Henry, Henry, Henry, Henry...how could somebody who did so many hard things so well, do so many easy things so badly?

[soundminded]

Taking apart AR2a and really getting into it, I've gotten a great deal of new found respect for it, and lost some. They gave you a lot for your money. Much more than I thought. But the crossover design stinks. There's no other way to put it. OK, I can live with three first order filters. What I can't live with is a midrange control pot which doesn't make the midrange speaker play any softer, just rolls off its high end. And it looks like two four ohm speakers in paralllel (combined DCR is 1.6 ohms and they ARE wired in parallel.) To confirm that I wasn't just imagining it, I took a pair of 8 ohm resistors in series and wired it in place of the pot. Sure enough, jumping it out only restores the high end of the midrange units which the pot attenutates, it doesn't make it any louder or softer at the other extreme. I checked the other unit and got the same results. What the hell kind of design is that? OK, now I've got to redesign the crossover network. Any suggestions? Bi-amping? Tri Amping?

BTW, it would have been nice not only for restorers and tinkerers 40 years after the fact but for people in the service department if he had been a little less stingy with wire. I've had to cut just about everything apart to work on any of it. And that's not good either. BTW, although the AR2ax woofer has a better (newer) magnet design, the AR2a woofer is better made in just about every other way imaginable. Looks like they were trying to cut costs on it when they made AR2ax.

[Carlspeak]

>Taking apart AR2a and really getting into it, I've gotten a

>great deal of new found respect for it, and lost some. They

>gave you a lot for your money. Much more than I thought. But

>the crossover design stinks. There's no other way to put it.

>OK, I can live with three first order filters. What I can't

>live with is a midrange control pot which doesn't make the

>midrange speaker play any softer, just rolls off its high end.

> And it looks like two four ohm speakers in paralllel

>(combined DCR is 1.6 ohms and they ARE wired in parallel.) To

>confirm that I wasn't just imagining it, I took a pair of 8

>ohm resistors in series and wired it in place of the pot.

>Sure enough, jumping it out only restores the high end of the

>midrange units which the pot attenutates, it doesn't make it

>any louder or softer at the other extreme. I checked the

>other unit and got the same results. What the hell kind of

>design is that? OK, now I've got to redesign the crossover

>network. Any suggestions? Bi-amping? Tri Amping?

>

>BTW, it would have been nice not only for restorers and

>tinkerers 40 years after the fact but for people in the

>service department if he had been a little less stingy with

>wire. I've had to cut just about everything apart to work on

>any of it. And that's not good either. BTW, although the

>AR2ax woofer has a better (newer) magnet design, the AR2a

>woofer is better made in just about every other way

>imaginable. Looks like they were trying to cut costs on it

>when they made AR2ax.

Lots of critique of the xover design. However, I noted the absence of comment on how the speaker sounded before taking it apart. Am I to assume you weren't happy with it? If that's the case I'm a bit surprised given most, if not all, AR2a and ax owners who post here are quite happy with the musical performance of the factory design - given all components are working properly.

Remember, it's all about the music

Carl

Carl's Custom Loudspeakers

[JoeB]

>Taking apart AR2a and really getting into it, I've gotten a

>great deal of new found respect for it, and lost some. They

>gave you a lot for your money. Much more than I thought. But

>the crossover design stinks. There's no other way to put it.

>OK, I can live with three first order filters. What I can't

>live with is a midrange control pot which doesn't make the

>midrange speaker play any softer, just rolls off its high end.

> And it looks like two four ohm speakers in paralllel

>(combined DCR is 1.6 ohms and they ARE wired in parallel.) To

>confirm that I wasn't just imagining it, I took a pair of 8

>ohm resistors in series and wired it in place of the pot.

>Sure enough, jumping it out only restores the high end of the

>midrange units which the pot attenutates, it doesn't make it

>any louder or softer at the other extreme. I checked the

>other unit and got the same results. What the hell kind of

>design is that? OK, now I've got to redesign the crossover

>network. Any suggestions? Bi-amping? Tri Amping?

>

The only crossover schematic I've seen is the one at classicspeakerpages.net/ar/documents/AR2ax.jpg which clearly shows the midrange connected to the center ('B') connection of the 16 Ohm pot. The high side of the pot is connected to the capacitor and the low side of the pot is connected to common. This is similar to the connection of the teweeter on an AR-3a. This should allow adjustment from full off to direct connection to the 6 UFD cap. The lack of midrange inductors limits the precision but you should still get full adjustability. Is it possible the pot is wired wrong in your speaker?

[soundminded]

As supplied by the factory, frankly it sounds awful. All of the ingredients are there but in the wrong proportions. There is a document in the library here explaining how to upgrade the AR2a to make it sound less "boxy." I think that's a reasonable characterization of it and there is far too much upper bass/lower midrange. So far I've determined that the loudness of the drivers is way out of whack. The woofer is too loud and the tweeter even at its maximum setting is too soft. And as I've said the midrange level control doesn't alter the midrange level much it only rolls off its high end as the series resistance increases. Having replaced the factory caps and disassembled and cleaned the pots (one unit only so far), I've tried bypassing the midrange control, setting the tweeter at max, and putting a 10 mfd capacitor across the woofer voice coil and an 8 ohm resistor in series with it. This helps a great deal although I'm sure I'll hear a lot about increasing the electrical Q and reducing the damping factor to under 1. As is my usual custom, I'll be adding indirect firing poly tweeters as well and adding an equalizer. This speaker needs all the help it can get at the high end. However, right now I have a hunch it has the potential to sound excellent. I'm kind of tempted to bring all of the driver leads outside the box and play with the crossover network externally. Somehow, having worked on more than a few of these, I am beginning to think all speakers should have external crossover networks, especially those in sealed boxes.

[soundminded]

I've seen this...and copied it down. Mine are NOT wired this way. The midrange potentiometers in mine are strictly a series variable resistance elements on both units, not a voltage dividers and there is no evidence that there was ever any solder applied to the unused terminal on either midrange pot indicating that someone might have gotten there before me. Besides, after my experience with the putty, I know nobody else has ever been inside. The midrange subenclosures and tweeters are sealed with the same putty. I'm quite confident I'm the first one to see the inside of this box since it left the factory.

[Carlspeak]

The only crossover schematic I've seen is the one at classicspeakerpages.net/ar/documents/AR2ax.jpg which clearly shows the midrange connected to the center ('B') connection of the 16 Ohm pot. The high side of the pot is connected to the capacitor and the low side of the pot is connected to common. This is similar to the connection of the teweeter on an AR-3a. This should allow adjustment from full off to direct connection to the 6 UFD cap. The lack of midrange inductors limits the precision but you should still get full adjustability. Is it possible the pot is wired wrong in your speaker?

This is also the way my experience with upgrading AR xovers shows. Perhaps Soundminded's 2a's may be an anomoly? Maybe they were wired on a Monday or Friday analogous to the auto industry's historical inferior quality on these days? Soundminded's problematic single 2a data point may not be representative of Henry's true intent.

Remember, it's all about the music

Carl

Carl's Custom Loudspeakers

[soundminded]

It hardly matters. Even at full midrange output, it isn't enough to balance the woofer. I put an 8 ohm 20 watt resistor in series with the woofer and a 10 mfd cap across the voice coil to further limit its high end. this helped a lot. I've also added a 1/2" poly crossed over with a 2.7mfd series cap and an 8 ohm resistor. There's great improvement already. All that's left I think is another indirect firing poly or two and one unit is done. It will be an excellent loudspeaker system just as I expected. I'm now driving it with a Pioneer SX950 receiver rated at 85 wpc which I bought in mint condition at a garage sale for $10 several years ago. They guy said his wife told him to get rid of it or else. I wondered what I was going to do with it. Now I know.