The "before" on these is as refurbished in 2005 by Graz. They were built in 1988, and the XO board is marked "Rev 1", for what it's worth. The inductor values do not match those shown on the Apogee Speakers web site. Tucked away next to the switch are spare red and blue wires that are used when connecting an active XO, along with 3 resistors just below the switch that are attenuators for the ribbon when a DAX is used.

It's nice to see that Apogee used 10ga wiring for the XO, and it is now connected to Boccino binding posts, so much of the heavy lifting is out of the way.

In these images of the left speaker, you will also note one of Graz' tweaks - the green angle iron at the left stiffens the remarkably thin (11/16") MDF speaker frame.

I propose to upgrade and bypass old caps with same-value better-quality new caps, pot the coils, and stiffen the structure between the panel and the ribbon, where the MDF is still pretty flimsy.

The speakers will wind up on stands about 10" above the floor and vertical, attached to 100-lb. stone slabs lying on carpeted floor.

HJ

The values of the inductors in the D sigs did seem to vary. I think the later ones had a smaller value bass inductor and a bigger value mid/tweeter inductor.
I would stick with the values of the ones in there looks like a 2.88mH bass and 4.?mH mid/tweeter.
I rebuilt a pair recently which were marked REV I and had virtually the same values and they sounded great.

Jon

The smaller coils are marked 30t and 1.77uH.

As Jon says, the two big coils are smaller values than documented, and the smaller ones are larger than documented.

None of these values are very close to those shown in the circuit diagrams on the Apogee Speakers webs site, and I'm a bit puzzled by the large discrepancies. Perhaps that's the Rev 1 engineering change.

The rest of the components are as documented.

I'll be replacing about half of the caps. Estimated cost of my economy XO upgrade project is about US$300, most it for a dozen 5mfd Dynamicaps to replace 5 and 10mfd Spragues. I anticipate improvements similar to what I got from my Slant 8 XO upgrades.

Graz tells me that there is not much point in trying to internally stiffen the mdf between the drivers, so that's off the table for me.

My Duetta Sigs already have upgraded binding posts and some heavy-duty internal wiring, so that part is taken care of. My system has Boccino posts, which are nice, but outlandishly priced. For an excellent budget alternative, try Cardas. The Cardas posts are also meant to be soldered, unlike the Boccino, which use a small set-screw.

So I started my 80:20 upgrade process by potting the inductors; I used epoxy to encapsulate and glue the coils together, in order to remove mechanical motion that can cause distortion.

I'm writing this because I'm bored while the epoxy hardens, so bear with me...

The theory about this is that whenever you run a current, like music, through a coil, it creates a magnetic field. In response to the magnetic field, the coil tries to physically expand due to the stresses, and then, after the current reverses, the field collapses, and so does the coil. When the field collapses, you have a coil in a changing magnetic field, and this generates current, which can modify a signal connected to it. This current created by the back EMF becomes non-linear if the coils are free to move physically.

Any coils that are free to move will have natural mechanical resonant frequencies. When something excites these resonances (like a loudspeaker element), they become a moving wire in a magnetic field, and thus generate frequency-dependent electrical distortion.

Also, any motion of the coils mostly dissipates as heat, the electrical energy that created motion is lost, and this is frequency dependent - low bass will have the most losses.

This mechanical energy storage in an inductor if the wire coils are free to move creates both electrical losses, and frequency-dependent distortion.

By encapsulating the wire coils in a hard insulating material such as shellac, varnish, or epoxy, one can remove some distortions created by loosely-wound coils.

The inductors used in the Apogees I have seen have been quite loosely-wound, and not potted with anything but a dab or two of silicone rubber or four nylon wire-ties to keep the coil from loosening or unwinding.

My approach is the the lazy-man's method; without removing the coils from the speaker, I soaked them in two-part epoxy, and let the stuff set. When it hardens, the wire coils are no longer free to move.

If you doubt that this is significant, ask George at Northridge Music; part of the magic of his custom-wound inductors is that they are potted solid by a vacuum process to eliminate even small bubbles inside the coil. His inductors are very solid and the coils don’t move – not at all.

While my hardware store epoxy is probably not the ultimate dielectric transformer potting agent available, I expect that I probably got 80% of the benefit of four new coils, at 2% of the cost.

I started by putting both speakers on their backs with the covers off.

For the largest coil, I cut off the front cardboard spool face, exposing the coils. I removed the dab of silicone that half-heartedly held it together, and then used 2" polyethylene "window-sealing" tape to make a snug dam all around the two large coils, sealing the back side of the coils to the cardboard spool or XO backboard. The purpose of this is to contain the epoxy so it cannot run out the sides of the coil and make a mess. The polyethylene tape can be removed from the epoxy after it sets.

I also needed something to avoid filling the center of the air-core inductors with epoxy. The largest coil had a core in it from the cardboard spool, but the next-largest one had no inside core. A toilet-paper tube made just the right size cardboard core, and I used 5-minute mix-on-demand epoxy to seal the core to the backboard.

I like Devcon FlowMix epoxy for this sort of project. It comes out of the double-hypodermic through a trick nozzle that mixes the two components on the fly. You can squirt it where and when you need it, already mixed in place. It, and the penetrating epoxy I used, are solvent-free, so they shouldn't cause problems for the insulation on the coil wire.

I used the FlowMix epoxy to soak and cover the two small coils, as it is just the right viscosity to wick through a couple of layers into the coil while it is fluid, while being viscous enough that it doesn't run all over the place. The small coils contained the epoxy nicely just a few drops of it running out.

For the larger coils I first soaked all four using just a few ounces of a very thin two-part penetrating epoxy that is sold as a wood-rot hardener. It is thin enough to soak into dry-rotted wood, and it did a fine job of going all the way into the large coils, with just a small amount running out around the bottom of the tape dam. I'm pretty confident that anywhere two wire coils are in close proximity inside these inductors, they are now solidly glued together with epoxy wetting and filling the gaps between them.

To finish the large inductors, I then used three tubes of the more viscous FlowMix on the up-facing (front) surfaces, so that it did not flow all the way through the coil, but remained in the front few layers of the coils to harden there. I expect that the thin epoxy did a better job of filling the lower part (back side) of the coils, while the thicker stuff won’t run all the way through.

Here's a picture of the XO with the tape dams still in place around the large coils while the epoxy is setting. Note the plywood protecting the drivers...

About $25 spent for epoxy so far. Soon - the capacitor upgrades.

than i built the x-over in a chassis with sirex

the results are amazing

quite different speaker - i don't need a sub anymore -
the bass is so low and very quick -

very musical instrument now, my Duetta Exclusive

I don't think that my perceptions of this are entirely wishful thinking, and the aural changes that I hear are not the same as my preconceptions of what I might hear from the change, so may not be entirely my imagination...

This has brought a another level of detail to my D-Sigs. I hear a bit tighter more detailed bass and more precise highs. Cymbals sound more realistic, but the real change has been through the vocal range.

It seems to me that the naturalness of vocals has been improved quite a bit: Norah Jones' "Don't Miss You At All" sounds more natural and un-reproduced to me. I've also noticed that the breath sounds on a number of jazz sax albums sound more as if they come from a human throat, rather than the plumbing.

I've not noticed any tonal shifts or other adverse effects from this change. Definitely worth doing on any Apogee with unpotted coils, IMHO.

The theory behind this is that one can greatly improve the quality of the original capacitors used in the crossover. This will result in reduced distortion caused by the almost 20-year-old OEM caps, and audibly increase the definition and clarity of the speaker without affecting tonal balance. Capacitors do degrade over time, and capacitor technology available today is much better suited to audiophile needs as well.

The technique is simply to replace old caps where feasible, and, at the least, to bypass or replace at least one of each group of capacitors with a new, high-quality capacitor. This allows the better-quality capacitor to effectively "fill in the gaps" where the older or larger-value cap is unable to accurately respond to demands for current.

In the Duetta-series, this is a particularly effective tweak because the frequency-shaping done by Apogee requires that the signal pass thru as many as three groups of caps in series. Even if each group of caps contributes only a small amount of distortion, that will become magnified by having three sets of caps in a row.

The new caps are the red ones. I used six 5mfd Dynamicaps, and three .1mfds to bypass the capacitor groups.

I used TRT Dynamicaps, an audiophile-grade metalized polystyrene caps that comes in reasonable sizes. I was unable to directly replace the 10mfd originals with Dynamicaps, as they are too fat to fit, so I used two 5mfd caps to replace a couple of 10 mfd Spragues (at the right).

Without affecting the value significantly, the .1mfd provide a higher-frequency signal path that helps maintain phase-accuracy and reduce any distortion from the larger-value caps.

You may also note that in the case of the bank of seven 10mfd caps, and in the 5mfd/10mfd combination at the left, I did not replace all the caps. Replacing one of each group is enough to get perhaps 80% of the potential benefit of replacing all of them, and a lot less costly.

While I was in there with a soldering gun, I also soldered four crimped joints between the stranded hookup wire and the solid-wire crossover buses. Even if these connections didn't oxidize or degrade during 20 years of use, soldering them will reduce the resistance somewhat.

The results of my 80:20 upgrade project have been well worth the effort - on the order of a pre-amp or amplifier upgrade, and much more of a difference than any speaker wires that I ever heard compared.

What I'm hearing after the cap replacements is that the speaker has become "invisible" - there is now almost no sense that the sound is coming from the speakers. The detail and clarity and ability to separate different instruments and voices is greatly enhanced. Breath-sounds and instrumental techniques are more natural, and it is clear that a whole level of distortion has been removed from the sound.

yes, complete new ones. They made with Corian on machine CEC. With the same material they also make
gitars