Factory Errors in AR 9 Schematic

[Guest Sean]

Upon dis-assembly of my 9's to perform some repairs / upgrades, i started studying the schematic that Mark has posted for these speakers. This is the same schematic that is also found in the original factory supplied owners manual.

What i found is the schematic shows a total of 7 coils in the circuit. If you look at the actual parts layout on the board at http://www.arsenal.net/speakers/ar/documen...ng/Scan15sm.tif there are 8 coils in the circuit. The error lies in the fact that the factory omitted a second coil in the schematic of the high pass section of the 1.5" upper midrange circuit. Judging by what the schematic offers, the high pass section would only be sloping at 6 dB / octave rather than the 12 dB that is stated in the written

text of the owners manual. Doing a side by side comparison of the actual circuitry and the schematic verifies that it really is a 12 dB slope and the schematic needs correction. Quite honestly, i'm surprised that nobody else has ever caught this error before.

On top of this, this specific set of 9's makes use of some slightly different parts values from what is shown in the schematic. I am working on documenting everything that i run into while doing "surgery" so that an addendum can be added to the AR archives on this website. It will be quite some time until until i can get all of this done and Mark can get the info posted, so those that were relying on info that was presented in the schematic should verify what is actually inside your specific set of speakers prior to performing any modifications / ordering of parts.

Please feel free to email me with questions in the mean-time. Sean

>

[Guest]

Sean,

Some years ago I replaced the woofers in my 9s (rotted foams). I then started using a Crown Macro Reference amp (1140 WPC at 4 ohms). This amplifier mastered the 9s, produced astounding bass definition (better than any bi-amp configuration I ever used) and did not run out of gas. Unfortunately, during one of my concert level re-enactments the 470uF capacitor in the woofer crossover exploded. It took me a while to figure where the smell was coming from.

Since then, I have been slowly renovating my 9s. I believe the teledyne AR-90 schematic accurately depicts the lower, upper, and high range crossover layout and values for the 9. The woofer, of course, is uniquely the 9. Thankfully, I discovered and joined this forum which is a tremendous resource.

I am interested in your x-over upgrades. In particular I want to replace the resistors and retain the factory slope adjustments, but am having difficulty finding non-inductive high wattage replacements. Plenty of manufacturers, no dealers.

Also, I have several physically large oil filled AC capacitors and I wonder if these are suitable or preferable to the "fancy" space-age capacitors available now. For example, a 6 microfarad 660V +-6% GE dielectrol cap is 6" by 2" (oval)in an aluminum can. I wonder if this and a 4uF would be okay in the tweeter? I have never seen or heard of anyone using this type of capacitor in a crossover.

Cheerz

Erzatz Mark

[ar_pro]

Regarding replacement caps for the 9...since the enclosure's carefully calculated internal volume must take the size of the crossover components (as well as anything else that is in, or projects into the cabinet) into account, what effect would replacement cans of a different size have on the performance of the system?

[Guest]

Although the oil caps are much larger than the original capacitors, I doubt very much that the bunch of them (4, 6, 8, 30 uF) would amount to even one half cubic foot. I imagine this is at most a 15% change in the internal box volume. The usual engineering variance assumes ten percent tolerances, so I think the capacitor size will be okay, but, the size is something to consider. Note, it would also be possible to locate (for example) the woofer network with its electrolytic caps and chokes external to the box and use the bi-amp input to feed the woofer.

[Guest Sean]

To answer your question about loss of cabinet volume, you can increase the density of fiberfill / fiberglass used within the cabinet to make up for the extra space consumed by larger caps. Slowing the air speed within the cabinet effectively fools the drivers into thinking that the cabinet is larger than it really is. You have to be careful though as this also effects the Q of the woofers at the same time.

As far as bass performance is concerned, this cabinet is already too small for these drivers in terms of ultimate bass extension. I'm sure that AR would have liked to have gone with a slightly bigger cabinet, but back then, the 9's were considered a "huge" speaker. As such, i would try to increase interior volume slightly, not eat it up.

With that in mind, you should be able to find some high grade electrolytic caps with a much lower ESR than what came in the speaker from the factory. At the same time, you should be able to obtain higher voltage parts that are 1/4 the size of the OEM parts used for the bass section of the crossover. A good source for these might be Mouser or Digi-Key. I'm thinking that a specific series of Panasonic caps will work great here. Nelson Pass recommends these for use in his DIY projects and i've used them when upgrading a few of my preamps before.

While my "rebuild / upgrade" is going much slower than planned, so far i've done the following:

1) All of the drivers, wiring, crossover boards, switched attenuators and damping material have been removed.

2) Once you do all of this, you can literally see how much the box "talks" when physically excited through vibration. It is in dire need of damping and internal bracing. While the factory fiberfill material that was stuffed in the top of the cabinet definitely helps, it is not nearly enough to get the job done.

3) With the above comments about minimizing internal space losses, i've installed three braces within the cabinet. I used wooden rods that come in whatever length you like courtesy of Handy Andy, Builder's Square, Menard's, etc.... These are used as hand rails in construction and cost anywhere between $0.80 and $1.65 per foot depending on material and diameter. They are not completely round but flat on one side. I chose 1 5/8" as being the best compromise between rigidity and volume. I think that this ran me about $1.08 per foot. If you duplicate my efforts, you'll need just under 6' of rail total. Just keep in mind that cabinet volume or sizes that i mention may vary ever so slightly due to production tolerances, so measure first and THEN cut. It is better to cut slightly large and shave the ends a bit to keep the braces in place.

4) Let me start off by saying that this becomes MUCH easier if you can CAREFULLY remove the factory installed brace that runs across the top of the woofers. This is glued in place on one side and stapled on the other. Once this is gone, you have a clean shot all the way up to the top of the cabinet. Between half-way crawling into the cabinet and being able to get your hands in through the front holes where the drivers were mounted, you should be able to do whatever you want to inside the cabinet.

5) I chose to mount two braces staggered at various heights within the cabinet. These two run from side to side. These need to be cut to almost exactly 12" in order to fit snugly within the cabinet. While i could have simply wedged these into the cabinet, i applied Elmer's carpenters glue to the ends of the 1 5/8" circular rods and placed a small amount of glue to where they would be placed inside the cabinet. Once this was done, i had to work the brace into position and then wedged them into place. Between the "compression" fit and the glue, they are ROCK solid.

6) I took care not to place either of the braces directly in line with that of the factory brace that is mounted directly above the woofers. As such, none of the braces should run exactly along the center of side panel. Nor should they be spaced at even intervals. I'm trying to break up the nodes along the length and width of the cabinet. As such, placement of the rods ( or whatever you use as a brace ) should be done in a fashion that does not allow any one section of panel to be the same size as any other area of unbraced panel. By doing this, you'll have broken up what was one major resonance into a multitude of resonances of much smaller amplitude.

7) I installed another brace from the front baffle to the rear wall of the cabinet. Since most of the baffle is filled with drivers leaving no space to anchor a brace, i went from the back of the 8's sub-chamber to the rear wall. If i remember correctly, i used a 9 1/2" run of the same 1 5/8" wooden rail and followed the same gluing / installation procedure.

8) When i figured out where the braces were going to go, i saw that i could tie the baffle to rear brace into place and then mount the middle non-factory side to side brace directly below it. This allowed the two braces to intersect. By placing the flat side of the rail on the front to back brace facing down and the flat side of the side to side brace facing up, I was able to get them so close that i was able to glue them together at the intersection. This raises the mass of the brace itself, making it less likely to rattle. It also increases the rigidity of the cabinet as it is now connected front to back and side to side in the middle of the cabinet.

9) Knocking on the cabinet shows that the cabinet is far deader, but still not what i wanted. As such, i installed 5 sheets of 10" x 10" asphault based damping sheets inside each cabinet at various points that were "open" and not braced. 3 sheets were installed at full size while others were cut in half and spread out as needed. You can obtain this from Parts Express under part #268-010 at a cost of $1.65 per sheet if you buy 4 or more sheets. I'll end up using 10 sheets in my two cabinets, but one could easily use 12 - 14 sheets if they wanted to. Another alternative to this is part # 268-030. This is basically the same thing but using a "high tech" material. It is thinner but provides twice the damping ( according to their advertisement ). Only problem is that it costs $3.10 per sheet in quantities greater than 4. This effectively doubles the cost of per sheet, but it also supposedly twice as effective. Between the braces and stuffing, i'm "assuming" that the asphault based sheets should suffice at half the price.

10) Rather than rely on the adhesive that comes on the back of each of these damping sheets by itself, i applied some glue to the insides of the cabinet where the sheets would be applied. I did not want to chance the sheets coming loose and rattling / vibrating / falling inside the cabinet once i got this thing put back together.

I ended up stopping at this point as the physical labor is over and we now get into the electrical phase of things.

I now have to order caps for the woofer section as i had already ordered all of the caps for the upper ranges. I'm using poly caps through-out except in the woofer section. I did run into one problem though as there is a cap labeled as being 6 uF's on the schematic and the actual part in the crossover is 8 uF's. As such, i had to order these separately. With that in mind, you might want to pull your specific set of speakers apart and make sure that the parts values that you need correspond with what the schematic calls for. I've been told that there were TONS of revisions made to the 9 crossover during production, so it is quite possible that they may all vary slightly. I would use whatever value is in your speaker and not worry about what the schematic calls for.

While this is a matter of personal choice, i will be bypassing all of the attenuation circuitry in the crossover. I'm doing this as i never use any of the attenuators, it simpifies the circuit and is "cheaper" Honestly though, this gets rid of excess wiring, 6 resistors, solder connections and allows the signal to go directly from crossover to driver without having to pass through the scrawny contacts within the three switches. The end result is a much cleaner and shorter signal path with less potential for signal degradation. I'll leave the switches in place for cosmetic reasons, but they will NOT be functional.

Along those same lines, i'll be doing point to point wiring and keeping the leads on all of the caps as short as possible. Internal wiring will be Kimber 4TC for the top section and 8TC for the woofers. I am not saying that this is the best wire for this specific situation, but it works well in my specific system. Binding posts are made by Axon and easily accept up to 8 gauge wire without a hassle.

I will only add that if you are going to replace / upgrade the resistors while you're tearing your 9's / 90's apart, you have to make sure that you get non-inductive resistors that will handle at least 22 watts. Since this is an oddball figure, you can probably just wire two 12 watt Mills resistors ( available at Parts Express ) in series to achieve the power handling capacity needed. Just keep track of what you are soldering together so that you end up with the correct values.

As far as changing / upgrading inductors goes, the most benefit would come from replacing those within the woofer section. Only problem is that the whole woofer "Q circuit" is based on using those EXACT coils that came in the speaker from the factory. Changing the inductors would alter the performance of the "bass network" due to alterations to series resistance. As such, i would NOT attempt this unless you were willing to open up what could end up being a bottomless can of worms.

I'm working on taking pictures of all of this as i progress, so Mark should be able to post these sometime down the road. Sometimes seeing is a lot easier to understand than trying to read things over and over again : ) Sean

>