AR2ax (Early) Tweeter Dome Suspension

[ReliaBill Engineer]

For an early model 2ax, I’m looking at these phenolic domes. The domes are relatively heavy, compared to a silk dome, treated silk dome, or thin polyamide dome, or even a titanium or beryllium dome. So I expect a stiffer suspension for this heavier dome. But these well-aged dome suspensions seem to have hardened and stiffened beyond their initial compliance. 
 

In looking at the suspension, it appears the 4 yellow globs are both a suspension and damper. Then over the top of it is a clear flexible sealant, that has hardened.  The voice coil is attached to the underside of the dome, in the magnetic gap.

 

Am I correct? Anyone in here willing to share their knowledge here? 
 

Im pretty sure that my tweeters are being restricted in motion. While I’m getting output, I don’t think I’m getting the full output that was available when new. 
 

I’ve rebuilt many vintage cartridge styli, from cantilevers, to tips and suspensions. I’ve also rebuilt tweeters using various parts I’ve found on eBay. I’ve formulated various room temperature curing elastomers that I use to form new cantilever suspensions for cartridge styli. I’m wondering if I should attempt to rebuild these tweeter suspensions.

 

[ReliaBill Engineer]

Perhaps I’ll have to be the pioneer here…

[Aadams]

Not interrogative but rhetorical:  What is your backup plan if you damage the voice coil or the phenolic dome?  I can't recall reading that either are practically replaceable. Seems like you would want a "training" tweeter.

[ReliaBill Engineer]

My backup plan is to replace the tweeter. I’ve got many in inventory to choose from, including a couple of Dynaudio 28 mm soft domes. 
 

But, I’ve done a lot of small work like this before. Fact is, my rebuild of a Shure VN35E stylus, and an (Micro Acoustics) MA 309, and Shure VN5MR and VN5HE styli was more demanding than this. 
 

Of course, I’d rather keep these original tweeters. So here’s my plan:

Using a 10x loupe goggles and powerful LED headlight, use a razor to cut the sealant around the dome, except for where the yellow blobs are. Scrape away the sealant from the base of the dome, but not ON the dome. Then apply my own elastomer “blobs” next to the original yellow blobs (X 4). Let it cure. Then cut away the yellow blobs. Then apply a clear sealant I’ll make from thinned clear silicone. A very, very thin coat. Let it cure. This way I won’t come close to the VC windings, nor mess up the centering and spacing of the original dome or windings.

Im actually a little surprised by how non-precise these domes were originally assembled. The yellow blobs are all different sizes, and the sealant looks to be painted on by hand.

Edited March 27, 2023 by ReliaBill Engineer
More info.

[RickB]

RoyC on here does repair those, his thoughts would be very interesting.

[RoyC]

This tweeter, which is original to the AR-2ax and AR-3, was designed to cross over at a very high frequency. The stiffer suspension is deliberate, and meant to minimize the dome's excursion to limit its effect on the mid frequencies. It should also be noted that the original tweeter is not only mechanically unique, it is electrically unique, with an unusually low DCR of under 2 ohms. If a modern tweeter (regardless of brand or quality) is used to replace it, the crossover must be modified to prevent it from creating a rather imbalanced and harsh sounding mid response.

A modern tweeter is really a different animal. In other words, don't mess up the original if at all possible. It is not just a simple matter of attempting to fit an aftermarket replacement tweeter into the unusually large cabinet hole.

The clear (butyl rubber) material around the perimeter of the dome can be peeled away mechanically with tweezers. Great care must be taken, however, not to damage the fragile aluminum leads or the suspension "blobs".

Roy

[ReliaBill Engineer]

Thanks for the heads up! 

Replacing the tweeter would be an absolute last resort. But if I should, I have several great candidates. But the impedance would require a new XO design, or at least a new value of capacitor. The tweeter output, as is, is woefully inadequate. It’s very restrained, even at max setting.

I had already known the heavier dome requires a stiffer suspension, and at 5 kHz+ the displacement was small. But I doubt the clear sealant is butyl. It’s hard as a rock on my tweeters. Butyl rubber doesn’t usually fail in that mode. Butyl rubber is what I’m going to use for the 4-point suspension. But a softer, more compliant mix. I may try a compound which uses micro encapsulated silicone grease; it has compliance yet damps extremely well. I’m going to try a thinned clear silicone for sealant. Membrane thin. It’s stable and won’t harden.  (I used to work for Goodyear Tire and Rubber as a technical engineer with the compounding engineer) Butyl rubber is used for the inner case of tires; it’s stable and doesn’t react to oxygen or nitrogen, it’s heat stable, and is stable under constant flexing.

I’ll do before and after vids of this track. This is before:

https://youtu.be/I9tDWEOLOgs

[JKent]

Hmmmmm. Roy said the tweeter crosses over at VERY high frequency. Bill said the tweeter output is woefully inadequate. I know my EARS have become woefully inadequate at very high frequencies as I age. Maybe it’s not the tweeter 😉

[ReliaBill Engineer]

2 hours ago, JKent said:

Hmmmmm. Roy said the tweeter crosses over at VERY high frequency. Bill said the tweeter output is woefully inadequate. I know my EARS have become woefully inadequate at very high frequencies as I age. Maybe it’s not the tweeter 😉

I tested the “ear” theory last night. Like you said, I needed to verify whether my ears were fooling me, or the tweeter output was low.

Using a frequency generator app on my phone, I injected signals from 1kHz to 15kHz into my AR speaker. At 8kHz the sound from the tweeter was barely audible. At 9kHz and above, the sound disappeared.
 

So I rotated the balance knob on my amp from right to left. Amp is in mono mode, to ensure L and R speakers see the same signal. AR speaker on the right channel, my Polk on the L channel. As I went up in frequency (1kHz steps) from 1kHz to 15kHz, my Polk was very audible up to 13kHz, the limit of my hearing. 
 

On the AR, the XO does seem to be at 5kHz, as I had output from the tweeter. But its output dropped precipitously after 8kHz. 
 

This should be an interesting project going forward. My wife loves the Polks, the 1989 RTA-8T, and is losing patience with my fiddling on the “old” AR speaker. What she doesn’t understand is that I hear greatness from this AR. But it can’t be realized with the tweeter not producing in the highest octaves. I know they’ll sound great once I get the tweeter working to its full potential. Without the upper octaves, the speaker will suffer from an apparent lack of “air”, detail, and balanced sound.

Edit: As an added tidbit, my father’s 1966 Wharfdale W70D speakers were his “cat’s meow.” He was very proud of them, bought new in ‘66. But they suffered from a beaming quality, and a sizzling paper cone tweeter with a felt surround. Already these little AR speakers have a promise of a more open, more spacious and detailed sound. My dad is still alive. In talking to him, he said he was sold on the “British” sound, and his Garrard turntable with Shure M44 cartridge. He considered the AR speakers, but got a “good deal” on the Garrard Lab 80 table and Wharfdale speakers….

[RoyC]

AR is known to have used butyl and (later) latex rubber compounds for this type of purpose...but the type of compound is irrelevant. These were manufactured in the mid-60's, and AR was the first to try various rubber and foam formulations.

If you scrape it, you will find it to be elastic and not "rock hard"...unless somebody applied some kind of additional material over it at some point. With that said, there is no doubt that it has stiffened with time, and the primary problem has as much to do with its initial degree of migration into the voice coil gap. If you are able to get it started in a "safe" area of the dome's perimeter, you can literally pull and peel it out. I have worked on many of these tweeters and have done this countless times.

Before doing anything, try listening to the tweeter with no cap in line vs with. You will likely be surprised at the subjective difference. The low DCR of this driver is as instrumental in its reticent response as the material around the dome. Unlike modern designs (like your Polks), these tweeters are only meant to enhance "sss" sounds, and little else. Your impressions are not unusual for folks new to early AR-dom.

I'm sure you can improve the response with your plan, but don't expect it to be a dramatic difference. Modern speakers simply have more upper mid and high frequency output.

Roy

[genek]

Sounds like a job for a supertweeter.

[JKent]

6 minutes ago, genek said:

Sounds like a job for a supertweeter.

And it is. I believe the 2-ax “mid” is a tweeter and the driver in question is a supertweeter. 

[ar_pro]

38 minutes ago, JKent said:

And it is. I believe the 2-ax “mid” is a tweeter and the driver in question is a supertweeter. 

If it's desired, a supertweeter will provide more "air" for an AR system - but in the far-field, I think it's not that noticeable.

[tripleguy]

With 2 speakers in the system, in stereo, I find the sound very pleasant. Yes, it's a bit surprising at first how much high freq the "midrange, carries. Cymbal freq is very wide, from 600-800 Hz up to 10kHz before disolving into harmonics. On an EQ mixing board, you would typically boost 6kHz to 8kHz freq. Sizzle is often not really desired so anything after 10kHz is generally dropped. The 2ax tweeters (old and newer) would or should do well inside of those parameters. 

[genek]

1 hour ago, JKent said:

And it is. I believe the 2-ax “mid” is a tweeter and the driver in question is a supertweeter. 

Yes, but it's a supertweeter that produces vintage AR highs. If someone wants to boost highs to modern levels, another one will be needed.

[JKent]

You’re right Gene. That’s why MicroStatic sold those add-on supertweeters specifically for AR and KLH. 

[ReliaBill Engineer]

20 hours ago, RoyC said:

AR is known to have used butyl and (later) latex rubber compounds for this type of purpose...but the type of compound is irrelevant. These were manufactured in the mid-60's, and AR was the first to try various rubber and foam formulations.

If you scrape it, you will find it to be elastic and not "rock hard"...unless somebody applied some kind of additional material over it at some point. With that said, there is no doubt that it has stiffened with time, and the primary problem has as much to do with its initial degree of migration into the voice coil gap. If you are able to get it started in a "safe" area of the dome's perimeter, you can literally pull and peel it out. I have worked on many of these tweeters and have done this countless times.

Before doing anything, try listening to the tweeter with no cap in line vs with. You will likely be surprised at the subjective difference. The low DCR of this driver is as instrumental in its reticent response as the material around the dome. Unlike modern designs (like your Polks), these tweeters are only meant to enhance "sss" sounds, and little else. Your impressions are not unusual for folks new to early AR-dom.

I'm sure you can improve the response with your plan, but don't expect it to be a dramatic difference. Modern speakers simply have more upper mid and high frequency output.

Roy

Roy, you are correct in every aspect! The sample that I scraped from my tweeter was a drop of grill glue that managed to get dropped near the tweeter perimeter, just on the edge of the elastomer sealant. Boy! Do I feel embarrassed!! I was being *too* careful in my “sampling” of the sealant.

Yup! I am learning much about AR-dom here. 
 

So….in your trials and travels, are these phenolic domes capable of more output by modifying their suspension, or renewing it? Or in other words, what is your experience in their change in output after *you’ve* changed out their suspension?

I haven’t yet molested the original tweeter. I’m being very cautious…

I’m going to attach a 1-minute clip of the speaker here. Only the caps replaced and pots cleaned. Tweeter at full output on its pot, midrange at 65% on its pot setting. I move my phone’s mic to within 1” of the tweeter. You can hear the tweeter’s output. Playing Bert Kaempfert’s “Wonderland By Night” from the 1960 Stereo LP. _________Everyone let me know what you think. Does this seem normal for this tweeter?

 

 

[genek]

19 hours ago, JKent said:

You’re right Gene. That’s why MicroStatic sold those add-on supertweeters specifically for AR and KLH. 

Most modern supertweeters are not going to provide the added "highs" that fans of modern speakers are after, because the sound most modern listeners are missing when they say ARs have no "highs" is actually mostly in the upper midrange (2kHz-4kHz) and presence (4kHz-6kHz) regions. Aperion ribbons, for example, don't kick in until 8kHz. By comparison, the Microstatic array can be set to go as low as 2.5kHz.

Longtime readers of this forum may recall that our late friend Carl Richard (Carlspeak) designed an "updated" version of the 3-series called the "Super 3a," that replaced the original dome mid and tweeter with SEAS drivers. Maybe what's needed here is a "Super 3t," with the HiVi Q1R and one of HiVi's dome midranges in a small box.

[tripleguy]

Its very hard to tell in the video Bill. The mid carries so much of the lower treble in those speakers. You'd almost need to kill the mid and take woofer out of circuit but the tweet isn't going to produce a lot of sound. I still think killing mid and doing a 6K-10K sine wave sweep will 1) confirm output and 2) demonstrate how little is there in freq much above 9K. On mine (smaller tweeter version) I have to be up close to the tweeter to hear anything above 9K. My ears start to roll off there tho. I'd need headphones to hear much higher. My friend has restored AR3s with the phenolic dome tweet (rebuilt). I can check his next time Im there.

One other thought: might be an issue with the mids since they carry a lot of high freq. My friend's AR3 mids needed a rebuild. Sound was muffled even after I changed pots and caps.

[Aadams]

The domed tweeters were added to prop up the power response of the mid/tweeter 2 series.  The domed tweeters were never intended to be on-axis spl demons. The graphs below may help.  The tweeter is the later 3/4 inch but the effect is very similar with the 1 inch tweeter.  The graph shows he mid tweeter of the 2ax responding above 10k on axis. AR6 on the right hand side.

 

[ReliaBill Engineer]

Pretty sure I figured out the problem. I don’t think my ears were lying to me. Output is greatly reduced. Looking closely at the tweeter dome, it was messed up at the factory 58 years ago. 
 

In these pics, on the right side vs the left, the voice coil is wedged in the gap; halfway out on the right side and flush on the left, and glued that way. Gently pushing down, there is no movement.

Below, VC windings visible near leads:

 

 

 

 

Below, windings visible:

 

Below, other side of dome is flush with mounting plate:

:

 

[ReliaBill Engineer]

So now I feel I have no option except to remove the dome. So I did. I’ll have to carefully clean the windings and the motor gap. Then put the damping back in and replace the suspension and sealant. 
 

Sealant removed:

The dome freed:

 

The damping batting inside the dome:

 

Dome completely freed:

 

 

[ReliaBill Engineer]

I spent last night formulating my plan for repair, the sequence and materials I’ll use. I tested using a silicone membrane for the sealant. Also the 4-point suspension. Hopefully I can get this done over the weekend.

[ReliaBill Engineer]

Anatomy of the early 2ax Dome Tweeter:

 

This dome tweeter has a unique build. The cloth-phenolic dome is light, although heavy by modern soft dome standards. It uses enameled aluminum wire on the voice coil, which is why you don’t see solder on the brass binding posts below the tweeter’s mount; the wires are twisted on the posts.

The suspension assembly is very unique. When in the voice coil gap, the stack of batting inside the dome supplies an upward (outward) spring pre-load to the dome. Of course, it also damps vibration and kills sonic energy inside the dome. The 4 liquid-foam deposits (yellow, applied as a foam liquid during assembly, then allowed to cure) anchor down the dome, against the pressure of the dome’s internal batting. The faceplate has 4 cutouts for anchoring the foam away from the voice coil. Then, a thin, sticky, clear elastomer sealant is applied on top of the foam, and on both the perimeter of the dome, as well as on the perimeter of the face plate. This elastomer sealant is allowed to cure. When all is cured, the dome is suspended by the batting that pushes the dome outward, and held in place by the 4 dollops of yellow foam and the clear elastomer membrane. 
 

Below, the polyester batting that is compressed below the dome. It provides damping and spring force outward:

 

 

The four foam cutouts in the faceplate for anchoring the liquid foam:

 

Below, the four foam suspending anchors to “float” the dome:

 

Below, with the clear elastomer sealant applied, that forms a dust seal, and a resilient retaining membrane, to allow the dome to achieve pistonic motion when a music signal is applied. 
 

[ReliaBill Engineer]

I got the gap cleaned and the voice coil cleaned. Pretty obvious that AR allowed the glue to get into the gap and on the voice coil, pretty much glueing the foam at the 4 points and on the coil windings. Very little clearance between the center magnet and inside edge of the coil.

Besides all of that, the stack of polyester batting inside the dome puts a VERY substantial preload on the dome. I put a HEAVY steel washer on the dome, and it isn’t enough to force the dome and voice coil down into the gap!

I tried a 1.2 volt battery on the leads, to center the voice coil…. It expelled the dome with tremendous force!  I won’t do that again.

The batting:
 



Using the heavy steel washer. I need a biggerweight!