ReliaBill Engineer

I decided it would be best to discuss this in here, rather than in the Acoustic Research section. This way it can be seen without potentially hijacking a thread, nor competing with genuine original AR tweeter discussions. In essence, when the original 4x tweeters blow, or are otherwise damaged and non-functional, many replace them with the Phenolic Ring Tweeter (PRT) sold in various outlets such as Parts Express, eBay vendors, Amazon, etc. However, while the PRT fits nicely, and is cost effective (around $20), it doesn’t perform the same as the original AR tweeter. The small AR-4x speaker is a nice sounding legacy speaker, despite its small size. A common failure in these is the tweeter. Out of pure curiosity I decided to buy a pair of these PRTs and have a look at them. They are reasonably well made. But a bit “harsh” in my opinion. Can these be made to be better sounding? Can they more closely mimic the performance of a genuine AR-4x tweeter? That’s what I want to find out. I may fail or succeed. Original 4x tweeter: PRT, stock: Typical listing:

I’m a bit more than surprised…. These PRTs have been offered up as “replacements” and “temporary replacements” for many years for the AR 4x tweeters. But yet, nowhere have I been able to find a representative FR and impedance plot for them. How is that possible? Again, if I’ve overlooked it, I apologize in advance. The AR tweeter FR plot:

FWIW…. I just bought a pair of 4x speakers. All intact, never worked on. So now we’ll see…. Supposedly both woofers work, tweeters intact, but may or may not work. Prime for repairing the originals (if required), and testing and comparing these PRTs.

I’ve been trying to find comparison FR and impedance plots of the PRT vs an original AR tweeter for the 4x. I either overlooked it, or I just can’t find it. Anyone know where they are?

I looked at that one, plus many more. Just based on appearance, they are very obviously not AR. There are a few paper cone tweeters that look a lot like the AR tweeter, but they all require pretty major mounting mods. I was going off of RoyC’s comments about the PRT sounding similar and not requiring major mods to mount in the baffle.

I think you’re being a bit too defensive, RoyC. The reasons I decided to experiment with this tweeter were already stated. But to recap, here are my reasons:

A smooth tweeter is a smooth tweeter. No “extreme” aberrations in the FR and a smooth, well controlled impedence curve, makes designing a crossover a breeze. I saw how much work Speaker Dave put into the 4x tweeter XO in an attempt to tame its response. (In combination with the woofer.) It would have been easier if the tweeter had a smoother FR response, don’t you think? My experience says “yes!” The mention of dome tweeters was only to illustrate a point. Not so much to state definitively the merits of one over the other. Pistonic motion was the point of the comparison I was making. Unconstrained (but damped) motion of a tweeter cone/dome provides the best chance for smooth FR output. Im still in the process of documenting my work on this tweeter. I haven’t even solidified my own process, yet alone listened to final results. So of course I haven’t provided any instructions to the readers. But in the end, it should be up to the readers to decide if they want to try it themselves, no? I think so. Yet already it seems a judgement has been made about the merits of doing this, and its level of complexity. Prejudged, if you will. I’ll finish this tweeter, then mount it for audition, and do some critical listening. I’ll compare it to the untouched unit. If it passes that test, I’ll have it tested with plotted results, mine and the original. I think that’s fair. BTW: I’m no quack. I’ve been doing this kind of work for many years. I was a product and equipment design engineer, engineering manager for 6 companies. Now as a reliability engineer for NASA SLS, it is my job to identify root causes of failures and non-comformance, predict and quantify failures, as well as design solutions. It’s very challenging in aerospace vehicles. But with audio components, a certain artistic flair helps.

Well, it wouldn’t fit under a “restore”, to the AR purists. Not using an AR dome, and not an AR voice coil would make it a Mod, no?

Ok. Maybe a place for Replacement Replacements? Mods to Replacements? You want me to start the “Mod To Replacement 4x Tweeter”?

My Amazon order arrived today. So tonight I hope to finish work on the tweeter. I removed my butyl roll edge. Looked great, but not the damping I needed. Didn’t Speaker Dave make his theory, progress and calculations before he published his schematic and “How To”, and even his listening observations? Pretty sure he did. This thread is about a replacement tweeter, not an original AR tweeter. So I’m not blaspheming an original AR component here. In a previous post I offered to send RoyC these modified PRTs for evaluation. And RoyC’s modified PRT was linked to previously in this thread. So I’m hardly doing anything or saying anything without precedent. So far what I’ve said/shown is what I’m doing, why I’m doing it, and final product is in progress. Testing will come later. That’s why I’m leaving one PRT untouched. While I don’t have a full suite of test equipment, I do know someone across town who does. Im a bit surprised at the hypersensitivity. I’m not in the market for selling my services or products, as others in here are. Once I’m done and satisfied with the results, I’ll freely pay to send these to RoyC for his observations and critique. It won’t hurt my feelings if he gives me a “👎”. Like you said, my saying they sound great carries no weight or credibility. As for the 2 midrange drivers, that’s a longer term project. They are both missing the original “guts”. No VC, no domes. So it’s more than just a restore; it’s a new fabrication. But according to you, when I’m done, it will be fruitless, since I don’t have the “street cred” to say they sound good and are viable repairs/replacements.

There’s more going on in this tweeter than you realize. First, RoyC said this PRT has a sound similar to the 4x AR tweeter, but some find it more harsh. That’s where I started. I also read through every post, every page in Tweaks And Mods of the 4X crossover efforts by Speaker Dave. That gave me some background. Other background I have is repairing and modifying countless tweeters over the last 40+ years, so I’ve seen many of the best out there. Also, I’ve taken apart, repaired, improved many phono cartridges; they are merely the reverse of a speaker driver. A driver converts an electrical signal into mechanical sound energy; a phono cartridge converts mechanical energy into an electrical signal faithful to the mechanical movements of the stylus. One is the reverse of the other, yet all the same principles apply. Also, I have dual degrees in mechanical and electrical engineering. After examining the PRT I see why it has distortion, sounds harsh. The cone is rigid-mounted. (The same problem an aged phenolic dome tweeter has in the 3a and 2ax speakers!) The clear dope on the outer edge of the PRT is way too stiff; the cone fights itself to move. It has a powerful motor (magnet and voice coil) so the voice coil is going to move while the outer edge of the cone cannot. That’s a recipe for distortion! It also has no damping for the back wave produced when the cone center moves. A tweeter is like any other driver or a phono cartridge; it’s a tuned motor system. The PRT has a motor (VC and magnetic gap) and a spring. The motor moves the cone. It moves the cone against a spring; the spring is the accordion spyder, a flat sheet of phenolic-treated gauze in the case of this tweeter. To get proper motion, the cone surround provides damping that resists bouncing of the spring-loaded VC. In this case, the surround of the PRT is STIFF and WAY over damped, to the point that the cone can’t move properly. The phenolic ring is an insulator for the VC wires. It allows running the VC wires on the surface of the mounting plate without electrically shorting the wires. AR used black electrical tape to achieve this on the early 3a and 2ax front-wired tweeters (and mids on the 3a). What I’m going to do is retune the surround on the PRT cone. Allow it to move, but match the damping of the surround to the spring force of the spider. It’s exactly the same as matching a shock absorber to the spring of a car’s suspension. In that case, a car will bounce uncontrollably on its springs when it rides over a bump in the road; the shock absorber allows one bounce, then return to neutral position. Same for a properly tuned speaker driver’s cone suspension. Thats why RoyC’s attempt to tame this PRT failed. He added his dope to the existing dope. Adding dope to a surround that is already too stiff won’t accomplish anything.

I’ve been working (designing, building, testing) crossovers since 1981. I wrote my EE thesis on the crossover I built in 1983 for my dad’s Wharfdale W70C speakers (1966). If you start with a driver with a smooth, well-behaved impedance and response curve, watch where your crossover frequencies and fundamental resonance are, it’s relatively easy. Back in 1983 I had to plot those curves by hand. Since the Linkwitz-Riley wasn’t out yet, I used a combined Butterworth-Chebyshev 2nd order 3-way response curve. But truthfully, these early ARs had/have very fundamental crossovers. Inductor on the woofer for low pass, high pass capacitor on the tweeter. The only more simple crossover than these is no crossover. Forgot to include these pics of the PRT tweeter. Installing the wool felt inside the cone, to absorb/damp any sound from the back wave. Also just a small amount of loose wool fiber. Below, the wool felt installed:

I’ve got materials coming to me via Amazon. If all goes well, I may buy 2 of the 4 ohm units and work on those. What I failed to effectively communicate to RoyC is that if these sound good in my Polks, they’ll also sound good in the 4X, or in any speaker, regardless of the crossover. Speaker designers and mfrs always start with a good sounding driver. Even they can’t make a silk purse from a sow’s ear. If they try, they will wind up spending lots of R&D time and increase the parts count in the crossover trying to tame a so-so driver. I think of the extra steps AR used to form the metal grill, add a fiberglass pad, then glue the grills onto the 2ax midrange cone driver, just to attempt taming its squawky nature from using a low cost driver.

Butyl rubber surround. Not fully cured until tomorrow. I did fire it up for a few minutes, to check it out. But the process for making the surround is too labor intensive. I’m going to try a different method.

Decided it’s a better fit here. This modified PRT sounds much smoother than stock. Below, stock PRT. The clear dope constrains movement of the tweeter cone. Surround removed (below): Below, after internal damping was added, and a new (very thin) butyl rubber surround was added:

Waiting for the surround to cure. I used wool felt as batting inside.

Germane to finding a replacement tweeter for the AR-4X: Not naive at all. I’ve listened to individual drivers for more than 4 decades. Every one of them, be it a tweet, mid or woofer, has a “characteristic” sound. For the interest of those using these PRTs, a partial dissection: I removed the cone surround: I didn’t expect to see it, but these tweets use a spider! The spider secures the dome to the rear metal housing, at the cone/voice coil interface. Also, there is no internal damping (batting) inside of the cone, in the void behind the cone. No felt, no fiberglass, no poly fiberfill. The cone cannot be easily removed from the VC gap without risking damage the cone. Hard to see, but in the pics below, one can see the phenolic treated “gauze” beneath the cone. I’ll be calculating the type, amount of damping, 1000Hz to 15 kHz. Then redesigning the surround.

“I bought 2 of the 8 ohm PRTs to play around with. I don’t know if I’ll have any more luck than you did smoothing them out. But it will be a fun learning experience. I’ll test them using my Polk speakers, since it’ll be easy to adjust the mounting, and I have impedance adjusting resisters, and I’m very familiar with the original sound of these Polks. Also, the efficiency matches as well as useable FR.”

Really? Are you just trying to be difficult and defensive? Or am I misunderstanding you? So….I guess after I get these sounding smooth, I’ll have to buy a pair of 4X speakers in which to install them. But even then you’ll be skeptical, regardless of what I say, show, or even record. Likely I’ll have to send them to you for auditioning. But I’m very hesitant to do that.

Relevant to the 4X? Aren’t these PRTs being used to replace original AR tweeters in the 4X? Haven’t you yourself, and others tried to make them more suitable as a replacement? Or did I misunderstand? Listening is always the final test. Test equipment has its place, but the final test is in the listening. Right? The Polk RTA-8T speakers I’ll be auditioning these tweets in are also 2-ways. They’ll make for a good test case for these tweeters. Sink or swim.

Not in the 4X. 1988 I restored a pair of McIntosh XR7 speakers.

I said this in an earlier post: “I bought 2 of the 8 ohm PRTs to play around with. I don’t know if I’ll have any more luck than you did smoothing them out. But it will be a fun learning experience. I’ll test them using my Polk speakers, since it’ll be easy to adjust the mounting, and I have impedance adjusting resisters, and I’m very familiar with the original sound of these Polks. Also, the efficiency matches as well as useable FR.”

Just got these in. Hopefully I’ll have time tonight to make a few changes. I bought 2. One will remain unmolested.

There were extensive tests done in the early ‘80s on tweeter pistonic motion. Laser interferometry was used to analyze various cone and dome tweeters, design and materials. Cone tweeters tended to have the highest distortion caused by vibration nodes; different velocities and phase from center to outer edge. Domes had much less; even less depending on material and surround used. Paper cone tweeters just cost less than well designed domes. Paper cone tweeters tend to be used on the bottom tier speakers of any given mfr, 1950 to now. (The paper cone mid on the 2ax is another example of price point compromise.) There are exceptions, of course. The paper cone tweeter I showed above in the wharfdale W70C was replaced with a modified dome in the W70D, but the W70C was an expensive iteration of the paper cone tweet. Even that didn’t work well. Poor dispersion and a harsh character to its sound. The 4X was not AR’s flagship model. Not by a long stretch. The below tweeter, while not paper, is a very good design for a cone tweeter. It’s actually smooth and well behaved.

I disagree. Distortion usually arises from lack of pistonic motion from the driver. It can be coloration, or most often from resonance(s) in the cone or dome. It’s a lack of faithfulness of the driver’s output compared to input. 2nd and 3rd harmonics produced in addition to the fundamental signal. Cone breakup is common from paper cone tweets. The outer edge of the cone can’t keep up with the center motion driven by the VC. Out of phase output from the outer edge compared to the center of the cone often results. That’s why AR used dome tweeters in its upper line speakers, so the outer edge IS the driven part where the VC is mounted. The PRT has an outer edge restrained and a center free to move by the VC. Can you imagine a woofer with a secured outer edge instead of a surround, trying to reproduce low frequencies? Or lower mids? Or upper bass? I remember the paper cone tweeter used on the ‘66 Wharfdale W70C. It had similar issues of harshness and beaming. It was a paper cone mounted on its outer edge to a grey felt ring. I wound up replacing it with a dome tweeter back in 1982 on my father’s speakers. I couldn’t get that Wharfdale tweeter to sound good.