AR-4x Tweeter Dissection

[Pete B]

I have a blown AR-4x tweeter here that measures open circuit and I'm going to

document at least some of the physical characteristics here. This was donated

by mrBruce4, thanks Bruce!

I was hoping that perhaps it had a buzz problem so that I could measure the

T&S parameters at a low level before the buzz but it is open circuit so this is

not possible.

This treated paper cone tweeter is essentially a scale down of an acoustic

suspension woofer, where the frame is sealed in back forming a chamber that

is filled with white fiberglass. The fiberglass actually touches the paper cone

with some pressure required to compress it into position.

The cone appears to be curvilinear or at least somewhere between conical and

curvilinear as it has a shallow profile. This helps to provide a flat response.

The edge of the cone lifted easily, not sure if this is because Bruce losened it

up or if the glue has simply dried out. I can lift up the cone and see the

fiberglass, but it is difficult to do much else. I'd like to desolder the lead in wires,

losen up or cut the spider if it has one, and pull out the cone, but I can hardly get

in there to work on it.

I'm about ready to cut the cone close to the dust cap, watching for the VC wires.

I was hoping to do this in a way where I might have a chance of repairing the

lead in wire and reassembling the driver for T&S measurements. Anyone know

how these were assembled? Does this driver have a spider behind the white

fiberglass? Anyone know the DC resistance of this driver?

It seems that the small cone tweeter was fairly popular in the 1960s and is

even copied today, this Pioneer tweeter looks very similar. Obviously, we do not

know if they have similar parameters or not without measuring them:

http://www.partsexpress.com/pe/showdetl.cf...tnumber=270-032

This Peerless cone tweeter was very common back in the day and it did also sound

very good: http://www.roger-russell.com/driverst.htm

[RoyC]

The edge of the cone lifted easily, not sure if this is because Bruce loosened it

up or if the glue has simply dried out.

Anyone know the DC resistance of this driver?

Hi Pete,

The cone edge glue is a common problem these days. DC resistance is typically just over 4 ohms. When face down, the right terminal is + (yellow wire, if still connected), which is reversed relative to the woofer in the crossover.

I have never dissected one, so I don't have any other info.

Roy

(drawing courtesy John O'Hanlon, blue wire in tweeter circuit is sometimes green)

[mrbruce4]

Thanks for the research/writeup Pete.

I hadn't done anything to the edge of the cone, so I expect it is the common glue problem that John mentioned.

Very glad that the dead "body" has gone to research!

[ARRAY]

I have a blown AR-4x tweeter here that measures open circuit and I'm going to

document at least some of the physical characteristics here. This was donated

by mrBruce4, thanks Bruce!

I was hoping that perhaps it had a buzz problem so that I could measure the

T&S parameters at a low level before the buzz but it is open circuit so this is

not possible.

This treated paper cone tweeter is essentially a scale down of an acoustic

suspension woofer, where the frame is sealed in back forming a chamber that

is filled with white fiberglass. The fiberglass actually touches the paper cone

with some pressure required to compress it into position.

The cone appears to be curvilinear or at least somewhere between conical and

curvilinear as it has a shallow profile. This helps to provide a flat response.

The edge of the cone lifted easily, not sure if this is because Bruce losened it

up or if the glue has simply dried out. I can lift up the cone and see the

fiberglass, but it is difficult to do much else. I'd like to desolder the lead in wires,

losen up or cut the spider if it has one, and pull out the cone, but I can hardly get

in there to work on it.

I'm about ready to cut the cone close to the dust cap, watching for the VC wires.

I was hoping to do this in a way where I might have a chance of repairing the

lead in wire and reassembling the driver for T&S measurements. Anyone know

how these were assembled? Does this driver have a spider behind the white

fiberglass? Anyone know the DC resistance of this driver?

It seems that the small cone tweeter was fairly popular in the 1960s and is

even copied today, this Pioneer tweeter looks very similar. Obviously, we do not

know if they have similar parameters or not without measuring them:

http://www.partsexpress.com/pe/showdetl.cf...tnumber=270-032

This Peerless cone tweeter was very common back in the day and it did also sound

very good: http://www.roger-russell.com/driverst.htm

Once upon a time I was a loudspeaker technician and somewhat recall the AR4X tweeter. In short, it's nothing special and you can replace it with much better. I've read a tonne of posts all over the internet about why the AR4X, and other drivers, let go at the basket/cone glue joint and here's why. It has to do with the kind of adhesive used. The baskets are made from steel and steel can rust if not properly protected. The adhesive itself has a property which wicks moisture from the air. This action eventually causes corrosion at the interface and since not much can stick to rust, the joint fails. Many speaker baskets use a better paint grade and application process so this problem doesn't happen. If you further dissect the tweeter, you'll find that packed in the aluminum former you'll find silicone grease placed there for resonance & temperature control. It's safe to say it does have a spider, but it's been so long I can't remember. The small magnetic drive, aluminum former, large cone area, spider and paper cap all make for lacklustre high frequency performance.

There are few good cone hi-fi tweeters out there not to say they aren't good performers, only because everyone wants domes, ribbons or something else exotic. Hi-Fi parts manufacturers don't waste much of their time on cone tweeters anymore.

[Pete B]

Thanks ARRAY I got it open and everything agrees with your statements.

I peeked in under the cone and could see one of the lead in wires. Knowing the path

they took, I used a tookpick (non metallic and handy) to get under the lead in wire

after having heated the solder joint at the input lug. One came out, the other seemed

to already be broken, this was the cause of the open circuit. I decided to simply pull

evenly on the cone, hoping for the glue to break, and sure enough after a few tugs

it came out. The glue joint broke cleanly at the spider. There is a small spider in there

and the glue around the outer edge was still holding strong. Looks like white PVA glue

and I decided to try alcohol to losen it up. Took about 3 applications but it certainly did

soften quickly, I scraped it away with a toothpick. The alcohol has to be applied several

times as it evaporates quickly. If I was to do another, I'd slip an eye dropper in under

the cone and let it drip down into the spider glue joint, soak, then pull.

There is an aluminum voice coil former which is good because it is only .5" in diameter,

which does not offer much thermal capacity, and the full lenght of the former is .375".

The length to the back of the cone is .25 to .3" as it was not evenly glued, or the cone

was not cut evenly. The VC wind height is .1" and the gap height is also .1".

There is a jelly at the top of the pole piece, looks exactly like milky/clear Bacitracin,

but is probably silicone grease as ARRAY mentions above.

There is a white fiberglass ring between the cone and the frame. Glue holding the edge

down had made its way onto the cone and joined the cone to the fiberglass in a very

small area toward the outer edge.

The dust cap is about .65" and it has about a 1/8" layer of fiberglass damping pressed

into the back, which seems to have absorbed some of the grease. The dust cap fell

off with very little pressure, probably due to the old glue and the grease that was

absorbed.

Here are some of the physical parameters in tabular form:

Date: JAN 3, 1967

Tweeter OD: 4.5"

Cone diameter: 2.05"

Suspension OD: 2.25"

VC former material: Aluminum

VC former ID: .5"

VC former OD: .55"

VC former length: .375"

VC wind height: .1"

Gap height: .1"

Front plate thickness: .1"

Magnet thickness: .275"

Back plate thickness: .165"

Dust cap diameter: .65"

Magnet weight: 2 oz (can anyone confirm this?)

Magnet type: Ceramic Ring

A few of the windings were loose at one end, however they are easily pushed

back into place and could probably be repaired with a very small amount of high

temp epoxy glue. The lead in wires are where the open circuit occurred and

the best solution would be to use a light gauge tinsel type of wire soldered to

the remaining VC wire and epoxy glued to the VC former.

My best guess at how these were originally assembed is as follows:

I believe that the cone voice coil and spider are first assembled and glued

probably on a jig of some type. A bead of glue is placed where the spider

contacts the frame, then the assembly is probably installed with shims between

the pole piece and voice coil former and the glue allowed to try. The cone edge

is also probably glued down at this time while the shims are in place. The shims

are removed and the dust cap is then glued in place.

This driver could probably be repaired, but it is a lot of time to put into such a low

cost item. It would be better to examine a modern tweeter of similar construction

to first see if it is of similar construction and motor geometry. Then measure and

compare to determine if one might be suitable as a replacement, perhaps with

crossover mods.

There's probably more value in rebuilding the .75" 2ax and 3a tweeter than this low

cost unit.

[dynaco_dan]

Thanks ARRAY I got it open and everything agrees with your statements.

I peeked in under the cone and could see one of the lead in wires. Knowing the path

they took, I used a tookpick (non metallic and handy) to get under the lead in wire

after having heated the solder joint at the input lug. One came out, the other seemed

to already be broken, this was the cause of the open circuit. I decided to simply pull

evenly on the cone, hoping for the glue to break, and sure enough after a few tugs

it came out. The glue joint broke cleanly at the spider. There is a small spider in there

and the glue around the outer edge was still holding strong. Looks like white PVA glue

and I decided to try alcohol to losen it up. Took about 3 applications but it certainly did

soften quickly, I scraped it away with a toothpick. The alcohol has to be applied several

times as it evaporates quickly. If I was to do another, I'd slip an eye dropper in under

the cone and let it drip down into the spider glue joint, soak, then pull.

There is an aluminum voice coil former which is good because it is only .5" in diameter,

which does not offer much thermal capacity, and the full lenght of the former is .375".

The length to the back of the cone is .25 to .3" as it was not evenly glued, or the cone

was not cut evenly. The VC wind height is .1" and the gap height is also .1".

There is a jelly at the top of the pole piece, looks exactly like milky/clear Bacitracin,

but is probably silicone grease as ARRAY mentions above.

There is a white fiberglass ring between the cone and the frame. Glue holding the edge

down had made its way onto the cone and joined the cone to the fiberglass in a very

small area toward the outer edge.

The dust cap is about .65" and it has about a 1/8" layer of fiberglass damping pressed

into the back, which seems to have absorbed some of the grease. The dust cap fell

off with very little pressure, probably due to the old glue and the grease that was

absorbed.

Here are some of the physical parameters in tabular form:

Date: JAN 3, 1967

Tweeter OD: 4.5"

Cone diameter: 2.05"

Suspension OD: 2.25"

VC former material: Aluminum

VC former ID: .5"

VC former OD: .55"

VC former length: .375"

VC wind height: .1"

Gap height: .1"

Front plate thickness: .1"

Magnet thickness: .275"

Back plate thickness: .165"

Dust cap diameter: .65"

Magnet weight: 2 oz (can anyone confirm this?)

Magnet type: Ceramic Ring

A few of the windings were loose at one end, however they are easily pushed

back into place and could probably be repaired with a very small amount of high

temp epoxy glue. The lead in wires are where the open circuit occurred and

the best solution would be to use a light gauge tinsel type of wire soldered to

the remaining VC wire and epoxy glued to the VC former.

My best guess at how these were originally assembed is as follows:

I believe that the cone voice coil and spider are first assembled and glued

probably on a jig of some type. A bead of glue is placed where the spider

contacts the frame, then the assembly is probably installed with shims between

the pole piece and voice coil former and the glue allowed to try. The cone edge

is also probably glued down at this time while the shims are in place. The shims

are removed and the dust cap is then glued in place.

This driver could probably be repaired, but it is a lot of time to put into such a low

cost item. It would be better to examine a modern tweeter of similar construction

to first see if it is of similar construction and motor geometry. Then measure and

compare to determine if one might be suitable as a replacement, perhaps with

crossover mods.

There's probably more value in rebuilding the .75" 2ax and 3a tweeter than this low

cost unit.

Hi Pete;

A very nicely done report, good job.

[Pete B]

Hi Pete,

The cone edge glue is a common problem these days. DC resistance is typically just over 4 ohms. When face down, the right terminal is + (yellow wire, if still connected), which is reversed relative to the woofer in the crossover.

I have never dissected one, so I don't have any other info.

Roy

(drawing courtesy John O'Hanlon, blue wire in tweeter circuit is sometimes green)

Thanks for the schematic Roy. 20 uF is large for a tweeter cap,

must be a rather low crossover point.

Do you know what the pot resistance is for the normal setting?

I don't actually have a pair of 4Xs but it is an interesting tweeter since it probably

lead the way for the large Advent type.

[RoyC]

Thanks for the schematic Roy. 20 uF is large for a tweeter cap,

must be a rather low crossover point.

Do you know what the pot resistance is for the normal setting?

I don't actually have a pair of 4Xs but it is an interesting tweeter since it probably

lead the way for the large Advent type.

Pete,

The pot's "dot" setting is about 4 ohms in series and 11-12 ohms in parallel (depending on the pot), although a setting near full "increase" is generally more satisfactory to me. The woofer's inductor is .88mh (#4) for earlier versions, or 1.18mh (#5) for later versions, and I believe the crossover is considered to be 1200hz to 1400hz.

Kent mentioned you are dissecting the typical KLH cone tweeter as well. I wonder if that one would be closer to the Advent tweeter, given Henry Kloss's role in both KLH and Advent. Any comparison with the 4x tweeter would prove interesting.

Keep us posted...

Roy

[dynaco_dan]

Pete,

The pot's "dot" setting is about 4 ohms in series and 11-12 ohms in parallel (depending on the pot), although a setting near full "increase" is generally more satisfactory to me. The woofer's inductor is .88mh (#4) for earlier versions, or 1.18mh (#5) for later versions, and I believe the crossover is considered to be 1200hz to 1400hz.

Kent mentioned you are dissecting the typical KLH cone tweeter as well. I wonder if that one would be closer to the Advent tweeter, given Henry Kloss's role in both KLH and Advent. Any comparison with the 4x tweeter would prove interesting.

Keep us posted...

Roy

Hi Roy;

I was trying to find any notes that I have regarding the AR-4/4X speakers.

Can't find any or those were damaged in the floods.

I have downloaded to my computer a photo of an old ad, possibly from an old Radio Shack catalog, but the resolution is so bad, I can only see that the AR-4's sold as low as $51.00 US.

I cannot read any written details as to crossover, etc..

I believe that the AR-4's had a higher crossover point than the 4X's, perhaps 1,800 hz.

I never saw an introductory sheet of the AR-4's.

The AR-4X's have a 1,200 hz crossover point.

Nice that you know the two different coil values.

I wonder why and when?

Perhaps when they changed from the Alnico to ceramic woofer magnets.

If I had continued on with revising the AR Chronological list, I would have posted at least these two different coil variations and serial numbers to try to identify timelines.

If the coil changes were not related to the woofer changeover, then I would have listed 4 variations and serial numbers.

As collectibles, we would probably want to know that our cabinets and contents are matching, well as near as possible.

I believe that Jeff is still working on revising it.

[RoyC]

The AR-4X's have a 1,200 hz crossover point.

Nice that you know the two different coil values.

I wonder why and when?

Perhaps when they changed from the Alnico to ceramic woofer magnets.

Hi Vern,

The crossover point would have to be a bit different with a different woofer inductor and the same drivers.

John O'Hanlon and I have been comparing notes for some time, and John is compiling a nice data sheet based on 4x serial numbers. It appears that the inductor change was around 1970...which is also about the time the 3a and the 2ax were also provided with larger woofer inductors. Cabinet stuffing type (to yellow shredded fiberglass) and amount (reduced) also seems to have changed around that time for all of these models as well.

Although there were a number of varieties of the 4x alnico magnet woofer, the inductor change does not yet appear to be related to the differing cone types. I'm not sure that the 4x ever had a ceramic magnet woofer, though it did have a foam surround version in its last days. Either way, the inductor change was much earlier than that.

Roy