AR-9 and AR-90 UMR crossover differences

[baffled_loudspeaker]

Looking at the AR-9 and AR-90 crossover schematics, there's a difference in the upper-midrange schematic. There's no 1.37 mH in parallel on the -9 and the 6uF parallel cap in the -9 is 8uF on the -90. It amounts to a +1.5dB difference in the passband for the -90. Anyone know why the -90 UMR is different, despite both models having the same LMR, UMR, and tweeter? The -90 has more components despite it being the lower-priced speaker, and it's not a demonstrably worse crossover, it's just louder. The rest of the two crossovers are identical.

I've attached a SPICE analysis of the two crossovers. The UMR driver is represented by a 4 ohm resistor. There's a second plot with the -91 just for comparison. (The AR-91 also has a different UMR crossover, but it's obviously cost-reduced)

 

[Stimpy]

I believe the AR-9 schematic, that listed a 6.0uF cap on the UMR, was in error.  Both the 9 and 90, should use the 8.0uF capacitor.

[Stimpy]

AR-9

[DavidR]

15 hours ago, baffled_loudspeaker said:

There's a second plot with the -91 just for comparison. (The AR-91 also has a different UMR crossover, but it's obviously cost-reduced)

It's a slightly different driver; able to reach lower. IIRC its p/n 200028 for the 9 and 90 and 200032 on the 91.

There was no need for the notch filter that is on the 9 and 90 models.

[baffled_loudspeaker]

Ah, thank you for the correction, Stimpy. I was working from that incorrect schematic! I've updated my SPICE simulation, see attached. The 8uF cap reduces the discrepancy between the 9 and the 90, but the the inductor on the 90 still gives it an additional ~+0.8 dB of output to the UMR. Why should these two crossovers be any different?

[Stimpy]

There is an extra inductor in there...!

AR-90

 

[baffled_loudspeaker]

Yes! But why?!

[fedeleluigi]

2 hours ago, baffled_loudspeaker said:

Yes! But why?!

Hello,

Actually there is no extra inductor in the AR-90 because the AR-9 also has it!

The AR-9 crossover schematic in the library is wrong. In my post linked below you can find the correct schematic of the AR-9 redrawn by me.

https://community.classicspeakerpages.net/topic/11407-my-ar-9-capacitor-recapupgrade-project/?do=findComment&comment=137305

 

 

[baffled_loudspeaker]

Haha well that explains it! Thank you Luigi! This should be updated in the Library section.

[AR surround]

14 hours ago, baffled_loudspeaker said:

Haha well that explains it! Thank you Luigi! This should be updated in the Library section.

Are you going to update the SPICE diagram to reflect that both speakers have the same inductors?   I ask because I am using AR9's as front speakers and AR90's as surround speakers in a home theater setup.  I had to add some additional series resistance to the tweeter and UMR of the AR90's to achieve a response closer to that of the 9's.

[baffled_loudspeaker]

6 hours ago, AR surround said:

Are you going to update the SPICE diagram to reflect that both speakers have the same inductors?   I ask because I am using AR9's as front speakers and AR90's as surround speakers in a home theater setup.  I had to add some additional series resistance to the tweeter and UMR of the AR90's to achieve a response closer to that of the 9's.

To summarize the thread, we've figured out that the -9 and -90 have the same UMR crossover. The popular -9 crossover schematic floating around the internet was incorrect in two ways. Luigi Fedele's updated AR-9 crossover schematic (attached) is supposedly accurate and shows that the it is the same as the -90. I say supposedly because I haven't verified it myself, but I have no reason to doubt the other contributors in this thread. We have an AR engineering drawing of the -90 schematic.

As to why you're experiencing greater output from tweeter and UMR of your setup, I would first suggest that this is due to acoustics. I say that because both transducers are too loud - they're separate drivers and separate sections of the crossover. The thing they have (nearly) in common is the acoustic path, particularly the distance from the speaker to the listener and angle / directivity. I'd experiment with these variables before modifying your speakers!

A good thing to keep in mind is the Rayleigh distance,  which is an approximation of where the nearfield of a baffled piston ends and the farfield begins. The farfield is where the pressure radiated from a transducer begins to decline at 1/R. The Rayleigh distance is R_0 = [surface area of the piston] / [wavelength of interest].

 

[baffled_loudspeaker]

For the record, this is my SPICE analysis of the voltage transfer function of the AR-90 crossover. This does not include the impedance of the transducers, only a resistor equal to the DCR. This is also not the acoustic output of the speaker, which is a whole nother can of worms. I intend to complete both of these in the future, time permitting. My true goal here is to replace this crossover with bi-amping or tri-amping (quad-amping?) and an active crossover.

[DavidR]

@baffled_loudspeaker > could it be that the AR90 LMR/UMR and tweeter are wired out-of-phase and the 9 is not ?

 

[baffled_loudspeaker]

23 hours ago, DavidR said:

@baffled_loudspeaker > could it be that the AR90 LMR/UMR and tweeter are wired out-of-phase and the 9 is not ?

Are you referring to @AR surround's loudness problem? Or something regarding the crossover schematics?

You are making me look closer at the -9 schematic, and that redrawn -9 schematic has the tweeter, UMR, and LMR transducers wired in opposite phase from the AR-90 schematic (below). @fedeleluigi, can you comment on the accuracy of your AR-9 schematic and the wiring colors and polarities?

 

[DavidR]

On 10/12/2023 at 5:54 PM, baffled_loudspeaker said:

9 schematic has the tweeter, UMR, and LMR transducers wired in opposite phase from the AR-90

The 90 is wired out of phase. The 9 is wired in phase, same as the woofers.

Just didn't know if that might account for your query re. 90 UMR not the same as the 9 in your 'sweep' - don't really think so but thought I'd throw it out there.

Another question for you: Are the UMRs all p/n 200028 in both sets of speakers? Sometimes they get swapped out for p/n 200032 as used in the 91 and other speakers; and then there's the 200044 UMR/Mid.