Rich W Posted May 2, 2016 Report Share Posted May 2, 2016 As a big fan of AR's 8” speakers, I read with great interest Rob Holt's modification of a pair of AR-4xa's using drivers from an AR-18. That got the wheels turning in my head . . . . Over the last few years, I've accumulated three pairs of AR-4x's from various sources. It is a superb speaker, echoed by many here on classicspeakerpages. It's one shortcoming, however, is that certain roughness in the mid to upper midrange. Speaker Dave addressed this problem in his fascinating post in the Mods and Tweaks forum, where he did series of crossover modifications to a pair of 4x's. I decided on a different approach, based on my experience with a restored pair of AR-7's, and some surplus AR-7 tweeters I happen to have on hand. My veneration for the AR-7 is well documented on this site. It has a way of sonically "disappearing" in a room, so that all you're aware of is an evenly balanced sound field with an incredible sense of depth, specifically that the music seems to emanate from a place behind the speakers. I attribute this quality to the Roy Allison designed 1-¼ inch tweeter (part number 200005-0), a design so successful that variations on it have appeared in many subsequent AR 8" speakers though the 70's and 80's. The first version of this tweeter, also used in the AR-6, AR-8 (and AR-4xa . . . read on!) has the virtue of retaining that wonderful classic AR laid back quality, with no hint of stridency or edginess. I suppose by now, you know where this post is heading - I decided to convert one pair of my AR-4x’s into AR-4xa’s. The AR-4xa has often been referred to as a “parts bin” speaker, intended to leverage off the pedigree of the AR-4x while offering “something new” to the buying public at a time when AR already had two great 8” offerings: the AR-6 and AR-7. It is however an excellent speaker in its own right, coupling the classic 8” woofer design (cloth surround in the early versions) with that superb sounding tweeter. The goal was to make my AR-4xa’s as close to stock in appearance and performance to the original as possible. Should the conversion prove to be a disappointment, I still had two pairs of 4x’s! The Crossover Relatively straight-forward. As it turned out, one version of my 4x collection had the correct #5 inductor required for the crossover modification. The 10 uF cap is a Dayton poly, a quality cap I had on hand. The first compromise with my “stock identical” approach was the tweeter level control. The 4xa crossover diagram calls for a three-position switch, and the existing 4x backboard cutout, while perfectly able to accept a switch, is designed for a pot and marked “increase/decrease”. Early history of the 4xa is a bit murky, but it appears that there was never an Aetna-Pollack type pot version – rather, from the beginning the 4xa always had a three-position switch. My solution was to use an 8-ohm l-pad, rather than the “correct” switch, utilizing just two of the three terminals, thereby functioning as a variable rheostat. The added advantage is of course an infinite range of settings. In practice, with my highly-damped room, I would set the control to maximum, with no attenuation of the tweeter at all. The crossover calls for wiring the tweeter in-phase with the woofer, unlike the AR-7, where the tweeter is wired out-of-phase. Here's the schematic from these pages: The Woofer While there has been some debate in these pages, early 4xa’s clearly utilized a woofer identical to the last version of the 4x woofer with a cloth surround. The 4x pair I used for the conversion was likewise a late version, with ribbed cones and mesh covering the rear magnet assembly. The surrounds were somewhat porous, so I gave them an extremely thin application of automotive gasket sealer thinned with toluene. The Tweeter I already had a pair of front –wired 1-¼ inch tweeters purchased from our favorite auction site. My original intention was to have them as back-ups for my AR-7’s should I become over-zealous with the volume level and blow one out. As my AR-7’s are now relegated to small-room/late night listening service, I felt “safe” using the tweeters for my 4xa’s. One added bonus – the seller was thoughtful enough to include the oval-shaped terminal strips, definitely keeping in line with the early-version of the 4xa I was attempting to recreate. Cabinet Modifications The “new” 1-¼ inch is not a drop-in replacement for the old 4x tweeter. It is significantly smaller in diameter, and the four screws/T-nuts do not align. It would have been simple and expedient to place a thin plywood plate over the old hole with an opening to accommodate the new tweeter and hardware. This method has two drawbacks: 1) The completed conversion would instantly give itself away as a retrofit, and a somewhat sloppy one at that 2) Using an adaptor plate brings the tweeter dangerously close to touching the speaker grill. My solution was to re-size the existing opening. First, I fashioned a backplate of thin plywood and glued it to the interior speaker wall, directly behind the existing opening. This supports a “ring” intended to fit in the existing opening fashioned from several layers of thin plywood laminated with wood glue. The inner and outer circumferences of the ring were cut with a circle-cutting Dremel tool attachment. T-nuts were then mounted into carefully drilled holes within the ring. The ring then was glued to the supporting backplate. Note the gaps in the backplate to provide clearance for the flat side of the T-nuts, and potential protrusion of the mounting screws. Duct sealer putty was used to fill the gap between the ring and the old opening. Finally, a ring of thin of basswood cut with an x-acto knife compass was glued to the recessed area where the flange of the old tweeter was seated. I also cut an opening to pass the wires which would be soldered to the back of the terminal strip. Once wired, the terminal strip was glued to the speaker baffle. The entire baffle was given a coat of matte black paint. See photos below: Once the cabinet modifications were complete, I sanded down the speakers, finishing them with 400 grit paper. There were some minor corner dings, which were filled in with a mixture of JB Weld wood filler epoxy and Mixol tobacco dye (thanks, J Kent) and sanded to match the cabinet. Several coats of Watco Natural Danish Oil completed the finish. A word about the grills. They were in perfect condition but were badly discolored. Rather than using new grill material, I restored the existing grills by spraying a solution of diluted bleach first and then Oxyclean spray afterward. After rinsing, the grills were patted down with paper towels to dry (I didn’t want the masonite to be weakened from being wet too long) and now look as good as new. The grill badges were carefully polished with Noxon, with very little pressure as to not to remove the lettering. I completed the assembly in the typical manner described in these pages many times. Cabinet stuffing was reused from the donor 4x. Here are some photos of the finished speakers, and their placement in the room. The large cabinet adjacent to each speaker creates a “quasi- 2 pi” environment, helping to smooth out bounce and cancelation effects (same as bookshelf mounted speakers surrounded by books). So the moment of truth. How do they sound? Absolutely superb, and exactly what I had hoped for. Below are response curves for both left and right speakers. These curves would be considered decent using the traditional method of micing a speaker at a meter’s distance. The fact that they were taken within the reverberant field, approximately 16 feet in front of the speakers is truly remarkable (and reflects my years of fidgeting with speaker placement in this space). If you ignore the left channel (purple) dip at 148 hz, both channels measure +/- 5 dB from 45 hz to 7.5k hz (the dip and generally rougher response in the left channel is certainly due to interference effects from nearby furniture). Changing the RTA resolution from 1/24 octave to 1/3 octave, I had all the data I needed to flatten the response further with a 31-band equalizer. Note that the high-frequency roll-off is considered normal in the reverberant field and that very little boost was applied 8k and above. FR curves and view of room from where photo and measurements were taken. The sweet spot in the room is unfortunately not at the sofa! Listening to one of my desert island recordings, Sibelius Symphonies Nos. 5 & 7, Colin Davis and the Boston Symphony Orchestra (Philips, 1975), there is a palpable sense of Symphony Hall itself, and you can literally feel the air surrounding the musicians change in their anticipation of the final chords of Symphony No. 5. I wonder how much you would have to spend on a pair of modern speakers to get the same level of performance. Link to comment Share on other sites More sharing options...
This topic is now archived and is closed to further replies.