Considerations on group delay of different cabs loaded with same driver

[ThisBass]

Inspired by a post that TB member Billfitzmaurice did to the subject mixing of speaker cabs
Mixing Speaker Cabs - Definitive Answer! | TalkBass.com

See post number #6
"...But there are differences in the phase responses of different drivers and cabs, even different cabs loaded with the same driver, ..."

I'm asking myself if issues with same drivers are indeed sometimes experienced by users or at the other hand considered if different cabs loaded with same drivers shall be brought together.

Example:
A 410 ported cab loaded with Deltalite 2510 shall be mixed with a ported 210 cab also loaded with Deltalite 2510.

The Deltalite 2510 shows a nominal impdance of 8 Ohm.
The impedance of the 410 cab is 8 Ohm (Series-Parallel configuration of the drivers), the impedance of the 210 cab is 16 Ohm (Series configuration of the drivers).

The cabs shall be daisy chained (Parallel configuration of the cabs). Will it be fine mixing these two cabs?
The total impedance of the stack is 5.33 Ohm so, unless running in bridge mode, probably any solid state amplifier won't get issues due to a lower deviation of minimal 4 ohm load.
The power spread to the drivers is balanced, every driver gets exactly the same power.

If you say now that everything is fine with this stack because of:
1) no critical total impedance for the amplifier
2) balanced distribution of power for each driver
3) anywhere exactly identical drivers

I say stop it, it's not knwon anything about the way how these two cabs work together on group delay.

I did a simulation for a 410 ported cab and a 210 ported cab, both cabs loaded with Deltalite 2510.
The 410 cab inner volume is 4.8 cu ft and Tuning 55Hz, the 210 cab was simulated as a small cab 1.6 cu ft and Tuning 65Hz.

The predicted group delay looks as follows:
yellow is 410, white is 210

So look what happens around 100Hz. The difference of group delay is appro 1.5msec. Keep in mind that cycle time of 100Hz is 10msec, and a total cancellation would happen at a difference of already 5msec. (at half delay time of one 100Hz cycle). So what happens if mixing these cabs? Very likely some cancellation but, more important the cabs play like they where differently placed in depth by around 1.7 feet.


The next simulation shows the same 410 cab, but 210 now is simulated to be a 1/2 of the 410. That means inner volume 2.4 cu ft and Tuning 55Hz

No issues with different group delay nothng at all.


So you may ask the question why this thread. Aren't there countless threads related to this subject? The answer is very simple. Most of the arguments against mixing different cabs / different drivers refere to:
1) unbalanced power distrubution to the drivers
2) The large danger of overpowering/destroying 115 cabs if paired with 410
3) unpretictabel sound issues but, the reasons why in general remain unspoken, tireless stressing that identical cabs or cabs with identical drivers keep far away from non pretictabel sound issues. But identical drivers may indeed cause same sound issues unless the cabs are not designed to work together.

The only way to keep far away from any unpretictabel sound issues is to use absolutely identical cabs. Keep in mind that mixing a 410 cab with 210 cab or the mixing of a 212 cab with a 112 cab does not mean per se you are totally free from sound issues although the cabs you consider to pair with are loaded with exactly identical drivers.

 

[Bass_Pounder]

Just goes to show that the cabs internal airspace per driver and tuning have just as much effect even when using the same drivers.

 

[JimmyM]

Unpretectical?

 

[ThisBass]

Unpretectical?

Sorry because of my totally wrong spelling, I was meaning unpredictable. Correction is already done.

 

[JimmyM]

Ah, now I get it!

Welp, score another in a long line of reasons I prefer to keep stuff matched. And I didn't even know this one before.

 

[DigitalMan]

Also raises questions about combo amps with "matching" external cabs.

 

[JimmyM]

Some combos have ext cabs that truly do match. But yeah, good point.

 

[Downunderwonder]

Milliseconds matter.

Then someone comes along and tells us how great his group delayed stack sounds.

 

[ThisBass]

To be honest but, some of the well known advocate of a same driver size already read this stuff in the meanwhile, so it makes me wonder they refuse to comment.

 

[Downunderwonder]

It's Saturday night in US.

And you did a whizzbang summary that only the converted would understand.

 

[ThisBass]

You are right, and I just tried to delete this stupid post above.
So please don't worry about that stupid post.

 

[agedhorse]

Yes, mSec CAN matter BUT you need to understand WHEN and WHY.

If you look closely at the group delay graph of the 410 and 210, the reason for the differences are because the electro-mechanical systems are not the same. The driver is only one of the factors in the system. Exactly the same argument applies to mixed driver sizes (or even same size but different driver in 2 different cabinets)... if the tuning profile of the SYSTEM takes into account the phase response, which is related to group delay, you should be able to recognize that closing your eyes you may not be able to tell that the drivers may not be the same size because the responses are suitable tailored to work together.

Another thing that you should recognize is that a group delay of 2mSec at 100Hz is probably relatively insignificant within the context of the broad band system performance. The time constant of 100Hz is 10mSec so the resulting phase shift will be less than 90 degrees at 100Hz. Now there are also boundary reflections that factor into the system response and these tend to be longer and of more practical significance.

I really encourage really understanding this topic and thinking through the relative magnitudes of the data points before assuming something is proof that "not identical" is bad. In fact, identical drivers themselves are not really identical and may differ by as much as 20%.

Geeez, I wish some of these myths would die

 

[Rick James]

Bill has always been very consistent in stating that group delay doesn't matter, if the cab has high enough group delay to be audible its a flawed design and the group delay will be the least of its problems. The problems with mixing different speakers is what happens in the midbass and mids, where they can augment each other at some frequencies and cancel each other out at others.

 

[agedhorse]

Comb filtering (constructive/destructive interference) occurs with identical drivers as well. It's the result of the summation of identical signals with 2 physical origins. Different sizes just makes the comb filtering different, not necessary worse (or better)

 

[bebass]

...

I really encourage really understanding this topic and thinking through the relative magnitudes of the data points before assuming something is proof that "not identical" is bad. In fact, identical drivers themselves are not really identical and may differ by as much as 20%.

Geeez, I wish some of these myths would die

Bold mine

Thanks for posting this.
It would lead to another topic: Measured versus manufacturer's data - per speaker.
Then we could buy speakers in matched sets - like tubes
For Sale: Matched set, 15" Drivers

 

[agedhorse]

Yes, in theory drivers could be "matched", but in practice it's not likely to improve performance for the reasons I mentioned.

There is a big difference between theory and practice ime.

 

[ThisBass]

Bill has always been very consistent in stating that group delay doesn't matter, if the cab has high enough group delay to be audible its a flawed design and the group delay will be the least of its problems. The problems with mixing different speakers is what happens in the midbass and mids, where they can augment each other at some frequencies and cancel each other out at others.

By itself group delay does not matter. That means a decent group delay that is very typical with every ported cab does not matter. To the contrary sealed cabs show a slight group delay only.

"...But there are differences in the phase responses of different drivers and cabs, even different cabs loaded with the same driver, ..."

Same drivers may cause some differences in phase response thus causing different group delay. It is not the group delay by it's own that may cause issues, The statement referes to issues if group delay becomes very different because the phase response is different.

I'm pretty sure that nobody would ask a question about phase response / group delay if somebody else would ask a question at the forum for mixing a 410 ported cab with a sealed 410 cab.

yellow is 410 ported, white is 410 sealed
Both cabs loaded with Deltalite 2510

No need to comment


As the deltalite isn't a well suitable driver for sealed cabs, here we go with Legend BP102
yellow is 410 ported loaded with Deltalite 2510,
white is 410 sealed loaded with Legend BP102

Allthough the amplitude response graphs (here not shown) let assume both cabs should work very well in combination, the obvios difference in group delay would interfere the total response at the lows.


And finally something that would probably bedeviled at TB forum discussions, a 410 ported mixed with 212 ported Deltalite 2512
white is 212, yellow is 410

Really perfect, the deviation below 40Hz does not matter nothing at all.

Note:
these are some samples only to show what may happen even on same driver size. The deviations on group delay may become more dramatical for different (matched size) drivers loaded into different cabs. The deviation depends on different mechanism of the driver parameters and the individual cab design that fits the needs of a certain driver.

 

[ThisBass]

Yes, in theory drivers could be "matched", but in practice it's not likely to improve performance for the reasons I mentioned.

There is a big difference between theory and practice ime.

Hi Andy,
many thanks for all your input here.
I hope you know that I'm a guy who does never claimed a stack has to consist of same driver size only.
As I told above, I was inspired by the comment of Billfitzmaurice to show what may happen even on identical drivers.
The forum is full of posts giving advice to the mixing of different cabs loaded with same driver size. Mostly the posts relate to power distribution to cabs and/or drivers but, no concern nothing at all if the different cabs may do it well together.
If someone would ask a question to mix a 410 with 210 or 115 then the answer at TB forum is very foreseeable. Nearly nobody would ask of what type the 210 cab was neither would anybody ask any questions about loaded drivers and tunings but, the TB police would by default leave blank the 210 and put into jail the 115.

 

[agedhorse]

Using post 17 as an excellent example of the PRINCIPLES behind combining 2 cabinets, I think it's pretty clear that group delay (and thus an analogy to phase response) is not just dependent on driver size but on the topology and the tuning of the enclosure. If there was an example of group delay using the DeltaLite 2512 in a horn loaded enclosure, I think that would be an even more dramatic comparison and in fact will be WAY different that any of the above graphs even though the driver is identical.

I have taken a bit of grief on the subject of mixing driver sizes, but I have ALWAYS qualified that IF the design takes into consideration the phase response, the results can in fact be excellent. I think the group delay plots above really help put this into more easily understood terms of the kinds of things an engineer might consider when designing a product line.

 

[fdeck]

I think I've determined that for basic sealed and ported cabs (did not study any others), the frequency response, phase response, and group delay curves are all joined at the hip, below about 250 Hz. Two speakers with the same frequency response will have the same group delay, and vice versa. And you can't have a sealed cab with the same group delay as a ported cab.

My proof, for anybody who enjoys math: Invalid Link Removed