AR-3a Repair due to Woofer Suspension Adhesive Failure

[Ron Purcell]

AR-3a S/N 3A-11146 (1969) Woofer Repair May 2019

 

High distortion was suddenly noticed from the right woofer on music and also 400 Hz sine wave test tone.

 

On disassembly, the spider was found completely separated from the basket frame, and the outer cone surround was separated part way around the basket frame.  This was due to drying out and failure of the old adhesive.  It permitted the voice coil form to rub against the magnet core during motion.  The voice coil winding however was completely intact and measured 2.536 ohms DC.  The spider, the cone and its surround were completely intact.

 

Replacement woofers of a newer design were available, but were not identical in construction or impedance, also having higher free air resonance frequency.  They were for the newer AR models.

 

An “AR re-coning kit” was available from Simply Speakers in Florida, containing a new cone, voice coil, spider, dust cap, coil spacing shim, flexible wire and proper adhesives for the various parts with instructions.  Kit was ordered and received 5/17/19.  The cone surround was a foam type and narrower than the old cloth one.

 

The kit’s voice coil measured 3.777 ohms DC - too high to match the existing crossover network.  It was also about 1/2” longer than the existing voice coil.  Fortunately the existing voice coil and cone assembly were OK to keep using.  The kit’s cone was also lighter that the old one, lacking the rubber mass ring.

 

Since the existing spider and cloth cone surround were also OK to keep using, all that was needed from the kit were the adhesives and the voice coil spacer shim.  The cone and basket frame were marked at the top for later alignment before removing the cone assembly.  The existing dust cap had to be cut off for the spacer shim to go in the gap during reassembly, but was found still usable and glued back in at the end. 

 

Old adhesive was sanded off the spider, the outer surround and the basket frame.  Mating surfaces were cleaned with solvent.  The magnet gap was cleaned out with a card and masking tape.  The gap was found deep enough that the voice coil form would never bottom out before the cone hit the magnet.  The voice coil form can therefore never be damaged.

 

For reassembly, the kit's voice coil spacer shim had to be cut into three long strips and taped together in a tight triangle at one end to create a tapered shape, because the inside of the voice coil had a lip of adhesive that blocked the shim from going in or out past the voice coil otherwise. 

 

The kit’s MI-3035 adhesive was applied in a strip to the spider, outer surround and basket mating surfaces.  With the spacer shim’s three strips inserted equally spaced around the gap and the taper facing straight up, the cone was pressed down with alignment marks facing each other until the cone surround and spider were in firm contact with the basket frame.  A curved piece of copper tubing was inserted through the wiring terminal screen holes and pressed down to insure complete spider contact with the frame all around.  Adhesive was left to dry 24 hours with shim strips in place, and then shim strips were removed.

 

Initially the voice coil made a slight noise when the cone was moved in and out, but it stopped after exercising with a high excursion 14-Hz tone (free-air resonance).  Apparently it was a stray particle that went away.  The old dust cap was reattached using the kit’s MI-2000 black adhesive.  Cabinet sealing compound was reformed into a 35” long round “dowel” and applied to the inner edge of the woofer opening.  Silicone was applied to the back mounting edge of the woofer for easier removal next time.  Using two drill bits in opposite screw holes as alignment dowels, the speaker was reattached into the cabinet after reconnecting the wires.  Distortion was gone and pink-noise EQ was essentially unchanged.

Hopefully this will help others having an adhesive failure problem to avoid having to trade an outstanding woofer for a merely good one.  The voice coil impedance, the cloth surround and the cone mass ring are all unique to the original AR-3a woofer's performance and longevity, and should not be unnecessarily sacrificed.