AR 9 Cabinet & Cross-over upgrade

[Guest dean128]

G'day,

Firstly, i'd like to thanks Dale for his help in most previous post. I am new to this forum so i got a bit confused.

I have been advised to upgrade my AR9s cross-over ( capacitors ) as well as lead-lining the cabinet to make it heavier and inturn reproduce more accurate low frequency reponse.

However, from inspecting the cross-over. It would be such a waste to throw away those beautiful, huge and quality original Caps. I have read an upgrade on AR94 on TNT site in which mentioned that "...old Caps dont cut it no more..." and therefore, should be replaced.

If anyone has done any upgrades to your AR9s and would like to share your experience that would be GR8.

Thanks & Look forward to your replies.

Dean

[tysontom]

>G'day,

>

>Firstly, i'd like to thanks Dale for his help in most previous

>post. I am new to this forum so i got a bit confused.

>

>I have been advised to upgrade my AR9s cross-over ( capacitors

>) as well as lead-lining the cabinet to make it heavier and

>inturn reproduce more accurate low frequency reponse.

>

>However, from inspecting the cross-over. It would be such a

>waste to throw away those beautiful, huge and quality original

>Caps. I have read an upgrade on AR94 on TNT site in which

>mentioned that "...old Caps dont cut it no more..." and

>therefore, should be replaced.

>

>If anyone has done any upgrades to your AR9s and would like to

>share your experience that would be GR8.

>

>Thanks & Look forward to your replies.

>

Dean,

You are right that many on this forum have made note of possible problems with crossover capacitors, especially the electrolytic caps of the type used in the AR-9. Quite frankly, if you feel that they are bad, or that they need checking, remove them from the crossover (or test them in the speaker by removing at least one of the two leads) and test or have them tested with the proper equipment to see if they are in tolerance. All (non-leaking) crossover capacitors that are within approximately 10% of their stated value are easily within tolerance, as this was the original spec. I would not change them until you have verified, objectively, that they do not meet spec.

Lead-lining the cabinet. Well, in my opinion, this is probably a huge waste of time and money, because the cabinet is already well-braced and relatively acoustically "dead" for all practical purposes. Lead-lining is akin to using #4 welding cable for speaker wires, and jacking these wires up in the air with sections of tennis balls in the fashion that was mentioned by Laura Dearborn in one of her hi-fi articles or in her book. Supposedly this enabled the current and energy to flow more freely through the wires, or something to that effect, with little resistance to flow. The cable analogy is a bit "tongue-in-cheek," but lead-lining the cabinets will probably never make an audible improvement unless the cabinet is poorly designed to start with. It could give some psychological boost, but little else.

Another thing to remember: the AR-9 has a definite interior volume, bracing and a specific amount and position of fiberglass stuffing. This should not be changed or altered. The speaker's resonance frequency and the special crossover-resonant circuit are interdependent in the AR-9; and if you alter any parameters, the speaker's bass response (and possible impedance) will suffer accordingly. As is usually the case, you are better served to try to repair anything that is broken, and insure that the speaker is left pretty much the way it was designed. It is an outstanding loudspeaker, and if you keep it the way it was designed, you will be able to enjoy those benefits.

--Tom Tyson

[rrcrain]

Most here probably know what my opinion is, but I'll repeat it. After 20 plus years, its highly unlikely that the original electrolytics are still good. They "might" be within spec, but their dissipation factor and other parameters probably have gone to heck.

I rebuilt the crossovers in my AR90 speakers and the results are nothing short of fantastic. Others have noted similar effects rebuilding the AR9 crossovers.

[johnieo]

There is absolutely no need to replace a device that is within spec. I submit that the specs on an npe are C, ESR, and voltage breakdown. Either increasing ESR (age), or decreasing ESR (replace with thin-film) can alter the results of the speaker's designer's hard work; the seriousness may depend on the complexity of the crossover network.

The real, practical issueis that inexpensive capacitance meters do not measure DF (ratio of ESR-to-reactance at measured frequency). Unless we work in a lab, have a good friend in a lab, or have a rich uncle, this parameter goes unmeasured.

[rrcrain]

There is a relatively recent post in the AR forum talking about your concerns. This gentleman rebuilt one speaker with poly caps, the other with good electrolytic caps. He admitted up front that he did not have the means to scientifically measure the differences and had to depend on his ears. Personally, I'll depend on my ears rather than a scope and analyzer.

In a nutshell, what he discovered was that the speaker with the rebuilt crossover using current technology caps (poly) sounded exactly the same as the other speaker rebuilt with "original equipment" caps. The main point is the anticipated lifetime of an electrolytic versus a poly cap. Most electrolytics have a shelf life of perhaps 5 years versus several decades for the poly.

Since AR, as all other speaker manufacturers did at the time, used electrolytic caps, and electrolytic caps have a shelf life of perhaps 5 years, its a good bet that a 20+ year old electrolytic is beyond its prime. Do keep in mind that manufacturers went to electrolytics due to economics and limits in technology at the time. The old caps using PCBs were banned, and those puppies supposedly were on par with todays poly caps.

[Guest Sean]

RR, let me start off by stating that i'm on your side : )

Having said that, electrolytics will last for many, many years and stay within spec. That is, so long as they are not subjected to extreme levels of heat and / or major temperature swings in a short period of time and / or high levels of in-rush current on a repeated basis. These are the three things that cause the most stress to any electronic component. I just pulled and tested the power supply electrolytics from a high bias amp that i have and they tested fine. This amp was made in 1993 and idles at 130* F. Unfortunately, the caps in my 9's and 90's weren't in quite as good of shape, so i replaced all of them. Then again, the 9's and 90's are appr 25 years old now, so you can expect them to be in less than optimum shape.

As i've mentioned before in these forums, upgrading the caps on the 9's, 90's, etc... is a very worthwhile venture. Not only does it bring the speaker within spec, it typically helps to level out the tonal balance a bit. If one is honest, and even if they aren't, it is quite obvious that most AR products have a relatively "warm, smooth and full bodied" sound to them. If one studies the frequency response charts that AR supplied with the 9's, 90's, etc... one can see that the lower frequency range is slightly elevated above that of the mids and treble. In effect, the response really is "full bodied" due to the elevated low frequencies, particularly in the "warmth" region, and "smooth", due to the slight shelving of the upper frequencies that takes place. By using caps such as Solen's, etc.. which are known for sounding a little "bright" ( we can alleviate much of that ), one can actually produce a speaker that sounds more open, spacious and may actually measure more linearly. To those that laugh or ridicule such statements, how many multi-thousand dollar sets of speakers have you done such work on and compared the results to???

As to Tom's comments about the AR 9 and 90 cabinet having "plenty of bracing" and being quite rigid, i disagree. These speakers have ONE brace in them and it is located just above the woofers going from side to side. While this is the place that a brace is most needed, the cabinet is quite flexible and resonates like mad due to the lack of bracing / internal damping. If you doubt this, pull the woofers and damping material out of the cabinet and then knock on the box. It's not near as rigid or "dead" as you might think. As such, i would recommend adding a brace from the baffle to the rear panel and from side to side up higher in the cabinet. If you do things right, you can tie these two braces together, adding even more rigidity to the cabinet. The use of wooden "closet rod" material works great and doesn't eat up much box volume at all. I place one of my braces going from the very top section of the 8's chamber to the rear panel and then ran another brace just above that from side to side. Since "closet rod" is 3/4 round, i was able to situate the braces where the two flat sections crossed each other. Not only did i glue each brace to the cabinet on each end, i glued the two together where they crossed on the flats.

Besides adding the bracing, sound damping sheets won't hurt anything and will help lower the Q of cabinet resonances. I would avoid working with lead and use some inexpensive but reasonably effective damping sheets that Parts Express sells. Try either 268-010 or 268-030. You can easily use 4 - 10 sheets per cabinet depending on how "gonzo" you want to go. My suggestion is to stagger the sheets apart a bit at random distances, which will help break up any major nodes that might be occuring.

While we are talking about damping resonances, here's one of the biggest tricks that you can do to drastically improve ANY speaker that uses stamped baskets. That is, apply material to the basket to damp the ringing that takes place. If you doubt this is happening, take the drivers out of the box and "flick" the basket with your finger. While the 12's will ring quite noticeably, the 8" driver will sing like your playing a bell. Personally, i've used "liquid nails" type caulk for this and it works wonders. If you're going to do this, spread a layer on the basket, let it set up till hard and then spread another layer on. If you doubt that this has any validity to it, try performing a frequency sweep on the driver before and after "damping the basket" and look at the difference.

As far as upgrades go, get rid of the factory wiring and bypass the factory attenuation circuit. These speakers already sound "warm and full", so you shouldn't need to attenuate any of the upper frequency ranges at all. This DRASTICALLY opens up the sound and shortens the signal path. You get rid of several feet of low grade wiring, the resistors and the switches. ALL of the lower midrange, upper midrange and treble signals had to pass through those junky switches, so get rid of ALL of that crap. Besides that, it makes the system SO much easier to re-wire it's not funny. You can leave the switches and board in place for cosmetics, but bypass all of that electrically.

As far as wiring goes, use the wire of your preference. I would recommend NOT using "zip" type cord as it is highly inductive and rolls off the top end. While i don't consider it to be the finest cable made, i used Kimber 8TC for the woofers and lower mid and Kimber 4TC for the upper mid and tweeter. This cable is obviously more than up to the task and i had GOBS of it laying around. If you want something cheap that works reasonably well, try using some Carol wire and cable that you can get from Parts Express. Use 100-728 configured as a Star Quad for the 12's and 8's and 100-710 for the domes.

Now that you've gotten that far, replace those crappy binding posts that these speakers came from the factory with. Try some of the larger all metal units from Axon or from Parts Express 091-340's. This way, you can use some real wire like the AR's deserve.

If you're really doing all of this, take the time to lay out hte crossover parts more efficiently on the board. If you're really careful, you can condense the two boards down to one AND reduce crosstalk between the various circuits. AR laid out some of the coils right next to each other on the same axis. This allows energy from one band to magnetically couple to another band via the inductive action of the coils. Coils CAN be placed in close proximity to each other, they just need to be oriented in different planes. That is, if one coil is placed with the windings laying flat, get the other coil as far away as possible and stand it on end. There have been articles published on this and just re-arranging the inductors like this was responsible for a reduction in crosstalk by appr -35 to - 40 dB's. This is a very sizable improvement in performance to say the least, which can only result in better sound and increased power handling.

Last but not least, make sure that the cabinet is sealed as best possible. In order to do this right, i replaced the rear plate that the binding posts were mounted to with a much thicker and sturdier board. Due to improper sealing from the factory and / or the antique glue that they used drying out, i had a very sizable air leak on my 9's and 90's in that area. Replacing that board with someting sturdier cured that problem since it was no longer able to "flex to the beat". As a side note, this board was what i used to mount ALL of the crossover parts on, giving me a direct point to point wiring arrangement from the binding posts to each section of the crossover.

If one does all of the above, the results are quite staggering to say the least. Your "antique" AR's, which are already pretty good speakers to begin with, will literally trounce many speakers that cost $10K or more. Feed them with some REAL wattage and you'll be amazed with the finesse, grace, detail and power that these "old pieces of junk" can deliver. Sean

>