Revitalizing the AR 3/4 inch tweeter using a waveguide

[Carlspeak]

Attached is a 6 page document detailing the results of using a small elliptical waveguide in conjunction with an original 3/4 inch tweeter of diminished capacity. A quick, non-optimized mod yielded a promising 4 dB increase in SPL response.

REVITALIZING_THE_AR_dome_tweeter.pdf

[genek]

Attached is a 6 page document detailing the results of using a small elliptical waveguide in conjunction with an original 3/4 inch tweeter of diminished capacity. A quick, non-optimized mod yielded a promising 4 dB increase in SPL response.

What's the measurement at 60 degrees off-axis look like...?

[onplane]

Attached is a 6 page document detailing the results of using a small elliptical waveguide in conjunction with an original 3/4 inch tweeter of diminished capacity. A quick, non-optimized mod yielded a promising 4 dB increase in SPL response.

Very interresting post, Carl!! In short .... nice work!!!

My expereince is that 3a's with better tweeter output just make for a far more enjoyable listening experience. Now, I need to clarify that I listen to zero vinyl. My music sources are CD's, FM and satellite.

Then my preference is for jazz and I like to hear clearly the cymbals. Frequently jazz drummers use brushes, so the sound can be quite "soft".

In closing, thanks ever so much for sharing your work, Carl.

Regards,

Jerry

[Zilch]

Excellent work, Carl!

The waveguide's directivity confers some immunity to the grille molding and cabinet edge diffraction effects, but the mid will still interfere with it, perhaps less than as originally configured, though.

Next up, try that little Dayton elliptical in stock form with a compression driver; I'd suggest Eminence, Selenium, 18-Sound, and B&C for "Econo," BMS and JBL neodymiums for upgrades.

I'd be tempted to stick Eminence APT-80 and whatever that is #270-049 in there for a trial, as well as the 4" square Selenium DT150. The "max dispersion" of the AR3a tweeter outside of the cabinet may be less significant with respect to in-situ system performance than has been traditionally assumed....

[genek]

I'd be tempted to stick Eminence APT-80 and whatever that is #270-049 in there for a trial, as well as the 4" square Selenium DT150. The "max dispersion" of the AR3a tweeter outside of the cabinet may be less significant with respect to in-situ system performance than has been traditionally assumed....

A while back I compared an original 3a tweeter to a 3/4" dome AR-218v tweeter supplied by Carl (no measuring equipment here, just ears). Don't know how much of a difference is required to be "significant," but it was audible. See here:

http://www.classicspeakerpages.net/IP.Boar...ost&p=76081

What I wonder is, whatever dispersion the 3/4" dome has/had, buried in the waveguide we're probably listening to the dispersion of the guide and not the dome, yes? If that's the case, why not swap in other drivers? If the guide can provide a comparable HF "spray" and the level controls are capable of adjusting the tonal balance down to original levels if desired, there doesn't seem to be much point in bending over backwards to keep the original driver back in there.

[Zilch]

What I wonder is, whatever dispersion the 3/4" dome has/had, buried in the waveguide we're probably listening to the dispersion of the guide and not the dome, yes? If that's the case, why not swap in other drivers? If the guide can provide a comparable HF "spray" and the level controls are capable of adjusting the tonal balance down to original levels if desired, there doesn't seem to be much point in bending over backwards to keep the original driver back in there.

That's where I'm coming from, and it's certainly possible to measure the directivity of the original tweeter in the cabinet and compare against any substitute. Inevitably, however, we're redesigning the speaker, and at some level, the outcome is not going to be the same as the original. The question then becomes what compromises are acceptable, and in this instance, a better-sounding speaker may not be among them; for restorationist purists, only a reissue of the original will satisfy....

[genek]

That's where I'm coming from, and it's certainly possible to measure the directivity of the original tweeter in the cabinet and compare against any substitute. Inevitably, however, we're redesigning the speaker, and at some level, the outcome is not going to be the same as the original. The question then becomes what compromises are acceptable, and in this instance, a better-sounding speaker may not be among them; for restorationist purists, only a reissue of the original will satisfy....

I think it's important not to confuse goals.

If what one wants is a "restoration," then yes, only original parts or exact clones of originals would be sufficient. Even swapping in the almost-identical-except-it's-ferrocooled 3/4" AR-10pi tweeter wouldn't do.

OTOH, if what one wants is just a "repair," then anything that can produce a tonal balance and dispersion equivalent to the original (regardless of whether one thinks that's a good thing or not) and also fits behind the original grill should suffice. And if it's a solution that can be installed without taking a router to the cabinet and making it impossible to reinstall the original part if desired and allows the user to dial in more response shaping than the original on demand, even better.

And , of course, for a mod or tweak, all bets are off and Frankenspeaker wouldn't be out of bounds, though I suspect that would be beyond what most folks here would opt for.

[Carlspeak]

.............What I wonder is, whatever dispersion the 3/4" dome has/had, buried in the waveguide we're probably listening to the dispersion of the guide and not the dome, yes? If that's the case, why not swap in other drivers? If the guide can provide a comparable HF "spray" and the level controls are capable of adjusting the tonal balance down to original levels if desired, there doesn't seem to be much point in bending over backwards to keep the original driver back in there.

Attached is a fresh set of measurements taken today, zoomed and printed at 1 octave smoothing to better see the general differences in dB level at various off-axis angles. I added a 15 kHz line. For some unexplained reason the dB drop was largest going from 30 to 45 deg. Otherwise typical drops were 3-4 dB with each angle change at 15 kHz.

Still, the WG drop off rate was not as great as the 1 inch dome in a conventional mounting in the 3a. On axis to 45 deg off with the WG was in the 8-9 dB range whilst with the 1 inch dome it was about 15 dB. This suggests the WG used with the original or similar size dome may be a better alternative

than installing a 1 inch dome for best off-axis performance.

Blue=on axis

Red= 30 deg. off

Green=45 deg. off

Brown= 60 deg. off

With regard to the quote above, there are not many 3/4 inch 4 ohm tweeters out there that the 7/8 inch throat diam. will conveniently fit over. If you opt for a 1 in tweeter with a WG, then you must shave the WG further to open up the throat more or consider a totally different one.

Again, the purpose of my experiment was to try to salvage the existing tweeter at low cost. Thus if one doesn't have an inventory of spare tweeters, this would presumably be worth considering.

I may repeat the measurements one more time next week with a grille on just to see what, if any, difference it makes.

[Zilch]

More good data, Carl; thank you.

I'm thinking maybe shave more off the throat to mate with bolt-on 1" compression drivers such as 18-Sound HD125 and B&C DE10....

[genek]

I'm thinking maybe shave more off the throat to mate with bolt-on 1" compression drivers such as 18-Sound HD125 and B&C DE10....

I think I'd be inclined to go with Zilch on this one. Debate on the response and dispersion of the originals aside, if the output of the original tweeter has fallen below usable levels without resorting to measures that mess up whatever advantage might have been left from the dispersion, it's time to put the original tweeter in a drawer someplace in case one wants to reinstall it at a later date for collector-value purposes and install something else, whether it be a HiVi dome or totally different tweeter tech. The best goal to pursue is whatever enables the user to dial in performance that comes as close to original as possible (and opens up a wider adjustment range if that's what's desired), fits behind the grill and is reversible to preserve the ability to reinstall good originals or modern clones of same if one ever wants to and is able to obtain them. It's not as if our living rooms are exhibition halls at the Smithsonian.

Are the mounting holes on that waveguide spaced evenly so someone turning their speakers on their sides could just remove four screws, turn the guide 90 degrees and screw it back down...?

[Carlspeak]

..............Are the mounting holes on that waveguide spaced evenly so someone turning their speakers on their sides could just remove four screws, turn the guide 90 degrees and screw it back down...?

Yes

I also plan another set of off-axis tests rotating the speaker in the opposite direction to see what differences may be evident. Next week.....

[genek]

Yes

I also plan another set of off-axis tests rotating the speaker in the opposite direction to see what differences may be evident. Next week.....

Good show. Please measure at 60 degrees off-axis as well as 45 if you can...?

[Carlspeak]

As promised, here's a couple more response tests showing the off-axis performance of the flush mounted WG when the 3a cabinet is rotated clockwise (CW) when viewed from above. The response test attached in post #8 was done with the speaker rotated CCW - which is the way I usually do it. The colored lines are coded same as in post #8.

However, as I thought about it more, I realized with this being an asymmetrical design with the tweeter tucked up in one corner of the baffleboard, I needed to see what effect taking resps. tests with the cab. rotated in the direction opposite to what I normally do. Lo and behold, the drop off was more severe with the tweeter going more away from the mic when rotated CW. (note: just to clarify, I don't use a fancy calibrated turntable like that sold by Old Cololny SL. I just rotate the cabinet sitting on top of another one and set the angles with a plastic draftsman's triangle). This method may not be ideal, but I still think is shows what's possible if one doesn't take into consideration an asymmetrical loudspeaker. So, a caution to hobbyists out there measuring their speakers: if the driver is configured asymmetically on the baffle board, be sure to take a second set of off-axis measurements.

The second response test was done using a normal 3a grille frame fitted with new, linen fabric. Rotation: CCW. The blue line is an on-axis 'control' without the grille. The red is on-axis with the grille in place. Response dropped about 1 dB overall from 2 kHz on up. Not bad. The green=30 deg. off-axis. Brown=45 deg. and light blue=60 deg. At 60 deg. off-axis, the drop at 20 kHz was 13 dB with the grille on vs a similar 13 dB drop with the grille off (post#8).

Thus, the presence of the grille frame and cloth did not seem to help or hinder the response of the WG (in this particular test).

So, I think that's the end of my diddling with the WG application. I'll put the tweeter back on the baffle board, install the woofer and continue to use this speaker as my 'control' for customer comparison tests with my second speaker modded with the Super-Mod kit.

The overall conclusion I draw from the work presented here is the Dayton WG can help boost the output of a tired, old 3/4 inch dome tweeter and, it's a low cost mod which is reversable to the extent the classic value of vintage AR speakers fitted with the 3/4 inch tweeter is not diminished. Thirdly, it appears to have better off-axis response than a normally flush mounted 1 inch dome tweeter.

My personal opinion, FWIW, is the off-axis response performance of the 3/4 inch dome tweeter, although somewhat unique and a highly novel development considering the era in which it was developed, is held in way to high a regard.

'cus, at the end of the day........

[Zilch]

My personal opinion, FWIW, is the off-axis response performance of the 3/4 inch dome tweeter, although somewhat unique and a highly novel development considering the era in which it was developed, is held in way too high a regard.

'cus, at the end of the day........

The legendary standing of the AR tweeter is well-founded relative to earlier designs, but diminishes to myth in comparison with modern ones. Technology has advanced appreciably during the intervening 40 years.

By any measure, even those done by AR themselves and published by Allison in 1970, the alleged benefits of the improved dispersion achieved by the dome, clearly evident in its out-of-cabinet response characteristics, are all but mooted by the cabinets and grille moldings housing them commonly used at the time. Modern constant-directivity waveguides are somewhat immune to these deleterious effects, as is confirmed by your experiments.

Thank you once again for undertaking this and similar efforts to establish the facts, and for publishing them here....

[genek]

So, I think that's the end of my diddling with the WG application.

Aww. I was hoping you were going to try measuring that setup with one of the compression drivers Zilch suggested.

[Carlspeak]

Aww. I was hoping you were going to try measuring that setup with one of the compression drivers Zilch suggested.

Oh, I think someone else will have that one covered.

[genek]

Oh, I think someone else will have that one covered.

The "cut open the entire top third of the cabinet and replace the mid and tweeter with a waveguide" is definitely covered. I was thinking more along the lines of the "flush fit in the existing tweeter opening and under the original grill" test, which is probably still wide open. Unfortunately, I don't have any spare speakers around here to try it myself.

[Zilch]

Unfortunately, I don't have any spare speakers around here to try it myself.

Yard sales and thrift stores; $20 will afford you the appellation, "Researcher."

[genek]

Yard sales and thrift stores; $20 will afford you the appellation, "Researcher."

Frankly, I much prefer my current one, "picky customer with disposable income."