AR Bi Amp Hybrids (ActiveXover Stacks) CR65, AR581w, AR51w, AR358s, AR98T

[Aadams]

This post is about a tweak that will work for any Classic AR with pots and not about the mod configurations that dominate this thread. 

I forgot to mention one big impediment, IMO, to making a properly working AR3a or AR5 produce a good image.  The tweeter and especially the mid output must be virtually identical across the stereo pair.  There is so much variation in pots that you cannot do this visually, it must be done by ear and it is not easy with one person doing everything.  I only noticed this recently when I made a change in the Satstack that produced a discernible leap in its imaging and which materially separated it from the AR5.  The Stack was suddenly razor sharp by comparison and I was not happy that the 5s were a little off balance.    The solution was to adjust the mid controls until the drivers produced the same output level.  Until recently I thought “close” was good enough but it is necessary, for imaging purposes, to dial them in exactly.   I am speculating here but I suspect the dome mid control variation may be a big reason why the 2ax has a better imaging rep than the classic domes.  The first 5.5 octaves of music are covered by the 2ax woofer whereas the domed systems barely cover 4 octaves in the woofer before crossing to the mid which may or may not be set correctly according to pot condition and listener care. 

Conclusion: Classic dome ARs are fit to excel at imaging but they require more attention because of pots or even Lpads.

 

Adams

[Aadams]

More Classic model tweak info:

The challenge for the 3a in this AR358 configuration, when compared to an AR9, will not be bass output but the imaging capability.  While it is fact, a properly functioning AR9 can play practically flat for almost 10 octaves, it is also fact that the boundaries of most music occur will within the range of a properly working 3a.  The AR9 was designed to be accommodating in a variety home listening environments.  It was and is, by comparison to most full range speakers, automatically flat and ready, out of the box, to provide an optimum full range stereo image with very little fiddling and futzing around if installed within simple and practicable guidelines.  The AR9 appears to be AR’s first great imaging loudspeaker.  Now days legend says the Classic and ADD domes could not image well and were weak, in near field listening comparisons to more modern speakers, and for this reason AR moved on with a tour de force system that still impresses.

Legend is wrong.    

In working with my AR5s and 3as I have discovered these old speakers can image very well in the near field if they are adequately restored to have proper output on all drivers and the rear gain controls have sufficiently fine granular adjustment to allow exact balance between left/right mid to mid and tweeter to tweeter.  The pots are functionally another set of balance controls that must be accounted for to achieve good imaging with the Classic ARs.

It would be a lot easier if there were no controls, as later happened, but these variable controls were a great idea before equalizers.  For imaging they are a nuisance but the imaging capabilities of classics are real and they can be astonishing with their very wide dispersion.

Adams

[Aadams]

On 4/28/2019 at 12:17 AM, Aadams said:

This may seem obvious but to be clear, if the sound is center left, for instance, you have two choices; lower the gain on the left speaker mid to PUSH the balance to the right OR raise the gain on the right speaker mid  to PULL the balance to the right.  This works.  When the sound is centered Turn ON stereo mode and check for centered image. 

The question has arisen, “Why not just turn the main balance control until the image is centered?”

 

Answer:  In the early 12 inch ARs, large, fundamental chunks of midrange frequencies are directed through the woofer.  There is no woofer pot so the main balance control is the only means to balance the woofers.  By balancing the tweeters and especially the mids you are aligning their output with the woofers so the midrange image is not skewed at the speaker box.

[Aadams]

 I have been procrastinating on putting the amps in place for my AR358/AR9 imaging comparison when I got the opportunity to compare an AR51W to an AR981W.  I have spent most of the day comparing the two systems.  The 98 is a bit more sensitive than the 5 but once I found the volume equalization point it is impossible to consistently guess which system you are listening to in this configuration.   I am doing the switching and even I lost track at times.  The sound field produced by either speaker is so similar at five to seven feet they are indistinguishable, except in direct AB comparison, which leads to this conclusion: The 98 upper range driver combination is formidable.  I don’t think I have heard such clarity from another AR speaker.  The only thing I have that is comparable is my SatStack but the little 4” drivers get messy when crossed over at 100hz as these speakers were today.

I listened to the AR98 briefly in a standalone config away from the wall and off the floor and it sounded thin, as expected.   I connected the 98 to the biamp system and changed the crossover from 200hz to 100hz, a point that is tolerable for the 98, to share the same bass subsystem as the AR5.  The last octave and a half of bass was eliminated as a point of comparison. 

Impressions

Imaging

Both systems image well.  On orchestral recordings The AR5 sounds a bit more expansive where the 98 is more concise.  On studio mixes the 98s had apparently clearer vocals but the instrument and performer placement were the same IMO. 

Timbre

So close it doesn’t matter but the mid-range seems a bit more assertive in the 98 and cymbals seem too forward on some recordings but this is just my opinion.   I could have fiddled with the equalizer to bring the 98 to the 5 but it would not have the changed the difference in clarity between the two.  Had I done the reverse and equalized the 5 to the 98 it would not have changed the clarity difference and I might have lost a tweeter.  The timbre may have played a role in the perception of clarity.  At one point I turned the 5 mids to full output, which didn’t affect the difference in clarity but then I had to rebalance the 5s, an inconvenience of the classic ARs, at least for imaging.

Too bad AR didn’t make this driver combination in an 8 inch 3 way, it would make a perfect top for a subwoofer.

[Aadams]

I spent most of this afternoon comparing the 981w to the Satstack.  Eventually, I was able to get the two systems volume and image equalized from the listening position.  Why would I bother? The answer: The more I listened, I increasingly felt that the AR98LS runs a little too hot in the mid-range and I thought I could use the Satstack (Boston CR65s), which are not shy speakers, to help verify what I thought I was hearing.

The bass crossover was raised to 150hz, a setting that seems to work well with the 98 and puts the CR65s back in their comfort zone.  The short version is, I discovered, in the octave centered at 4k, the AR98ls needs taming to bring it into line with the CR65s.    

I don’t think this is news.  I have read in CSP where the LSi version of the 98 fixed some irritating sonic aspect of the LS and that the mod can be applied to bring the LS to the LSi spec but I will use an equalizer.  

The bottom line:  A hot midrange in the 4k range skews voice timbre and explains why cymbals and gated snare are so prominent on a 98ls.  Still it can image very well and clarity is superb.  

[ligs]

"The bottom line:  A hot midrange in the 4k range skews voice timbre and explains why cymbals and gated snare are so prominent on a 98ls.  Still it can image very well and clarity is superb."

Let us assume both AR 98LS and CR65 have the same axial response around  4K hz, AR 98LS  would sound louder just because the wide dispersion characteristics of the 1.5" dome midrange. I think our ears receive directional clues from the first arrival sound but our perception of loudness is from the total acoustic energy(power response) from all directions. Many audiophiles prefer an overall frequency response for a speaker which is sloping very slightly down from bass to mid to high range. 

 

In addition, our ears are most sensitive in the midrange. Proper voicing the midrange, hence , is critical for naturalness and musical enjoyment. 

 

[Aadams]

59 minutes ago, ligs said:

Let us assume both AR 98LS and CR65 have the same axial response around  4K hz, AR 98LS  would sound louder just because the wide dispersion characteristics of the 1.5" dome midrange.

Ligs, good to see you back. 

You may be right about the energy.  It takes three CR65s stacked just right to get the same expansive sound of one AR3a with the mid turned down about a 1/4 turn.   Excluding the 58s and 915,  which were "value" speakers,  the LS series was the first , I think, in which AR used a mid dome wide open with no damping material, attenuation circuit or mechanical focusing.  As with previous domed mids, this one probably measured flat in the test environment but in reflective surroundings needed toning down but I am just speculating.   My AR9 mids with their concave horn covers and acoustic blanket are set wide open and I don't hear this brightness problem but an AR3a mid wide open in the same environment can be quite annoying and must be turned down. 

I have been running with grills off awaiting new material.  The grills originally had a foam inner surround that may have subdued the reflected energy.  When I recover the grills I will replace the rotted foam material on the inside edges and we shall see if there is a difference.  I would prefer not to use the the equalizer.

Adams

[Aadams]

Ligs I believe you were correct.

I finished repairing the 98ls grills today and re-installed sound absorbing material around the inside edge to replace the original foam that had rotted.  I put the grills in place and discovered that the grills of the AR98ls are an integral part of the speaker system.   They were instantly transformed into what I had expected them to be from the outset. 

I lowered the bass crossover to 100hz and compared them to the AR5s that I like so much.  It is now very easy to dial them into the AR5s using only volume controls and with zero equalizer adjustments.  When grills are off, I must lower the entire band between 500hz to 5000hz two to three decibels to get them where I want which is like turning down the mid control on an AR3a, practically the same adjustment over the same band of frequencies.  They also are bit more focused, akin to an AR9 but they also sound  expansive, similar to an AR Classic dome but without the muss and fuss of attenuators.

The material with which I lined the grills is ½ inch polystyrene pipe insulation.  I don’t know how it compares in performance to the original foam, but it works for my needs, is readily available and easily fits in the space with double sided tape.

The 1985 Audio review of the 98ls was performed with grills off.  The reviewer must have had a very dead listening space or ears of steel or didn’t care.  The review sounds perfunctory anyway, like he was making a deadline but he did give it good marks for imaging and I agree.  The big leap here is a single front facing 12" with the mid range neutrality of a properly setup, flush mounted 3a or an AR5, but with a bit more clarity and lot more power handling.

So here is another free tweak that has nothing to do with mods.  If you own an LS series and you are missing the interior grill surround foam then you are probably not hearing the speakers in the way they were intended.

I will be moving these 98s into the room with the AR9s to compare imaging. Combined with the AR58s for bass they will become AR9858s but I still haven’t moved the amps.

Aadams

[ligs]

I am glad you can contour the sound of speakers to your liking with simple adjustments. 

Here is an article talking about the effect of felt ring(or block) on the tweeter, suggesting a cleaner sound is possible with the  proper use of felt. 

https://www.speakerdesign.net/felt/felt_ring_vs_blocks.html

 

George

[Aadams]

On 5/11/2019 at 7:12 AM, ligs said:

am glad you can contour the sound of speakers to your liking with simple adjustments.

Ligs

I was a bit hasty when I said the grills solved the problem.  My premature conclusion was based on listening almost exclusively to vocal recordings, when I should have included a good selection of orchestral.  Over the last two days I have been listening exclusively to orchestral music and eventually realized French Horns were sounding more like trombones, Trumpets sounded like Piccolo Trumpets and Violins had way too much sheen on the high strings when heard through the 98s.  Nothing offensive to the ears but it just wasn’t right.

Bottom line is, even with the grills and absorbent foam, I must lower the entire 500 to 7k band about 2db to bring the 98 into line with the 3a, 5 and the CR65 stack.  Other than that, my previous opinion is unchanged.  It is an excellent imaging speaker with an expansive sound but not at the AR Classic level of expansiveness.

 This equalizer issue makes it much more inconvenient for me to directly compare the 98 to the AR9.

I wish AR had given the 98ls more attention before they released it into the wild. 

Aadams

[Aadams]

Problem solved. 

The AR98ls now has the sound profile of an AR5 without using an equalizer.  Eventually, I identified the problem of over-brightness as being totally in the dual dome mid/tweet unit and not the entire 500-7k band but more like 1500 – 7k. 

 My solution was revolutionary and there was lot of trial and error but in the end, I settled on using one layer of felt cloth and three layers of muslin placed in the top 7 inches of the grill which covers both domes but not the 8 inch LMR driver.  The 98ls is now, to my ears, a very enjoyable speaker.  It still has an expansive quality and its clarity and imaging are better than the 5 but not worlds better. The difference is only obvious at the moment of the AB switch and afterward for as long as you can remember which speaker was last switched.  Once you refocus on the music it is easy to lose track of which speaker is operating.  

This AR98ls is the only AR 4way I have heard with a single front facing 12” woofer.  It has a 200hz crossover at the woofer which, with its 4 way configuration, gets the bass out of the mid-range but to get the smoothest bass it is designed to be placed on the floor, its back squarely against a wall or flush mounted in a wall or bookshelf, vertical or horizontal.   Floor placement puts the tweeters and mids between 16 to 29 inches off the floor, which seems low to me and is at odds with the AR recommendation to place the tweeters at ear height for listening.   There is a shelf/floor switch near the input terminals, but it has nothing to do with bass frequencies.

As a passive subwoofer only, the 98ls is not very useful.  The 200hz crossover is an impediment because an active crossover point should be at least an octave away from a passive crossover point to limit interference effects, which I have discovered can be audible.  For the 98ls the active crossover should be 100hz or below.  This means any AR 12inch 3 way from an AR3 through the LSi series would be a better choice for duty as a passive subwoofer because, in those units, the entire bass frequency band, up to the lower mid-range, is available for a crossover choice.

The best use for the 98ls IMO, is as a full range speaker, if you can deal with its placement requirements or as a top for a good subwoofer which would give the 98ls greater freedom of placement in the room.

It could also be used as a bottom for a good modern two-way speaker crossed over at 400hz but this doesn’t relieve the placement constraints of the front firing woofer and mid.

I am still working toward a comparison of the 98ls to the AR9 but have not decided if the 98 will be full range or a bi-amped top.

[Aadams]

The AR98LS has been modified to achieve its highest purpose at my house.  It is now a passive subwoofer, which means the internal crossover has been bypassed and the woofer directly wired to the amplifier with the active crossover set at 200hz in this case. 

The  woofer sits very low in the 98LS cabinet which was intended to be placed on the floor and against the wall,  but I turned it sideways so the baffle is perpendicular to both reflective surfaces.   The bass is very smooth.  No peaks or dips that I can detect by ear from the listening position.

The 98ls system required too much effort to fit into my scheme of things.  Without an equalizer it is not a great sounding speaker, in fact it is harsh. With an equalizer, the adjustments that must be made aren’t of the minor fine-tuning category but gross adjustments in bands that are just not normal for a linear speaker.  

The modification I performed is completely and easily reversable.  No harm was done to the original speaker system save one tiny hole that can be made to vanish with little effort. 

The AR1w pair has been re-jumpered to an AR3a which now sits next to the AR5 and Satstack. 

Now, with the 98Ls off the menu, I am back on the original plan regarding the AR9 and 358 which can also be flipped into a 581w.

[ligs]

"The AR98LS has been modified to achieve its highest purpose at my house.  It is now a passive subwoofer, which means the internal crossover has been bypassed and the woofer directly wired to the amplifier with the active crossover set at 200hz in this case. 

The  woofer sits very low in the 98LS cabinet which was intended to be placed on the floor and against the wall,  but I turned it sideways so the baffle is perpendicular to both reflective surfaces.   The bass is very smooth.  No peaks or dips that I can detect by ear from the listening position."

Did you notice any difference in the bass quality of the AR woofer driven through the crossover vs driven directly by the amplifier?  Without the resistance(perhaps as high as 0.8 ohm) from the crossover inductor, the AR woofer may sound more controlled and more extended  in the bass. 

[Aadams]

2 hours ago, ligs said:

id you notice any difference in the bass quality of the AR woofer driven through the crossover vs driven directly by the amplifier?  Without the resistance(perhaps as high as 0.8 ohm) from the crossover inductor, the AR woofer may sound more controlled and more extended  in the bass.  

Ligs

I don’t know for sure because the placement of the 98 when running through the crossover was always baffle toward the listener and above 100hz.   After the mod I placed it in exactly the same spot where the 1w sat but woofer on the floor as described.  The bass gain remained exactly where the 1w gain was set.

My initial subjective perception is, after playing Flight of The Cosmic Hippo and a good sampling of other music, the bass is not quite as strong at the low extreme as the 1w but the 98 bass is overall smoother.  When the 1w was in the same position, woofer down but 1 ft above the floor, it set off objectionable room modes at the listening position and thus always sat inverted and was raised as shown below which was quite satisfactory.  I have never installed a 3a/1w cab directly on the floor.

Adams

 

 

[Aadams]

Today I setup the AR58s pair in the bi amp configuration as tops for the AR98ls subwoofer and compared them all afternoon with the AR5s.

 I got my 58s about 4 years ago after reading in the CSP that it was perhaps the ultimate evolution of an AR bookshelf speaker.  My infatuation subsided, when I realized the AR58s is hard to live with unless you are willing to use an equalizer to tame the mid-range driver to compensate for its lack of an attenuator switch.  The ads described the 58s as a “value” system for those wanting top line AR sound for a low purchase price.  It was in fact an AR91 in a plain wood case with no attenuators and no acoustic blanket and was probably a means to clear from inventory the last of the tweeters and mids that were used in the vertical series.  Used with an equalizer it works well.

Another issue for me is the front facing 12” woofer that is crossed over at 700hz, mixing low bass with the lower mid-range in the same driver, which,  IMO,  muddies human voice and is in fact the opposite direction AR took with all of its TOTL  12” systems.   

In the bi amp configuration, the 58s woofer works as a large LMR driver between 200 and 700hz, the whole system effectively working as a big satellite for a subwoofer. 

The AR5 and 58s were on a switch connected to identical electronics and source.  The bass below 200hz for both systems was identical.  The systems were sitting adjacent, about 3ft above the floor, at least 6 inches from a rear wall and no other boundaries closer than 4ft.

I guess I was hoping for a miracle, but it didn’t happen.  The mid is still too bright by about 3db but using the equalizer allowed me to dial it in to be pleasant sounding and make it very difficult to distinguish from the AR5, with very good imaging, though not expansive like the 5 or the 3a.   The 58s can be an excellent top for a subwoofer but, like the 98ls, doesn’t fit into my scheme of things so it will go back to subwoofer duty. I don’t mind adjusting gain when switching but having to remember equalizer settings from speaker to speaker is too much. 

On a positive note, the 12” woofer works very well as an LMR up to 700hz, if it doesn’t double as a subwoofer.  Male and female vocals were very clear and sounded virtually identical to the AR5, once the midrange driver was turned down via the equalizer.  

[ligs]

"I guess I was hoping for a miracle, but it didn’t happen.  The mid is still too bright by about 3db but using the equalizer allowed me to dial it in to be pleasant sounding and make it very difficult to distinguish from the AR5, with very good imaging, though not expansive like the 5 or the 3a.   The 58s can be an excellent top for a subwoofer but, like the 98ls, doesn’t fit into my scheme of things so it will go back to subwoofer duty. I don’t mind adjusting gain when switching but having to remember equalizer settings from speaker to speaker is too much. "

Even with 3 db adjustment in the mid, you still favor AR5 over AR58?

[Aadams]

On 5/30/2019 at 11:18 AM, ligs said:

 Even with 3 db adjustment in the mid, you still favor AR5 over AR58?

Ligs

 The short answer is I don't know, because I must move too many knobs and  buttons to do a fast A/B switch. I do prefer the expansive sound of AR classics especially on live recordings but they require a fair bit of attention to make them image well where the 58 can image with minimal attention. Running full range, the 58 is average even when equalized, but crossed over at 200hz with the woofer as an LMR and the HMR subdued 3db over its entire band using an equalizer I could easily live with it.  I took them down because they required to much fiddling to easily switch and they were taking up space. That said, I have plan that may get them back as permanent tops but first I must install Lpads in some 3as to know for sure.  Right now life is sucking up hobby time so the going is slow.  Thanks for your interest. 

Adams

   

[ligs]

I have 4 AR 91/92/58 exposed dome midranges in my parts bin. I may do something with them someday yet. I wonder what are the real differences between exposed dome midrange and fried egg type(AR3a, 5), in terms of sound quality.

 

George

[Aadams]

On 5/30/2019 at 12:49 PM, ligs said:

I have 4 AR 91/92/58 exposed dome midranges in my parts bin. I may do something with them someday yet. I wonder what are the real differences between exposed dome midrange and fried egg type(AR3a, 5), in terms of sound quality. 

 

George

From the 58s backward ,according to people who know, there should be zero difference in sound quality.  For same impedance, it is practically the same driver from the 3a to the 58s with different faces for different dispersion characteristics and different power handling capabilities. All of them can be made backward compatible across the entire range but there are limits to forward compatibility.   Anyone out there, feel free to correct me on that.

Adams

 

[Aadams]

On 5/30/2019 at 11:18 AM, ligs said:

Even with 3 db adjustment in the mid, you still favor AR5 over AR58?

Ligs

I put the 58s back as a top but this time comparing to an AR3a.   As you know the 58s has no attenuators and IMO is not bearable without an equalizer.  To improve comparability, I configured the 3as, which have new Chris tweeters, to mimic the 58s by replacing the pots with Lpads and resistors, then setting them to full increase which removes the Lpad from the circuit.  The output at this setting is very bright, much like the 58s.

Both speakers are setup as tops for a separate bass system crossed over at 200hz.  All electronics, including equalizer settings are identical.  To compensate for no attenuation controls, the 12db /octave equalizer was set at 3db down on the 1k, 2k and 4k sliders, falling completely within the span of the 58s mid driver which is crossed over between 700 and 7500hz. Setting up in this fashion allowed rapid switching between the two speakers, needing only a minor volume adjustment to compensate for the lower sensitivity of the 58s.

In this configuration the 12inch woofer of both systems is an LMR driver. Bass performance is not a factor.

All listening was done from 5 to 7ft with the speakers set away from the wall 30” off the floor.

After listening the entire afternoon, I would say there is no practical sonic difference between the 58s and the AR3a when setup in this way.  

I spent a lot of time seeking real substantive differentiation and if there is any it is this; The 58s is a bit focused compared to the 3a but this can only be noticed when listening to recordings done in live venues that are through performed, basically all large scale classical recordings and most stage performances before a live audience.    This is the only perceived advantage of the 3a over the 58s in the tested configuration.  To answer your question, I prefer the 3a or 5 sound because I listen to a good bit of orchestral music but it would be easy to make do with the 58s in this configuration.

Used as a top for a subwoofer, the 58s has at least two advantages over the 3a. 

1. It requires zero effort to balance the driver output for imaging.

 2. Its vertical driver alignment does not require toe in to get the best image.

 

Something else I learned is Lpads with resistors are superior to the cleanest 50 year old pots.  They provide finer control and they have a built-in bypass position.  I will be installing LPads in my AR5s.

 

Adams

[Aadams]

Lately, I have spent so much time comparing various AR speakers, tweeters, and subwoofer choices that it occurred to me that I may have drifted far away from the original premise and lost my bearings.  It rained all day so I used the afternoon to circle back to the original idea for this thread which is using small speakers to duplicate the AR sound. 

The Stack is back and is now sitting atop an AR3a with which it can easily be compared for imaging and overall balance.  The 3a has rebuilt tweeters and good pots.  The view you see below is close to the listening position.   

With the stack configured in the pictured position there is virtually no difference in sound.  When the switch is made the volume adjustment is so rapid that you truly cannot hear a difference, though I believe the stack is somewhat clearer on vocals but it is awful close.

The 12” in the AR3a is serving the LMR from 200hz to 525hz and is as clear as the little 5inch in CR65.

Bottom line is I am not off track.

Adams

[ligs]

On 6/7/2019 at 6:19 AM, Aadams said:

 

"Something else I learned is Lpads with resistors are superior to the cleanest 50 year old pots.  They provide finer control and they have a built-in bypass position.  I will be installing LPads in my AR5s."

Adams

The old pots are just  variable 15 ohm resistors. The amplifier sees different loads (resistances) depending the settings. Lpads , on the other hand, provide  constant loads(typically 8 ohms)to the amplifier. 

 

[Aadams]

If you like the AR9 then you will probably like the A58s mated with an AR1w in an AR581w configuration.  The short version is, I have listened to them, side by side, for several hours over the last two days and the sound is virtually identical.

All speaker attenuators were set to zero.  The signal path was identical except for power amps.  The AR9 was running full range on its own amp, while the 581w was crossed over high and low at 200hz.  The attached photo shows the AR1w, which is an unjumpered 3a, turned with the baffle perpendicular to the wall which places the woofer perpendicular to the floor and wall, in the same fashion as an AR9 and yields excellent flat bass.

The music was mostly pop/rock vocals with a bit of classical near the end to check imaging and timbre of orchestral instruments.  None of the music used extended below 42hz.  Listening distances were between 4 and 8 feet.  This was not about filling the room with music but comparing the character of their sound up close.

To begin, the 58s was configured to run full range, in order to gauge the amount of difference between it and the AR9.  Stacked 30” off the floor and baffle aligned with the AR9 baffle meant it would sound anemic in the bass, however human voices were clearer than I expected but the imaging was opaque and lifeless.  The sound was flat in front of the speaker with no dimensionality in comparison to the AR9.  After reconfiguring the system to convert the 12” woofer to an LMR and engaging the 1W, everything came to life. 

Near the end of today’s session, I rechecked my ears by listening again to the full range 58s against the AR9 to see if I had deluded myself.  I had not.

I would guess these results for the 58s would be the same for an AR91 and AR92.

 

[Aadams]

The AR581w has been disassembled and reconfigured to create a comparison of the AR3a and the AR9 with the AR3a configured as a 358s.  The two systems have been playing side by side for several hours to compare the character of the sound from 4 to 8ft, basically an imaging comparison.

The short version is, the AR3a yields good performance in imaging, comparable and, at times equal, to the AR9 but, only after a lot of fiddling with speaker position and Lpads.   

The setup was identical to the 581w test in the previous post with the only difference being the 3a and 58 swapping roles.   

The AR3a pair has rebuilt tweeters, Lpads with resistors, and new npe caps.  

All attenuators were initially set to 0 but the 3a mids had to be turned down slightly to get a timbral match with the nines and then rebalanced between the speakers to center the image. The 3a also had to be turned toward the listening position to produce its best image. 

The AR3a, in this configuration, at these listening distances, compares very well to the AR9 in all respects except (1) the AR9 mid, which extends strongly to 7500hz, occasionally extracts a bit more ambience on some live venue recordings, compared to the 3a mid/tweeter combo covering the same range and (2) the amount of futzing around required to achieve max imaging performance when the AR9 requires near zero effort, by comparison. 

Honestly, without a good imaging speaker with which to compare I am not sure I could set up a any AR classic to its max imaging potential. 

Adams

[Aadams]

On 5/30/2019 at 11:18 AM, ligs said:

Even with 3 db adjustment in the mid, you still favor AR5 over AR58?

Ligs

You asking this question has led to a big shake up in my speaker world.  The quick answer is I no longer prefer the 5 to the 58s but for a reason that was unexpected.

Once I realized how similar the 58s and 9 could sound, I decided to make the 58s a permanent top and take it out of woofer service.   When it came to readjusting the attenuators on the classic models, the lpads and pots on the 3as were able to meet the increased output levels required to fit into the new setup but the AR5 pots are too weak and unreliable near the upper limit so I took them out of service until I get time to install Lpads and resistors. 

The 3as are now stacked as an AR31w configuration permanently next to the 9s and the 58s is the top for the woofers in the 98 cabinets.    

If you hadn’t challenged me the potential of the 58s would not have been revealed. and this improvement would not have happened.

Thanks