voice coil))

This means that for a 15NMB420 with
Qts = 0,29
R(voice coil) = 5,6

you reach a Qts(new) of 0.7 with a series resistor of 8 Ohm.

IF you wire two 15NMB420 in series you roughly insert a series resistor of 100 Ohm at the resonance frequency as can be seen in the impedance graph.
This equals to a Qts(new) of roughly 5.6 !!
(a little bit lower actually, due to the not so high Qms of 6 in the case of the 15NMB420 )

Not a good idea unless you are designing instrument combos IMO.



Greetings
Michael

[john k...]

Originally posted by mige0
Hi

IF you want to keep Qts under control any series OR series parallel wiring is contra productive.

For highish Qms drivers Qts changes roughly like:

Qts(new) = Qts(original)*(1+R(series resistor)/R(voice coil))

This means that for a 15NMB420 with
Qts = 0,29
R(voice coil) = 5,6

you reach a Qts(new) of 0.7 with a series resistor of 8 Ohm.

IF you wire two 15NMB420 in series you roughly insert a series resistor of 100 Ohm at the resonance frequency as can be seen in the impedance graph.
This equals to a Qts(new) of roughly 5.6 !!

Not a good idea unless you are designing instrument combos IMO.



Greetings
Michael

This is incorrect. The damping is controled by the reverse current generated by the back EMF of the drivers. The reverse current is given by Ib = Vback/(Re + Rg) where Rg is the amplifier's output (Ro)imepdance and any other series impedance. When connected in series Rg = Ro + Rg of the second driver so Ib = Vback/(2Re + Ro). At first glance this looks bad but then you must considet that for two driver in series Vback also doubles. So Ib = 2Vback/(2Rg + Ro) . If Ro is small Qts will not change significantly. If the amp's output impedance is high the series connection will actually move Qts back towards the short circuited voice coil value.

The problem with series connections is that unless both driver are identical the back EMF of one driver can modulate the second driver and vice versa. There is also the problem of nonlinear distortion from on driver modulating the other as well. Series connections can be problematic in this regard.

[soongsc]

Originally posted by john k...


This is incorrect. The damping is controled by the reverse current generated by the back EMF of the drivers. The reverse current is given by Ib = Vback/(Re + Rg) where Rg is the amplifier's output (Ro)imepdance and any other series impedance. When connected in series Rg = Ro + Rg of the second driver so Ib = Vback/(2Re + Ro). At first glance this looks bad but then you must considet that for two driver in series Vback also doubles. So Ib = 2Vback/(2Rg + Ro) . If Ro is small Qts will not change significantly. If the amp's output impedance is high the series connection will actually move Qts back towards the short circuited voice coil value.

The problem with series connections is that unless both driver are identical the back EMF of one driver can modulate the second driver and vice versa. There is also the problem of nonlinear distortion from on driver modulating the other as well. Series connections can be problematic in this regard.

I have experieced this where two drivers were connected in series, and they just would not move together, resulting in very bad bass reproduction. If they are in one single enclosure, things are worse.

[mikey_audiogeek]

Hi Brett,

Thanks Mike. What amp are you using?

A John Swenson - style transconductance amp using Magnequest iron. I should also point out that the OB's are in a room about 4m x 4m so i don't need a lot of SPL. Next step is to double up on drivers. Lynn is of course right about the "merits" of EQ however the transconductance approach is cool, plus I'm not using much power and plan to increase Sd by doubling the drivers. One aspect of the current drive approach is that it also rolls off the output below Fs, reducing excursion - at the levels I listen at (100dB peak) you can barely see the cones move.

Cheers,
Mike

[Brett]

Originally posted by mikey_audiogeek
Hi Brett,

A John Swenson - style transconductance amp using Magnequest iron. I should also point out that the OB's are in a room about 4m x 4m so i don't need a lot of SPL. Next step is to double up on drivers. Lynn is of course right about the "merits" of EQ however the transconductance approach is cool, plus I'm not using much power and plan to increase Sd by doubling the drivers. One aspect of the current drive approach is that it also rolls off the output below Fs, reducing excursion - at the levels I listen at (100dB peak) you can barely see the cones move.

Cheers,
Mike

I'm not aware of this, but some googling turned up some posts at the Magnquest forum. The schematics didn't show though, and I'd love to see them if you have them, especially the 9W version mentioned here.
http://db.audioasylum.com/cgi/t.mpl?f=magnequest&m=5167

I like building and tweaking amps as it relaxes me, so something different would be nice.

I'm going to be using 3 Eminence B15's and worked out that in parallel 2R should be enough to get the Qts up where I want it. I have 8 secondaries on my large OPT, so it should handle it fine when I configure it that low a secondary Z.

Some years ago I built Gary Pimm's PP47 amp and loved it. As I have a stash of NOS Mullard 47's, perhaps that might be it for the mids. 4W in a room the same as yours should be adequate even when I want to headbang.

[dobias]

John K...
Thank you so much for your description. I may take awhile to digest what you have decribed but I feel as if you've given me the answer to my question.
Now I need to decide if I want to live with my 6 ohm woofers in series wiring.
Your explanation was an education I've been looking for.
dobias

[mikey_audiogeek]

Hi Brett, sorry can't help, you could ping John on the Magnequest forum and get him to repost. Otherwise the post in the attached link gives all the info, and Gary's webpages are back online.

Cheers,
Mike

John Swenson transconductance amp post

[Brett]

Cheers Mike, I'll email him through AA.