AR surround

A colleague in town has giant Legacy Whispers driven by McIntosh 500 watt monoblocks. A magnificent system, but the Whispers are a very unforgiving speaker. Specifically, they expose everything in excruciating detail...the good and the bad. Some stuff sounds downright ugly.

Bret (Diamonds & Rust) had recommended using bypass caps across every cap, so I did so. I added the bypass caps afterwards and I could tell that they made a difference on both the UMR and tweeter. The improvement to the tweeter was more subtle than on the UMR which sounded significantly better than without the bypass caps. Member DavidR added bypass caps to his AR91's and said that they made a remarkable improvement. The Dayton bypass caps that I used cost $0.74 each, so it was a small investment with big returns. Had the experiment failed, I could have snipped them out yet not spent a lot of money.

Here is what I used on my AR9's and I am extremely happy with the results. The Dayton F&F Bypass Caps across the high range series caps made a difference especially with the UMR 24uF E-Cap. I do not know why as I own no sophisticated instruments with which to take measurements. 4uF: 4.0uF Dayton PMPC 250V 1% + Dayton F&F 0.01uF 400V 5% bypass capacitor 6uF: 6.2uF Dayton PMPC + Dayton 0.01uF bypass 24uF: 22uF + 2.2uF Mundorf E-Caps + Dayton 0.01uF bypass 8uF shunt cap: 8.2uF Dayton PMPC + Dayton 0.01uF bypass 40uF shunt cap: 40uF Dayton DMPC 250V 5% + Dayton 0.01uF bypass 80uF: 82uF Jantzen Cross-Cap 400V 5% + Dayton 0.01uF bypass 30uF shunt: Dayton DMPC + Dayton 0.01uF bypass 470uF: 470uF Mundorf E-Cap + Dayton 0.10uF bypass 2500uF: 3x680uF Mundorf E-Cap + 1x470uF Mundorf E-Cap + Dayton F&F 0.10uF bypass

Aadams, Did angling the Boston A40's away from each other eliminate any problems with comb filtering? (Is this the same as lobing?) I ask because once I had what turned out to be a not such a good idea. I made a larger channel speaker with two Boston A40's, but the comb filtering proved to be a disaster. AR

Actually, I would have used a $1.50 100uF +/- 10% NPE from Parts Express. I've got P.E. 150uF + 200uF NPE bundles for the 350uF woofer shunts in the AR90's.

I tried Solen caps only in the 24uF UMR spot in the AR90 and AR9, so I have no evidence that substituting a 25uF Dayton PMPC would have yielded different audible results. I wish I had read the audio forum fluff, or remembered the PM from Bret back in 2007 bemoaning the use of Solen caps in his AR10pi's, before I bought the Solens. I do know that the 22uF + 2.2uF combo of Jantzen Cross-Caps sounded substantially better than the single Solen. However, I only achieved excellent results when I heeded Carl and Roy's suggestion of trying a NPE and switched to Mundorf E-Caps. I suspect that the Jantzen Cross-Caps sound fine in the 24uF spot on the AR915 Midrange because the crossover differs from that of the AR9 and AR90. Or perhaps it is because I added the Dayton F&F bypass. Or it could be as simple as the height of the midrange dome in the AR915 sits lower than that of the AR9 and AR90 and fires vertically off-axis for a seated person.

If you intended to get me riled up ra.ra, I congratulate you on your excellent work. I installed those 24uF Solens in both my AR9's and AR90's UMR circuits. If you want to see where this is all written up, see the thread, AR90 Crossover Recapping - Taming the UMR: http://www.classicspeakerpages.net/IP.Board/index.php?/topic/9406-ar90-crossover-recapping-taming-the-umr/ I had a bone to pick with those four Solens for which I paid something like $50 for the lot. So they were either going to make themselves useful ganged together as the woofer shunt cap in this AR915 or find themselves subjected to capacitor "burn-in" over my fire pit!!! The Jantzen Cross Caps while also somewhat problematic for me in the AR9 and AR90, seem to be OK in the AR915 24uF spot with the Dayton F&F bypass cap and 2 ohm attenuation resistor installed. If you want to know why I used Dayton F&F bypass caps (at under a buck a piece), take a look at the May 29th post from DavidR in this thread, Latest Acquisition - AR91s: http://www.classicspeakerpages.net/IP.Board/index.php?/topic/8163-latest-aquisition-ar91s/&page=2 So that is the saga of how I ended up using 15 caps and an extra resistor in my AR915.

I recapped one AR915 which is now in service as the center channel speaker between the AR9 fronts. The AR915 sounds superior to my last center channel speaker, an AR48s. Having drivers essentially matched to the AR9's along with the larger size of the AR915 obviously has helped. The details: - 100uF woofer shunt cap (lower right in photo) is an assembly of four 24uF Solen MKP's plus a 4uF Dayton DMPC plus a 0.10uF Dayton F&F bypass. The Solens were first employed as the 24uF UMR series caps in my AR9's and AR90's but proved incompatible. Had they not been sitting around in the parts bin, I would have gone with a Bennic 100uF NPE. - 40uF spot is a Dayton DMPC plus 0.01uF Dayton F&F bypass. - 24uF midrange spot is 22uF+2.2uF Jantzen CrossCap combo originally tried in an AR90 plus a 0.01uF Dayton F&F bypass. - The other two caps (4uF and 8uF) are Dayton 1% PMPCs plus 0.01uF Dayton F&F bypasses. - The midrange is attenuated -3dB by using a 2 ohm resistor (lower left corner) in series with the 24uF cap bundle. Deciding to install this resistor right off the bat was an educated guess based on a lot of trial and error on the AR9 and AR90 recapping efforts.

Thanks ra.ra. I found the drawing. I didn't notice it earlier because the lead-in starts with AR58. The connection points are clear. Both crossovers in my AR915's have only two resistors; it's just that the left speaker has a third resistor sticking out in space and the right speaker has the second resistor disconnected. It should be an easy path to recap / reconnect everything now. I just don't understand how these crossovers got so mucked up.

I just acquired a pair of AR915 speakers in pretty good cosmetic shape. The AR915 is similar to the AR91, but without midrange and tweeter switches and has fake veneer. I opened them up to do an inventory of parts that I need for a recap and found some problems. The first photo is of the left speaker crossover. It shows a 2 ohm resistor hanging in space as shown by the arrow. It appears that another resistor (1 ohm) has been substituted in its place. Anyone know if AR may have done this? Or is it more likely that someone messed with the speaker in the aftermarket? Note also that two of the caps are Mylars rather than Callins NPE's. The second photo is of the right speaker crossover. It also has a 2 ohm resistor hanging into space as shown by the arrow. However, there is no 1 ohm resistor substituted as with the left speaker. Also, there is only one Mylar cap in this unit. I find this bizarre since the serial numbers are only 3 units apart. Looking at the AR91 crossover schematic - included here for reference - it appears that dangling end of that 2 ohm resistor should connect to the terminal circled and marked "T" in the photo. Can someone correct me if I am wrong? Lastly, does anyone know how to remove that blob of glue in the middle covering the binding posts? Soften it up with 91% isopropanol perhaps? Thanks for the help.

Tom, I agree with DavidR that some technical info on the watershed speakers (AR-1, AR-3/3a, AR-LST, AR-9, Magic, etc.) should be included in the book; perhaps as appendices so those not interested could skip that material. I would expect that your publication would muster even more interest in our beloved classic AR speakers.

Having achieved success with the AR90's, I finished the rework on my AR9's last week: - Replaced the Solen 24uF cap in the UMR with Mundorf E-Caps. - Replaced that 6 ohm resistor which is in series with the choke across the UMR. - Replaced the resistors in the attenuation network. I installed the same value attenuation resistors in the AR9's to yield the same results as with my AR90's: -1.75dB and -2.5dB attenuation on the tweeter and -3dB and -4.5dB attenuation on the UMR. If I were to do this over, I would change the AR9's UMR resistors to provide only -1.5dB and -3dB of attenuation. The balance of the AR9 is very different than that of the AR90's. So although I prefer -2.5dB attenuation on the AR90 tweeter, I like the AR9 tweeter set at 0dB. With the UMR, I use -4.5dB attenuation on the AR90 but find even -3dB to be too much of a reduction for my taste with the AR9. - Replaced the big 2500uF and 470uF Callins woofer caps with bundled Mundorf E-Caps. These reworked AR9's are very nice and produce a really big sound. The AR90's also produce a big sound, but the 9's are really over the top.

Thanks to all of you for your kind comments and for all of your help. I could not have gotten through this without the input and advice of those of you who enthusiastically participate in this forum. John

Because the vendor didn't have the 12 ohm Mills in stock.

Many of you recall from my post titled "AR-9 / AR-90 Crossover Re-capping Adventure" that I had a heck of time trying to find the proper combination of the attenuation resistor switches to achieve acceptable results. People probably thought I was crazy or losing my hearing. The culprit was the 24uF Solen MKP 400V capacitor installed inline with the Upper Mid Range (UMR.) It was all over the place depending on program material. It sounded anywhere from great to shrill and exhibited the uncanny ability to drill down into a recording, extract whatever grain is there and put it front and center. This cap seems to be incompatible in this particular circuit with this driver. Stimpy suggested adding a Dayton 0.1uF F&F bypass cap as a fix. Doing so decreased the shrillness of the Solen, but also increased the grain to the point of rendering many recordings unlistenable. My goal is not simply to obtain the best sound from the best source material, but also to get acceptable results from average recordings. So this very unforgiving capacitor had to go. I decided to try one other poly cap before going to an NPE as Carl, RoyC and others had suggested. I had read that the Jantzen Cross Caps are considered comparatively smooth sounding, so I installed them in place of the Solen. Although the improvement was a step change, I still wasn't quite where I wanted to be. A 16 hour burn-in period using high volume pink noise made no difference. So I installed Mundorf E-Caps (Plain) in matched pairs purchased from Sonic Craft: http://www.soniccraft.com/index.php/mundorf-ecap-ac-c-23_318_358 Like magic, everything was now where I wanted it to be and has remained so after several days...no "it sounds different on Tuesday afternoon than it did Monday night;" no waiting for burn-in; no endless effing around with the attenuation switches. Great recordings still sound great, average material is very listenable, and I perceive no loss of detail in going from the poly cap to the Mundorf. Remarkably, the Mundorf sounds like it has more "air" and "depth." I surmise that this perception is related to how the E-Cap integrates into the entire speaker system rather how it directly compares to the poly caps. Notes: 1) I confirmed with Carl that the E-Caps with their 70VDC/50VAC rating is a suitable replacement for the stock 100VAC Callins NPE. 2) The 24uF Solens were still installed in the AR9 surround speakers and got the 16 hours of pink noise burn in. It still isn't acceptable but for some unknown reason has always sounded more tame in the AR9 than the AR90. 3) I use -4.5dB (estimated) attenuation with the E-Caps in the UMR. I've always used -3dB going back to 1980. The attenuation network is now set up so that I can go back to -3dB if desired. The AR9's are next. I may try the Cross Caps first just as an experiment before going to the E-Caps.

Perhaps the 6uF Callins capacitors in the tweeter signal path failed altogether. One went bad in my AR9's, but I was getting very low output and some low level noise from the tweeter as opposed to no sound at all. The same capacitor in the left speaker was incredibly off spec (400%) but still hanging on for dear life. I understand that the 6uF Callins are especially notoriously for going bad. I don't know how long it would take for those caps to die altogether or if they would eventually cause the tweeters to fry.