I See the Puzzle of the N800-F Remains Unresolved @ Lansing Heritage...
That linked thread on the Lansing Heritage board is so long as to be almost un-followable. Too bad it couldn't be broken up into logical chapters for ease of comprehension. Anyways...
I see that the puzzle of the potted Altec N800-F is still not totally solved. I've had a few of those apart and reverse engineered the circuit and measured both the components and nominal response. My experience comes mostly from refurbing a pair of Altec Carmels. These use 2x12' 16ohm LF drivers wired in parallel @ 8 ohm with a 16ohm 511/806. Also, worked on a pair of Valencias. Based on this work, my observations:
The N800-F schematic posted on the LH board and reproduced below is not quite correct.
The basic circuit is correct but the level control is shown as an "L" pad. Based on my experience and deciphering the comments from the LH board thread, the control is actually a simple 25ohm, linear taper pot used in a somewhat unconventional manner to partially simulate the constant input impedance input characteristics of a true L pad. The actual circuit is as shown below. It's a relatively common Altec xover topology from the late 50s/early 60s. This image comes from
How to Build Speaker Enclosures by Badmaieff and Davis. The usual arrangement was for the control to be wired up as an attenuator so the "hot" lead going to the HF driver was connected to the CCW side of the pot. Thus max HF level occurs non-intuitively at full CCW rotation.
Notice how the the HF driver is connected to the pot. One of the drawbacks of this arrangement is that the input impedance is not constant. At max level with a 16ohm load the hi-pass portion of the xover sees 11.8ohms, 15.2ohms, at mid rotation and 13.4ohms at max atten. More about this a little farther down.
My measurements essentially confirm the component values in the first image above. I got ~3.1mH for the inductors with 0.4 ohm DCR. The lo-pass cap is consistently specd at 10.0 or 10.5uF. The hi-pass section is not consistent. I've discovered at least two versions. One uses 10.0 or 10.5uF which is sorta conventional and results in a nominal xover F ~750Hz assuming an 11.8ohm load. The other version uses a 4uF cap in the hi-pass section. This results in a xover around 1200Hz again @ 11.8 ohm. It also produces a calculated dip in response of > 3dB centered on 1200Hz. From date codes on the caps (1962 and 1963), the 4uF hi-pass iteration is the later one. Seems that Altec might have recognized the need to tame the mid range a bit and combined with the relatively higher eff of the HF driver, maybe a bit of simulated HF boost.
One other issue with the N800-F xover is that it presents a variable load on the hi-pass portion of the xover depending on attenuation, which in turn alters the xover point as well as the HF level. This effect is quite noticeable to my ears on both the Carmels and the Valencias. I simulated the electrical response of the xovers in the following two images. These assume 8ohm lo-pass load and 16 ohm hi-pass load and don’t include acoustic response. The show what happens in theory at max ccw , mid and max cw rotation of the attenuator.
Everything above applies to the earlier, potted versions of the N800-F. I have no idea what's going on with the later, unpotted versions. +The may indeed be using true L-pads for level control.
Considering how the N800-F was applied to the Carmel and Valencia and who knows what other models and impedances, I have to wonder if Altec was really paying attention the to the details back in the day. Or maybe it’s subtle genius. Regardless, I'm probably going to redo the Carmel xovers as 2nd order LR @1000Hz using actual impedances and not nominal, Will probably add some form of HF "equalization" too although not like the std Altec bridged "T" config.
Got any leads on where to obtain correct foam for LE14A, the “early” version with the reverse roll, lansaloy surround. None of four possible sources so far have been able to provide anything close. I’m leaning towards brake fluid. No lectures please