Recapping the bass section of the AR9

[Diamonds&Rust]

Well gang, I’m back with another long-winded story. At least you have ample info to pick-apart.

I have replaced the capacitors in the bass section of my AR9s. I used Jensen non-polar electrolytic capacitors because poly caps would have cost more than I had in my piggy-bank. Even the Jensens weren’t cheap at about $ 60.00 per speaker.

Because of the way the bass section in an AR9 works, I have been more than a little afraid to do this.

It was only because of the dramatic improvement replacing the 2500uF capacitors made in my 10pi speakers (telling me the original 2500uF caps were bad, very bad) that I decided to try this. The 10pi speakers are no longer in the room I am most likely to be in; the AR9s took their place.

Due to my acquiring (with more than a little help from my friends) a really pristine pair of NHT 3.3s which are in my big room now, the 9s had to move and the 10pi had to go elsewhere. The 10pi’s were sounding sooooo good, too. I finally had them positioned just right after the cap change and was really enjoying them. I just have too many good memories associated with this set of 9s to get rid of them. [being sentimental can get in the way of good thinking sometimes]

First, I dragged an old equalizer and real-time analyzer (RTA) out of the closet and hooked it up to my system with the 9s. While there are all sorts of limitations to the usefulness of using an RTA, and I know better than to trust it, and I know mic position is extremely important, and I “get” that this isn’t perfect, I had to do something and this seemed a reasonable starting point. Among other things, the RTA told me that I had a big hole in the 180Hz band.

Fiddling with the EQ confirmed that adding (substantial) 180Hz signal and (a little) 90Hz signal did fill-in a hole. But it sounded a little “funny” boosting that band with the EQ. I think I was just forcing the lower mid to play louder at the low extreme of its range and wasn’t doing much to the woofer. It wasn't bad, exactly. In fact, it could be enjoyed that way. But I knew I was overcoming a problem by introducing other problems. That bugs me.

The resonance frequency of the lower midrange speaker in its enclosure is listed as 175Hz. Since, I reasoned, the roll-off of the lower mid should be substantial and quick below 175Hz - the missing 180Hz had to be missing mostly from the woofers.

Looking at the schematic, I assumed the 470uF capacitor was partly responsible for the roll-off of the woofers. I re-read "Engineering the AR-9" from the forum's library. To quote Tim Holls, “ As the signal frequency increases, the impedance of the resonant circuit also goes up and the speaker system output goes down! By selecting precise values for the components and using a choke coil with low internal resistance, “Q” and impedance are smoothly controlled to yield extended bass response with full amplifier protection.”

As far as I’m concerned, the important words were “precise values.” The idea that these thirty year old components had held their precise values was laughable.

My thought was that whatever I screwed-up by not getting the ESR just right could be no worse than leaving a bad capacitance or inductance value in the speakers. Remember, I *knew* identical Callins 2500uF caps in my 10pi’s were bad. There was no reason at all for me to believe that just because these 2500s were in a 9 that they would have the good manners to conscientiously stay within spec. Surely when Tim Holls said, “precise values,” and he used a capacitor, capacitance was a major spec to get precise.

Once I got the old caps out, amazingly my cheap capacitance meter reported that the old capacitors were almost *exactly* on spec. “What the hell?” I replaced the old caps anyway because my 10pi capacitors were also “in-spec” according to my meter, but had obviously gone bad.

After I replaced the 470uF and 2500uF capacitors I got the RTA to tell me that there was no more hole at 180Hz. My ears confirmed that boosting the 180Hz as I had prior to replacing the capacitors sounded terrible.

Before anyone begins chastising me for not having a laboratory anechoic chamber with a Bureau of Standards and Measures calibrated pink noise generator and the latest in RTA technology complete with a six hundred acre open field above which I could suspend an AR9 at 300' in order to accurately measure the bass response; whether the bass is flat in my room or not is not dependent on anechoic measures and is ***completely beside the point***.

THE point is that by changing these capacitors which measured almost perfectly with my meter, the response of the speaker system changed. And now, with the introduction of measures from an RTA, whether you think they are true and accurate measures or not, you don’t have to take my word for *hearing* the difference. I not only heard it -- I measured it. And yes, I moved the microphone around (about a foot in any direction) when I was testing. Yes there were differences. No, they weren’t in the frequency range in question. No, they weren't huge in any range.

According to the RTA the only negative change is a very small dip in response at 125Hz which was not there before. It’s a very small dip and judging from the 90Hz band’s being almost “flat” with regard to the rest of the cabinet, I would say this is most likely a room effect. Why I did not see it before the capacitor change I have only one guess and it is probably wrong - so I won’t posit it here. The 90Hz reading I wouldn't trust at all, really. It's probably the room.

The sonic effect (god help me for saying it) is that the woofers handle transients better. That is, they are “faster.” (Not literally, but as the term is commonly used.) The woofers are also “quieter” and just as I experienced in the 10pi, I have the impression this is due to better amplifier control. I have to play with positioning a lot more before I “pronounce” an effect on imaging, but the lower midrange (it sounds higher than 200Hz, but must not be much higher) is way, way more articulate. It may be that the lower harmonics are playing well again, so it makes sounds with higher primary frequencies seem more detailed... I don’t know, I’m just taking shot-in-the-dark guesses.

Obviously my experience with Jensen NPEs from Europe may not be what you’d get if you stuck a bunch of $0.50 Madisound capacitors together to come-up with 2500uF and 470uF. I’d be amazed if it were the same.

I don’t know how the old caps were failing that wasn’t “C”. Whether inductance went up substantially at the frequencies of interest, or if the ESR rose substantially and caused the compensation circuit to prematurely shut the woofers down, I just can’t say.

When I first put the 9s in this room, I was very sorry I had to give-up the 10pi’s. After putting even more money, time, and effort into them, I can say that I *think* I am going to enjoy the 9s in here even more than I did the 10pi’s. That is how it should be, but I really like that early 10pi tweeter.

For all of you out there with AR9s who are afraid to change the woofer caps, as I was, there is no need to be fearful. Whatever I might have made worse, it is a tiny fraction of what I changed for the better. Even using NPEs this is not an inexpensive little undertaking, but I don’t feel the least bit as though I paid too much for the improvement I got. It was a bargain, relatively speaking.

With a pair of 2500uF Spragues in Don's 10pi's being bad, my pair of 2500uF Callins in my 10pi's, and now the 2500uF and 470uF Callins from my AR-9s being bad, I have completely convinced myself that all of these old capacitors should be replaced. Period. All of them.

Bret

[ar_pro]

.

[ar_pro]

Excellent write-up, Bret - it's very difficult to adequately analyze & describe some of the things that we hear - even from systems like the AR-9, with which we are so familiar.

Now, here's a bit of a curve ball -

To date, we've rebuilt seven pairs of AR-9 speakers, and each & every one had woofer caps that measured within spec.

We've used NOS replacements when they were available from H&R, as well as built-up caps, and have even swapped caps from one system to another, and no matter what replacement we've used for the woofer caps, nothing seems to have had any sort of impact on the woofers' sonic quality - the woofers' output seems balanced with the rest of the system, and extension seems unaffected.

BUT as soon as *any* of the capacitors in the upper half of the crossover are changed, look out!

Built-up woofer caps matched with new upper-range capacitors have *consistantly* resulted in a low-end response that at first, appeared tighter, but on further listening, was clearly the result of a diminution of LF extension. Relative overall efficiency also appeared to increase, with the system playing more loudly with a given level of input.

After checking audible in-room LF extension with an audio generator, it became clear that the *original* or NOS caps were the only combination that permitted the AR-9 maximum extension! As soon as the built-up caps were removed, the originals or NOS combined with the upper-range replacement caps for optimal performance.

The apparent slight increase in volume I've kind of chalked up to the amplifier's interaction with the truncated-response characteristic of the AR-9 with built-up LF caps.

The phenomenon went away with vertical bi-amping, although this arrangement had no effect on LF extension.

And so, I'm left with vintage LF caps & replacement HF caps, for what sounds best to me!

Oh, and NO level controls!

[Diamonds&Rust]

We've used NOS replacements when they were available from H&R, as well as built-up caps, and have even swapped caps from one system to another, and no matter what replacement we've used for the woofer caps, nothing seems to have had any sort of impact on the woofers' sonic quality - the woofers' output seems balanced with the rest of the system, and extension seems unaffected.

BUT as soon as *any* of the capacitors in the upper half of the crossover are changed, look out!

Built-up woofer caps matched with new upper-range capacitors have *consistantly* resulted in a low-end response that at first, appeared tighter, but on further listening, was clearly the result of a diminution of LF extension. Relative overall efficiency also appeared to increase, with the system playing more loudly with a given level of input.

After checking audible in-room LF extension with an audio generator, it became clear that the *original* or NOS caps were the only combination that permitted the AR-9 maximum extension! As soon as the built-up caps were removed, the originals or NOS combined with the upper-range replacement caps for optimal performance.

I can't tell you how happy I am that you chimed-in on this. I have not tried built-up caps in the AR9.

I didn't want to get into it because I can't *prove* this, and have no measurements to back me up, so I decided to be quiet rather than be punished for not having a lab available.

I have tried a sort-up built-up capacitor arrangement and it didn't work worth a flip. I've tried "by-passing" those big caps and that didn't work, either. In the AR9s woofer arrangement, I suspect that whatever is different about a cascaded bundle and a single large capacitor is to blame, but I don't know why. I guess, although I can read and semi-understand the "automatic transmission" layout of the 9s crossover per Holls' explanation, I really do not understand all the implications and inner workings of it.

This single, large, Jensen cap may be as close as I can imagine to a NOS cap. It might be important that the big one is labelled 2750 and measures 2900. That's as close as I know exists. Unless I'm sadly mistaken, I'm not missing any bottom extension with these non-bundled, single, large caps.

I don't know what to blame for the failure of a bundle.

Clearly my old caps were shutting down the woofers too soon. Can I blame that on rising ESR? Do you have an opinion?

The apparent slight increase in volume I've kind of chalked up to the amplifier's interaction with the truncated-response characteristic of the AR-9 with built-up LF caps.

The phenomenon went away with vertical bi-amping, although this arrangement had no effect on LF extension.

And so, I'm left with vintage LF caps & replacement HF caps, for what sounds best to me!

Oh, and NO level controls!

You know, I'm always afraid of the heat I'll take for saying bad stuff about cheap caps in large bundles. Someone's liable to accuse me of "hearing things." (of course I hear things - that's why I've got AR9s and 10pis instead of Panasonic "Thrusters") So thank you for confirming what I experienced and what I would have guessed about the 9.

I suppose I had better officially modify my stance, to:

IF you can find like-kind capacitors, don't be afraid to change them. If you are going to build a cheap bundle to replace your old capacitors - DON'T.

Bret

[Guest PhoenixRising]

Excellent information from both of you. Thanks for sharing your experiences.

I've had my AR9's for almost 10 years, and compared to just about any other speaker I've heard (new or old), they still sound better to me. However, with all the chatter online about recapping vintage speakers, I'm tempted to try it, just to see if I'm missing out on something.

My 9's were definitely restored to some extent by the previous owner. The mid drivers were swapped out (presumably to replace faulty originals), the cabinets were obviously refinished, the binding posts were upgraded, the grill cloth was replaced, etc...

However, I don't know how to tell if any of the caps were replaced by the PO, and I can't seem to find a reliable pic of the original crossover components to compare mine with. Any suggestions? Could I take some pics of what's in there now and have you guys tell me if they are original components or replacements?

I am at least proficient with a solder iron (being a long time guitarist who likes to swap pickups/electronics on a daily basis!), so physically doing the replacement is probably within my abilities. Reading the schematic, on the other hand, could be tough.

Should I attempt this myself? Stick with what I have and forget it? Take it to a professional for a recap? I hate to spend too much on this since I'm happy with them as they are.

Any thoughts from you pros would be appreciated. Thanks.

[johnieo]

However, I don't know how to tell if any of the caps were replaced by the PO, and I can't seem to find a reliable pic of the original crossover components to compare mine with. Any suggestions?

Greetings:

Electrolyic capacitors of that era are usually date coded. You might look for codes on the capacitor of the form YYWW. For example 7925 means 25th week of 1979. All of the caps if original should be clustered in a reasonably tight time frame. If you find one dated years later, you know it has been replaced, no question.

The seals in electrolytics leak very slowly over many years allowing water to evaporate from the liquid electrolyte and change both C and ESR (alternatively expressed as Dissipation Factor). The leakage was really bad in some brands -- and very small in the best of the brands. Personally I would not spend money on NOS for that reason. New miniature electrolytics often do not have date codes, but are instead are rated by their lifetime, e.g., 2000, 3000, or 5000 hr.

Cheers,

[Rlowe]

Greetings:

Electrolyic capacitors of that era are usually date coded. You might look for codes on the capacitor of the form YYWW. For example 7925 means 25th week of 1979. All of the caps if original should be clustered in a reasonably tight time frame. If you find one dated years later, you know it has been replaced, no question.

The seals in electrolytics leak very slowly over many years allowing water to evaporate from the liquid electrolyte and change both C and ESR (alternatively expressed as Dissipation Factor). The leakage was really bad in some brands -- and very small in the best of the brands. Personally I would not spend money on NOS for that reason. New miniature electrolytics often do not have date codes, but are instead are rated by their lifetime, e.g., 2000, 3000, or 5000 hr.

Cheers,

In addition to John's comments the following pics may help. I collected them a while ago. The silver cap cans should be Callins and the black and red caps were typically used by AR. I don't know if the blue cap on the bottom board is original, the others look as if they are.

[ar_pro]

Panasonic "Thrusters", that's great!!

Seriously, I thought I might be nuts when the effects of built-up LF caps only became apparent when the upper-range caps were also replaced.

Concern for damage to the irreplaceable original drivers forced the move to newer, much-closer-to-spec upper range replacement capacitors.

But clearly, something takes place with this arrangement of new upper-range caps & built-up LF caps that's not desirable, and places the result outside the boundary for what, to me, constitutes a proper-sounding AR-9.

Going back, I fully expected out-of-spec LF electrolytic caps to be the norm, so having never found a "bad" LF capacitor, and listening to the poor result with expensive, built-up capacitors, *not* having to replace the originals is an easy decision with which to live.

At this point, and purely from the practical aspects of doing the labor, and living with the outcome, I've found that replacing the upper-range capacitors is worthwhile, while removing LF caps that are still in-spec is less than desirable - again, the mysteriously truncated LF extension, and apparent increase in overall output.

Maybe we should all say a quiet prayer of thanks that the LF capacitors that AR chose were as good as they were - because once that reference is lost, who'll know what an AR-9 is supposed to sound like?

[ar_pro]

PhoenixRising, these images are from an unmodified, very early pair of AR-9s:

[Guest PhoenixRising]

PhoenixRising, these images are from an unmodified, very early pair of AR-9s:

Thanks guys for the pics and info. I love this forum!!

Mine are apparently all original, as the pics I took today will show. I'm really interested in recapping at least the upper range since you guys have seen dramatic improvements with this.

Can you guys recommend a source of info to educate myself on all things caps? What to buy and from where, and the basics of the installation?

[Diamonds&Rust]

Panasonic "Thrusters", that's great!!

Seriously, I thought I might be nuts when the effects of built-up LF caps only became apparent when the upper-range caps were also replaced.

Going back, I fully expected out-of-spec LF electrolytic caps to be the norm, so having never found a "bad" LF capacitor, and listening to the poor result with expensive, built-up capacitors, *not* having to replace the originals is an easy decision with which to live.

Hold on there, Polyanna!

ALL the LF caps I've ever touched are bad. Yours too, man. They could suck the paint off a golf-ball. The %*$# meters that show they are good are what's crap around here. Notice the stupid meters show that the built-up cap values are the same as the old caps, but do they make the same noises? NO, I SAY! 'er, uh.... NO, I HEAR!

What we need here is a serious piece of calibrated lab gear put to good use.

You agree with me even if it embarrasses you. Or I agree with you. In any case I suspect we're in good company. There IS a reason. When we find the @#$% reason, THAT will be our finest hour. We will never surrender. We will fight for our island...

...but I digress.

For whom does the bell toll? It tolls for thee...se stupid NPE capacitors. They stink. They are smelly. I hate them. Hate, hate, hate.

Ad Hoc, Ad Loc, Ad Quid Pro Quo. So little time, so much to know. Pepperland. Love Music Love. Blue Meanies. Help! <if you don't remember it, your hair's too short and isn't grey enough>

NPEs are definitely Blue Meanies.

Even yours that measure great at the couple of frequencies the meter might be working at are *not* still good. Hell, mine measure fine and they could knock a olfactory impaired buzzard off a carrion wagon - at the height of the Black Plague, even.

I'm going to find-out how much money it would take to get a cap manufacturer to make us some low ESR 2500uF 60v NP capacitors. I'm just betting it'd be cheap per cap, but the minimum order would be 10,000 pieces. Got $50,000 laying around with nowhere to go? Sometimes the affordability of "stuff" surprises me, so we'll see. You don't know if you don't ask.

At this point, and purely from the practical aspects of doing the labor, and living with the outcome, I've found that replacing the upper-range capacitors is worthwhile, while removing LF caps that are still in-spec is less than desirable - again, the mysteriously truncated LF extension, and apparent increase in overall output.

Maybe we should all say a quiet prayer of thanks that the LF capacitors that AR chose were as good as they were - because once that reference is lost, who'll know what an AR-9 is supposed to sound like?

I can't think of anyone except me and thee.

Wouldn't it be great if we had some decent, unassailable data on this? <he says indulging in a bit of heavy-handed foreshadowing>

Nice photos everyone's posting, though. Those caps remind me of "Chucky" or "Michael Myers" or "Pinhead" or... maybe THAT'S whatever happened to Rosemary's Baby.

Bret

[Diamonds&Rust]

Mine are apparently all original, as the pics I took today will show. I'm really interested in recapping at least the upper range since you guys have seen dramatic improvements with this.

Can you guys recommend a source of info to educate myself on all things caps? What to buy and from where, and the basics of the installation?

They look all original to me.

My best advice about learning everything capacitor? You've got it all here, just not laid-out nicely. Heck, if you'll search you'll find John O'Hanlon's dissection pictures, failure mode explanations, and even perfectly plausible explanations for why some of them fail before others!

You've got Pete B's experience re-capping an AR-11 and what he found and what he had to do.

You've got rrcrains, and Rlaski's, and Roy's, and ar_pros', and my, and, and, and, and Sean's, and Ken Kantor's take, and... We go into dielectric absorption and what dissipation factor is and Pete can tell you how to calculate things backwards and John will tell you how it is done forwards, and soundminded and Ken had an interesting "exchange" about "knees" and there's a discussion of how resistence contributes to phase shift and hundreds of hours of debate (I wish I had some of that time back) about "special" capacitors and what might be different about them, and...

Well, you get the idea, yet here we are debating it again. I make my New Year's Resolution not to debate this again, at any level, in 2008.

Yes, absolultely, recap your speakers' upper cabinets. Leave the bass cabinet alone unless you get lucky and find some relatively fresh, large aluminum can, non-polar electrolytic capacitors. For the upper cabinet you are going to *have* to experiment to see what sounds best to you. Maybe it's a Dayton poly cap with a little resistor on it, maybe it's a Solen with a little resistor, maybe you can still get North Creek Zen capacitors (at a multiple of the cost) and use the recommended by-pass (film and foil on the tweeter, Harmony on the mids), and maybe you will discover a whole new thing that just sounds fantastic.

Every single one of us who has recapped a 9 or 90 who has a capacitance meter has found at least some of those black and red capacitors are really badly out-of-spec. For the sake of your old, irreplacable drivers, change the caps to something within specifications. Anything in spec is better than anything out of spec. The upper midrange caps seem to be the worst, but your mileage may vary.

It's a lot of fun screwin' around in those cabinets! Have at it and remember to report back what you did and how it all came out.

You think they sound good now... but just wait until you get them playing correctly again. Of course, it's taken me a little over five years to get there and I didn't even know I still had a problem until I got another great set of speakers to compare against. But now? Now I'm pretty doggone sure I can't make them better than this. Maybe. I've still got a thing or two to try. It's a hobby.

Bret

[Pete B]

Can you guys recommend a source of info to educate myself on all things caps? What to buy and from where, and the basics of the installation?

http://www.classicspeakerpages.net/IP.Boar...capacitor+myths

[Diamonds&Rust]

Going back, I fully expected out-of-spec LF electrolytic caps to be the norm, so having never found a "bad" LF capacitor, and listening to the poor result with expensive, built-up capacitors, *not* having to replace the originals is an easy decision with which to live.

ar_pro,

You have a piece of data that could be important in an ongoing effort to discover *why* this seemingly reasonable solution doesn't work.

Do you remember, just approximately, how many capacitors of what type (poly, whatever) you used for your built-up caps?

(e.g., five 100uF poly; or two 200uF NPE and one 69uF poly)

Also, did you twist leads and solder or do the whole "ring tongues bolted together" trick or what?

There has to be a reason this doesn't work and I think investigating it is interesting. It may prove practically pointless, but it is still interesting.

Bret

[rrcrain]

One of the very unofficial tests I did on the old NPEs I took out of AR90s was to attach the caps to an electronic 1,000VDC megger. This simple test quickly demonstrated the old caps were leaky, unable to hold any charge at all while the new poly caps held a a dangerously high charge over night and had to be discharged thru a resistor. An electronic megger uses an extremely low current, so low it won't shock the user unlike the older crank type meggers, and it takes several hours to charge the large poly caps. And yet, these old caps tested in spec on a digital cap meter.

[Guest bluelobster]

"You can use Bennic NP caps - since they tend to be a bit under

their stated value you can use that "space" for a good quality

film bypass. For the 470 uf cap I'd use a 330 uf and a 140 uf

with a 5 uf film bypass. For the 2500 uf use (2) 1000 uf and

a 500 uf with a larger film bypass, like a 40-ish uf. You'll

want to get an extra pair of 40 uf and 50 uf Bennics so you

can "trim" the composite cap to a good tolerance - 5% or better

is good, I went for 3%. "

This is from Jim McShane, who is getting these for me to redo my "new"

speakers. Jim is an AR9 fan as well, he told me once that he had family

that was involved in there development.

Not only that I know he has donethe research and development for these

replacements himself.

I'm following his lead, using my Harman Kardon Citation II to drive them.

Yes tubes!

http://pages.prodigy.net/jimmcshane/

[Diamonds&Rust]

"You can use Bennic NP caps - since they tend to be a bit under

their stated value you can use that "space" for a good quality

film bypass. For the 470 uf cap I'd use a 330 uf and a 140 uf

with a 5 uf film bypass. For the 2500 uf use (2) 1000 uf and

a 500 uf with a larger film bypass, like a 40-ish uf.

You can, but I wouldn't.

Wait until you hear it.

You're better off leaving the old caps in the speakers.

And no, I don't think I know more about the AR-9 that he does.

Bret

[Guest bluelobster]

Well Bret

Since I've never heard the AR9 before and to listen now is not a fair assessment.

Like old tube amps the parts inside are tired, so they are not a true representation

of what I would hear if they were as new.

I'm rebuilding them and I really only want to go in once. I reread your post above I'll

make this deal with you. The woofer caps aren't that hard to get at, so I'll first try my

new caps installed, giving them time to settle in and my ears to become accustomed to them.

Then I'll put the old caps back in and give it a fair listen. If the old are better, they stay!

By the way I NEVER implied, suggested or insinuated Jim knew more than you.

I am open to suggestions and try to take all input into consideration, thank you for yours.

Mike

You can, but I wouldn't.

Wait until you hear it.

You're better off leaving the old caps in the speakers.

And no, I don't think I know more about the AR-9 that he does.

Bret

[Diamonds&Rust]

I'm rebuilding them and I really only want to go in once.

Then do what is successful. Find a new, high quality, 470uF capacitor. At least one brand is available. I have only *tried* the single Jensen so can only vouch for it. I have only tried it because it is all I can find. I didn't get the Jensen because it is magical, I got it because I could. I am listening to them now. They are the best I have found. They are not terribly expensive (relatively speaking).

I have tried built up capacitors of brands we discuss here often and they do not sound fine; ever; in any use; be that a 100uF or a 470uF or a 24uF bundle. The A/B I'm afraid I've encouraged you to try is not the best course of action either; therefore it is a waste of your time and money. I've already wasted mine.

I do not have the engineering background to say why. Does that matter?

ar_pro has had the same experience I have had. Two guys with the same experience does not make a "law of nature" but our observations agree too well in too many different trials for it to be entirely accidental.

And no, Mike, I wasn't taking offense or meaning to imply that you said *anything* about what I might or might-not know. I was, however, trying to stop a certain objection which might have been raised by someone who has never touched an AR9 or tried built-up capacitors in them.

Please don't assign any motive to my post. I don't have one of those egos that says I have to be right all the time. You might have a solution even better than the Jensen. The one you listed isn't it. If I can keep you from making a mistake, I'll plead like it is important to me to be right.

So make me a different deal, Mike - try what actually sounds good and compare that to either the old cap or the solution you originally proposed. If you don't agree with my ears, I'll buy the Jensens from you and pay the shipping. You just lose the time installing them and listening to them (gotta give 'em a few days unless you form them outside the speakers).

I'm not trying to be combative, or right, but you can't profit from the mistakes of others if others don't admit their mistakes.

Bret

[Guest bluelobster]

Bret

Fair enough, I'll reconsider or at least leave my options open.

Mike

[ar_pro]

I wish that I'd taken some photos of the installation, as I was very pleased with the cosmetic result.

I was so certain that the new built-up 470 (5x100) & 2500uF caps (5x500) would be an improvement, that I pulled the originals, and installed a custom plywood board with screws & glue, and a terminal strip to make the appropriate connections for the new caps.

Idiot. Moron.

Having been so happy with the necessary replacement of the upper range caps, I assumed that replacing the woofer caps would yield a similar fine result. To make a very long and frustrating story short, here are the subjective conclusions that we drew, representing in order, the best-to-worst-sounding arrangement:

Best: AR-9 with all-original drivers, level switches disconnected, upper-range caps replaced, and original woofer caps.

Very Good: Same as above, with clean level controls left in place.

Acceptable, On A Good Day, With A Tailwind: All-original parts, competently refoamed drivers, with an untouched crossover (assuming you can find a "live" spot on the three level switches).

Disappointing: As "Best", but with built-up woofer caps.

Crapulent: Anything with unoriginal, "replacement" drivers.

In the end, so terrible was the result, that we wound up removing the beautiful replacement boards, and re-installing the original LF capacitors. This was three summers ago, but it's possible that the boards are still somewhere in my son's garage, and if we can locate them, I'll make them available to the forum for testing.

[Diamonds&Rust]

In the end, so terrible was the result, that we wound up removing the beautiful replacement boards, and re-installing the original LF capacitors. This was three summers ago, but it's possible that the boards are still somewhere in my son's garage, and if we can locate them, I'll make them available to the forum for testing.

Your entire message was very interesting and enlightening. I know exactly what you mean. Every time I thought I'd changed something for the better I found-out later that it was much better if I put it back the way God and AR intended.

The quoted section, above, makes me very hopeful. There is a reason this build-up method doesn't work and identifying it may help solve many other mysteries. I've got all my fingers crossed that you can find them.

Bu t Ittt IS hArd To T Ype LiKE This.

Bret

[Kuja]

Wow!

All this info came just in time, since I'm about to start working on LF part of my AR-90 crossovers.

So far, it seems that everybody agrees that building up those big LF caps is a big no-no.

But, what about the lower/upper midrange and tweeter caps?

I had to build up those.

It seems that those cap values AR used back in the day, can't be found today.

At least they are not available in my country.

I tried Clarity Caps, ICEL, Audyn and Mundorf and 90% of those original AR-90 capacitor values are not present in their current catalogues.

So I had to build up almost all values.

I have found a single bipolar 330 uF cap for a LF crossover section, but what about the other caps?

Here is the picture of my new upper crossover board.

What do you think?

[Diamonds&Rust]

So I had to build up almost all values.

I have found a single bipolar 330 uF cap for a LF crossover section, but what about the other caps?

What do you think?

I haven't experienced "the problem" with small built-ups using as little as three capacitors and I've never heard it using two.

If you hear a problem, we'll look for a culprit. But...

I think you're going to like this half of your rebuilt crossover. I'll bet it sounds terrific just like it is.

[Kuja]

Hi there again!

I went very carefuly through this thread about AR9 recapping again, reading about those bad experiences with using built up LF caps in AR9.

But then I realised that there are no actual experiences with built up 350uF LF cap in AR90!

Are we only assuming that the same bad thing will happen if we use built up LF caps in AR90?

I checked the AR-9 schematics that I've downloaded long time ago from the library.

It seems that the LF crossover sections are VERY different between the AR9 and AR90 models.

In AR90 there is only one 350uF cap in parallel with the woofer.

In AR9 we have that huge 2500uF cap (paralleled with some inductor) which seems to be connected in series with the woofer and then we have this 470uF cap connected in parallel.

I attached crossover schematics from both models.

So, are there any real world experiences regarding those same problems using built up LF 350uF cap in AR90, or we can only make an educated guess that it will happen too in AR90 case?

Thanks again for the great write ups!