Results: Rebuilt AR9s: cabinets, drivers, cross-overs, wiring

[kkc]

About two years ago, I purchased some AR9s. I was hoping that I would not have to do any DIY, but when they arrived in Sweden, there were a few problems. Inside the cabinet, some heavy inductors and caps had torn loose and rolled around. Since I had access to the XO I took some measurements and realized that many of the caps were completely useless. On top of this some drivers had not been refoamed correctly ... of course it had to be the big 11-inchers...! And one of these even had separation of the paper at the coil area.

Clearly these speakers were begging for some major rework. This is what I did:

1. Replaced the woofer with the separated coil with a NOS identical - thanks to the seller sending me a replacement
   
2. Refoamed those drivers that needed refoaming
   
3. Removed all stuffing, added braces between the tweeter and UMR and behind the LMR. All braces were basically two 1 inch rods crossing at 90°.
   
4. Changed all wiring. Rubbed down the oxidation on all contacts. I used 0.8mm dia. wire for the tweeters, 1mm for the UMR, 2mm for the LMR and the thickest, heavy OFC for the woofers. Gold plated spade connectors were used
   
5. I have outboarded the cross overs - see pictures below. Changed all caps and resistors. I upgraded the switches to heavy duty marine types... these are overkill at a rating of 15 or 20A, but I wanted switches with a high contact pressure. I was tempted to leave the switches out, but I am glad I didn't. I find I use them when listening to CDs that have way too much high end for example.
   
6. All driver cages were damped down with heavy bitumen. The bitumen is about 4mm thick so at the top edge of each driver, you have to thin it a bit so that the driver will sit snug in its cabinet opening.
   
7. Bitumen was also applied on the outside of the LMR compartment. I resisted to place bitumem on the inside of the main AR9 cabinets, but if I ever open them up again, I would not hesitate to add bitumen to the inside of the main cabinets as well.
   
8. By the way, it's really easy to place the wool back into the cabinet. Basically there are four parallel rolls in each cabinet. Remove these carefully so you do not destroy the roll structure. When you want to put them back in, get some smooth, flexible, low friction plastic sheeting. Roll up the wool in the sheet and then pipe the wool + the sheet into the cabinet. Gently pull out the sheet. Repeat for remaining three rolls. Using the X braces you effectively will have four channels in the cabinets... so it's very easy in the end.
   
9. For the components - I reused the inductors. The caps I replaced with combination of Clarity SA, Mundorf Supreme and Mundorf Electrolytic caps. The two big woofer caps and the one big LMR cap were replaced with Mundorf electrolytics but then bypassed with some polypropylene caps to make the driver responsiveness faster. I had to bundle four Mundorfs plus one polypropylene cap for the subwoofer circuit. Since many on the forum had expressed negative experiences, I also secured the old 2500uF in the new cross over box .... but as it tunrs out the new caps are just great and I will not need to use the old 2500uF caps. I also bypassed every cap sub-circuit with a Vishay MKP 1837 0.01uF (9 per speaker) as these are supposed to clear up the harmonics (I think mostly on the high end). The resistors were all replaced with the Mundorf premium ones... I got these cheap, so why not. Most components were connected using screw terminals.
   
10. All components were mounted on cork board with silicone gel ... basically to reduce vibration.
    
11. All electrical connections were sprayed with anti-oxident lubrication spray.
    
12. I used hard rubber balls cut in half (four halves to each speaker) under each speaker to isolate the speaker from the floor for better bass and LMR. Partially countersunk dowels ensure that the rubber halfs stay in place and not move around, by connecting the rubber half with the screw holes in the base of each speaker.
    
13. I replaced the sticky foam rubber gaskets between all drivers and the corresponding cabinet mating surface.
    
14. Finally I removed the silver trim around the sub-woofer grills - the ARs look a lot nicer now.
    

So how do they sound.....!

They sound really, really good. The electronics are still breaking in. The bass is all there... I can shake, rattle and roll if I feel the need. I have Channel Islands D200 and D500 monoblocks, fed by a passive preamp. The D200s pump 200w into 8ohms and the D500 about 800w into 4ohms. Both amps have absolutely no problems driving the AR9s and similarly the AR9s will take anything that the D500s will put into it without breaking into a sweat.

I am still breaking in the speakers and the D500s (just bought them a month ago), so the sound is still changing but all in all:

Highs are very clear and airy, the mid range is fantastic, full of texture and nuance. Voices are lifelike as are organic/real instruments - violins, guitars, pianos, real driûms sound so incredibly real and full of texture it's amazing.

Lows are bloody good......movies are fun to "feel". I was rattling the windows even on a 2/10 setting at the lowest frequencies on a test CD. These are frequencies I cannot hear but I some how feel. One can definitely feel the low frequencies. Whats great is the control that the amps exert over the woofers. It's tighter on my smaller AR94 speakers, but its seriously powerful on the AR9s - although not as sharply controlled as I would like. I guess this is due to the heavier drivers and the BIG inductors. The polyprop cap does help with speed however.

Things I would now do differently (and may do in the future anyway...!):

1. add bitumen sheeting all over the inside of the big cabinet ... not just the LMR sub-box and the driver cages - make the cabinet heavier and "deader"
   
2. consider adding one more X brace just above the woofers - I didn't do it because I was afraid of taming the bass ... but I needn't have worried
   
3. try to source OFC wiring for all drivers - not just the sub woofers. The 2mm LMR wiring is way to stiff. I would use flexible multi stranded OFC for LMR.
   
4. I used heavy brass terminals for connecting the drivers to the outboarded XO... I am not happy about this but the copper terminals are very expensive. I used copper for the speakers cable connections.
   
5. Depending on how much I use the switches, I may do away with em --- the nice thing about an outboarded XO is I can make mods with just a few minutes and a screwdriver.
   

All in all these are really amazing speakers. I can see why AR were comparing real live performances. The horn/dome UMR does have a distinctive sound - which is good and bad. But lets see how with some more burning in the speakers settle down.

Thanks to all the foum members for your helpful posts and comments. Thanks also to Dusty, the designer of the amps for tons of advice.

AR9_crossover_circuit_2010_02_27_v0.5_kkc.pdf

kkc.

[kkc]

Thanks Howard.

By breaking in, I am principally referring to the capacitors, although to an extent it will also apply to drivers. On my AR94s, I replaced the tweeter and even right out of the box, they sounded very good, but a year on the difference is huge. Now they sound fluid, wide open, balanced, sweet and incredibly musical.

On the AR9s, I suppose that the drivers may loosen up because they probably had not been used for a few years but mostly because of the new foam surrounds.

There does seem to be a difference before and after using a demagnetising CD e.g. Densen audio's CD, but it's hard to say if that's because it really does demag the circuitry or because the drivers get a good thrashing for 3 mins or maybe even both.

In any case, I was really referring to the capacitors. Generally during the first year or so the materials inside e.g. teflon will change properties as the caps are fed with electricity. In the last month, my speakers have had major changes in the way they sound. At first the AR9s were just a shade better then the AR94s, then they cleared up. Also in the beginning the bass was not so good. Now they are a lot more detailed ... and even sound a little bit compressed.... but they will change in the coming weeks and months before settling down.

I think I will only know in a years time if I want to change any more things.

I suspect I may run into issues where the drivers weaknesses start to show up as I have used pretty good quality caps. We'll see.

I think generally manufacturers have a tough time with burn-ins. It costs money, means manufacturing space is occupied by finished goods that could be shipped out.

Regards, kkc

[dingus]

wow, very nicely done! i am thankful that my 9's needed only simple driver replacements.

what was your reasoning for relocating the crossover external to the cabinet? was this made necessary by the added bracing, or was it part of the plan from the beginning? i too found that removing the chrome from the woofer grills to be a cosmetic improvement. as for the drivers, the tweeter in the pic looks like it could be a non-original replacement (i dont see the tell-tale white ring around the dome).

[kkc]

what was your reasoning for relocating the crossover external to the cabinet? was this made necessary by the added bracing, or was it part of the plan from the beginning? i too found that removing the chrome from the woofer grills to be a cosmetic improvement. as for the drivers, the tweeter in the pic looks like it could be a non-original replacement (i dont see the tell-tale white ring around the dome).

Hi Dingus.... I should be thanking you. I found your posts and responses very helpful...!

Regarding the tweeters, they do have rings, but the rings are under the foam and closer to the outer edge of the metal plate that holds the tweeter. I was told that they are original. Other than that I don't know anymore. I can post a photo if it's helpful for anyone.

Out boarding the crossover was planned from the beginning for a number of reasons:

1. I knew that the new caps would never fit in the AR9 without significantly changing volume. The polypropylene caps are pretty big.
   
2. I wanted to have an easy ability to tinker with the XO - especially after many of the comments on the sub-woofer cap replacements
   
3. i wanted to isolate the AR9 vibration from the XO
   

The downside of outboarding the XO is that the wiring gets longer, it looks messy and I have to use brass connectors which is yet another point in the signal path.

Thanks, kkc.

[dingus]

i dont know how i could have helped, but i'm glad if i did.

i understand your choice with relocating the crossover now. though as you say a bit messy, it is much more convenient. as to the tweeter, i am confident that it is original. to make sure you can check the part# stamped on the back - 200029-1.

[rrcrain]

Nice work. Very nice, actually.

Regarding the break-in concept. I assume that breaking in involves the speaker drivers "loosening up" to some extent, until they reach a point where the drivers meet specifications. What I am wondering is: what causes the drivers to stop loosening up at just the right point? Why don't they just keep getting looser and looser as the system is played over longer periods of time until trouble develops? And how can manufacturers guarantee that drivers meet factory specifications (particularly when good driver manufacturers will offer up a full list of driver parameters when they build and sell the things) if they will not reach that point until hours after the speakers have been played in their new listening rooms? What stops the break in change?

And if the break in does not involve the drivers "loosening up," just what does it involve? Does the magnetic field evolve? Do the metal parts change in some way?

Finally, most speaker rebuilders use measurement tools to ascertain if things have worked out with the rebuild. It appears to me that one way to evaluate the break-in phenomenon would be to measure the drivers right after construction and then do it again after the break-in period. That way, one could actually see measurements that determined the effect of breaking in a speaker.

Howard Ferstler

A number of weeks ago, I read a news article about some research researchers were doing using dogs and their vocalizations. They recorded various canine vocalizations during play as well as agressive behavior over food and toys. To their ears, everything sounded identical. More importantly, none of their test equipment cold detect a difference. Yet, when played back, the dogs reacted differently when an agressive vocalization was played back versus a play vocalization.

What this tells me is that in spite of how good our scientific measurements are, they are missing something the ear and brain hears.

[RoyC]

By breaking in, I am principally referring to the capacitors, although to an extent it will also apply to drivers. On my AR94s, I replaced the tweeter and even right out of the box, they sounded very good, but a year on the difference is huge. Now they sound fluid, wide open, balanced, sweet and incredibly musical.

My Toyota pick-up drives like a BMW now that the electrical system has "broken in". Sorry, I'm not buying small differences, much less "huge differences".

Very nice restoration, however!

Roy

[ar_pro]

Nice job, kkc - yours is the first outboard AR-9 crossover mod that I've seen - congratulations on your hard work!

Can you elaborate on the construction of your additional cabinet bracing?

Did you consider spiked feet, instead of the rubber balls?

Have you tried separating the upper & lower crossover sections of each speaker, and stereo bi-amping? I've noticed some very interesting results using low-frequency equalization with this type of configuration.

[kkc]

Nice job, kkc - yours is the first outboard AR-9 crossover mod that I've seen - congratulations on your hard work!

Can you elaborate on the construction of your additional cabinet bracing?

Did you consider spiked feet, instead of the rubber balls?

Have you tried separating the upper & lower crossover sections of each speaker, and stereo bi-amping? I've noticed some very interesting results using low-frequency equalization with this type of configuration.

Thank you ar_pro.

The bracing was applied in two places. I wish I had taken photos. Each pair of braces consisted of two 1 inch wooden rods at 90° to each other, crossed in the middle like a "+". Where the two pieces crossed I had flattened the wood so I could stick them together. The braces were applied between the tweeter and the UMR and one more time right behind the LMR box. I was fortunate to get a very snug fit.

The speakers are still changing quite a lot and I have noticed that I am getting some mid range boominess. I could probably have placed more bracing to make the cabinet even more "dead". I may do that in 1-2 years time.

One more point, I mentioned in my 1st post that the bass was really good and listening to some music this evening I realized that the speakers are about 1-2 feet away from the side walls ... this would serve to amplify the bass. When we move to our new home at the end of the year I am sure I will get different acoustics. The place we live in now has heavy concrete walls, completely dead room whereas the house we will move into will have wooden walls but with 46cm mineral wool insulation and very dense cement/wood fibre boards on the inside.

Spiked feet couple the speakers to the floor and therefore transfer the vibrations to the floor. Since I live in a flat, I did not do that as my neighbors may not be very happy. Using split rubber balls isolates the speakers to the floor and reduces the vibration flow through. On the AR9... because of the heavy bass this was very easy to notice. With no spikes or rubber feed, vibrations did get transferred to the floor. As soon as I added the rubber ball halves... it pretty much disappeared.

It's funny you mention the equalization. I did not do that, but when I connected the speakers and started using them... I really appreciated how much power the four woofers must be drawing and the first thought that occurred to me was that I should have split the circuit. My old DM110i speakers from Bowers & Wilkins had a split circuit and I feel that they made a difference. Here's the nice part ... with the external crossover... it's really easy for me to experiment and I might just do this at a later stage. On the other hand, I am totally amazed by the D500 mono blocks.

These Class D amps are simply the best I have heard, not at all cold or warm. Just neutral and very faithful to the recording. The efficiency of these is above 80% and at 800W into 4 Ohms... that's a lot of power. BTW, thats 800W per mono block. I believe the AR9's are rated at less than 150W but they happily take a lot of power without any distortion.

Cheers, kkc.

[ar_pro]

You're right, kkc - spiked feet would definitely couple the speakers to your neighbor's ceiling, so the rubber balls are an excellent solution.

Thanks for the details on your internal bracing - the AR-9 cabinet is quite rigid, but undamped through much of its height, and I've wondered what effect additional bracing or the application of damping material would have on the speaker's character.

It's counterintuitive (at least by current standards) to think that large areas of undamped surface would be of any sonic benefit, but the proof is in the pudding - the AR-9 sounds terrific as designed, but it seems as if your modifications may have taken it to a new level of performance. If I have the opportunity to rebuild another pair, I'm going to try some of your methods.

The Channel Islands amplifiers are just what this speaker needs - even at moderate listening levels, the 9 has an overall top-to-bottom completeness that is only realized with substantial reserves of power, as your D500 provides.

Interestingly, though, without crossover work (at least replacement of the upper-level caps, and attention to the level-control switches), the system, while handling high levels of power with a significant spl, tends to sound compressed - almost as if it's hit some sort of invisible wall - this has been true of every unrestored AR-9 system that I've heard in the past 15 years, or so.

Your completed re-build and current setup seem ideal - now it's just a question of adjusting those level-control switches!

[kkc]

You're right, kkc - spiked feet would definitely couple the speakers to your neighbor's ceiling, so the rubber balls are an excellent solution.

Thanks for the details on your internal bracing - the AR-9 cabinet is quite rigid, but undamped through much of its height, and I've wondered what effect additional bracing or the application of damping material would have on the speaker's character.

It's counterintuitive (at least by current standards) to think that large areas of undamped surface would be of any sonic benefit, but the proof is in the pudding - the AR-9 sounds terrific as designed, but it seems as if your modifications may have taken it to a new level of performance. If I have the opportunity to rebuild another pair, I'm going to try some of your methods.

The Channel Islands amplifiers are just what this speaker needs - even at moderate listening levels, the 9 has an overall top-to-bottom completeness that is only realized with substantial reserves of power, as your D500 provides.

Interestingly, though, without crossover work (at least replacement of the upper-level caps, and attention to the level-control switches), the system, while handling high levels of power with a significant spl, tends to sound compressed - almost as if it's hit some sort of invisible wall - this has been true of every unrestored AR-9 system that I've heard in the past 15 years, or so.

Your completed re-build and current setup seem ideal - now it's just a question of adjusting those level-control switches!

Interesting comments: I did not place a third brace because I too was afraid of changing any designed in sonics. In the AR94s the subwoofer is in the center of the enclosure (tweeter at the top and fullrange at the bottom). When I upgraded these, I mistakenly placed the subwoofer at the bottom (both the full range and subwoofer are exactly the same size). The bass literally disappeared. I was really shocked. I understood then that the sub is probably using the center position to give it some flex to work with. Given this I think, rather then add a third brace, just applying heavy bitumen on the insides of the cabinet may be better. This would still allow the woofers the side walls to resonate with but with a sharper response, less boom....?

I realized a few days ago that my speakers are about 2 feet away from the room corners. This would amplify the bass quite a bit. When I move to the new house at the end of this year I expect to have a lot more room. By then the speakers will have broken in as well, so I will be able to report a more accurate description of effects.

Compression is something that does concern me... at times the music is too rich... but the caps are still breaking in and I know from the AR94s that a year on and there was a lot of clearing up in the sounds. In the AR94s, I upgrade the tweeters with Dayton silk domes and this completely cleared up the high end scratchiness and compression.

Tony Gee (he makes some amazing speakers) warned me that upgrading caps also can hi-light inefficiencies in the drivers (and cabinets).

I am expecting one of the Goldmund speaker designers to visit me in a few weeks - she is keen to hear the AR9s... so if anyone has any questions or simple measures ... I can ask her. One thing is for sure, the D500s are as perfect/neutral at reproduction as any amp I have heard... so I am reasonably confident that the measures will reflect the AR9s and not the amps. These amps really are fab.

Rgds, kkc.

[kkc]

My Toyota pick-up drives like a BMW now that the electrical system has "broken in". Sorry, I'm not buying small differences, much less "huge differences".

Very nice restoration, however!

Roy

Hi Roy,

For driver burn-in see Driver Burn-In - AV123 Site

For electronics ... there is a reference at the above link... but I am sure there are articles on the web too. I'll post if I find...

The same site also has pretty good description of low bass performance on different types of drivers.

Rgds, kkc.

[RoyC]

Hi Roy,

For driver burn-in see Driver Burn-In - AV123 Site

For electronics ... there is a reference at the above link... but I am sure there are articles on the web too. I'll post if I find...

The same site also has pretty good description of low bass performance on different types of drivers.

Rgds, kkc.

Hi kkc,

I've been at this for a very long time, and the capacitor debates are endless. I have no problem with mechanical changes in a driver over time (for better or worse), but with respect to capacitor "burn in", I am firmly planted on the other side of the fence. There is no way to substantiate the claims you are making.

One of the better debates on the matter took place in the Audiokarma forum awhile back, and starred some of our own esteemed CSP members:

http://www.audiokarma.org/forums/showthread.php?t=240347

BTW, I'm also not one to purchase the expensive caps you used, or fool with the "thickest OFC, heavy" wire for the woofers, etc. Other than using mainstream film caps, I rarely stray from the types of materials originally used. Different philosophy...Whatever floats your boat.

Please don't misunderstand me. I think your restoration is a thing of beauty.

Roy

[kkc]

Hi kkc,

I've been at this for a very long time, and the capacitor debates are endless. I have no problem with mechanical changes in a driver over time (for better or worse), but with respect to capacitor "burn in", I am firmly planted on the other side of the fence. There is no way to substantiate the claims you are making.

One of the better debates on the matter took place in the Audiokarma forum awhile back, and starred some of our own esteemed CSP members:

http://www.audiokarma.org/forums/showthread.php?t=240347

BTW, I'm also not one to purchase the expensive caps you used, or fool with the "thickest OFC, heavy" wire for the woofers, etc. Other than using mainstream film caps, I rarely stray from the types of materials originally used. Different philosophy...Whatever floats your boat.

Please don't misunderstand me. I think your restoration is a thing of beauty.

Roy

Hi Roy... No offence taken and thanks for the feedback. Rgds, kkc.

[djcheung]

Hi KKC,

I know this is a very old post.

I wonder if you remember how to remove the resistor block that have switches attached at the front of the speakers.

Best regards,

David.

[kkc]

Hi KKC,

I know this is a very old post.

I wonder if you remember how to remove the resistor block that have switches attached at the front of the speakers.

Best regards,

David.

Hi David

It is a while back. The selector switches are attached to one side of the same board that also has the resistors. I seem to remember that the block is glued airtight. To remove, just tap away with a tool e.g. a long screwdriver and a hammer. But before final re-assembly don't forget to reglue something back so that there are no open holes into the cabinet - the cabinets are meant to be airtight...!!

If you don't intend using the level switches (frankly, you don't need them...!), you can just keep the resistor block in place and snip all the wire connections.

Good luck. kkc.

[djcheung]

Hi KKC,

Thanks for the advice.

So they are just glued onto the box !!!

When you say tap, do you mean from inside the box after removing the Lower Midrange Unit to gain access?

And then get the screwdriver inbetween the resistor block and the inside speaker wall.

I have just bought these speakers so may want to rebuild the crossovers but I haven't heard them before so don't know

what to expect at the sonic front.

Best regards,

David.

[kkc]

Hi KKC,

Thanks for the advice.

So they are just glued onto the box !!!

No. The switches and resistors are fixed to a small wood panel. The panel is then glued to the main speaker box.

When you say tap, do you mean from inside the box after removing the Lower Midrange Unit to gain access?

And then get the screwdriver inbetween the resistor block and the inside speaker wall.

Just open up and start looking inside. It's very straight forward ... you'll work it out. The glue is 30 years old and somewhat dried out, so comes off easy.

I have just bought these speakers so may want to rebuild the crossovers but I haven't heard them before so don't know

what to expect at the sonic front.

Best regards,

David.

[harry398]

I know this is a Older rebuild.  Great job!  great job showing how the caps were Off spec.

So I am wondering if you are still enjoying the AR9?  how is it?

Did you get the woofers to seal tightly?

Harry

[kkc]

ty harry 398.  yes the speakers are great.  sound is very good.  sealing the woofers is not at all an issue.  The glue is very tacky and seals and dries extremely well. The felt or sponge between the metal flange and the wood panel also provides a good seal.  Changes I would make:  solder the components.  I used screw connectors to connect and this still allows oxidation to take place in between the touching wires.  However be aware that when soldering a lot of wires into one touch point or when soldering thick speaker wire (as I used for the woofers) one would need a high power soldering iron.  A low power will not be able to heat up the copper wire enough.  to protect caps, a heat shunt between the solder point and the cap itself would be a good idea.  I quite like the Mundorf silver based solder.  Alternatively, if you can make a spot welder from microwave oven parts (it 's dangerous - your choice, your risk) then you could possibly spot weld the wire connections without any solder. I would also use full copper connectors on the speakers connectors (I just used pure copper connectors on the two main speaker posts).  Another experiment is to install an omnidirectional tweeter on the top.  I hear these are really nice at spreading the high notes.  No idea how they would sound with the ARs.  I have a set of AR94 (smaller brother to AR9) and these sound amazing too. Positioning, amp pairing, cable all make differences.  If replacing the tweeters, I would use the Israeli tweeters.  The AR94Si rebuild is here: