Upgrading a pair of Heathkit AS101/Altec Valencia Speakers

[ZilchLab]

Originally posted by revintage
Please enlighten me! What is CD compensation?

The rolloff of high frequencies above ~3.5 kHz is characteristic of compression drivers, so characteristic, in fact, that most pro active crossovers and some pro amps have compensation circuitry built into them which may be engaged for the high frequencies when used with compression drivers on horns that require it.

In the case of the Behringer CX3400 used for those measurements, you push the "CD Horn" button to, in effect, "boost" the highs and flatten the response.

It's clear from those curves that it would require ~20 dB of compensation to make Altec 806As on 811B horns play flat to 20 kHz, which is simply not practical. However, the high frequency response typical of early Altecs can certainly be extended using this approach.

Similar may be accomplished passively, as Altec did in Model 19 and other subsequent designs. In fact, there is some (insufficient) compensation built into the simple 2-pole N800-F and N800-8K crossovers used in Valencia and other similar Altec two-ways. They are NOT textbook crossovers. The series crossovers such as N1500-A (604E) incorporate CD compensation, as well....

[ZilchLab]

Originally posted by kevinkr


The lower curves shown match my recollections quite well.

To be fair, 806A running on the N800-F is a bit better; you can see the built-in compensation in operation in the measurements below.

My point is, unless you prefer rolled-off highs and no extended bass, there seems little benefit to accrue from swapping caps and/or amps with these systems until their more fundamental deficiencies are recognized and addressed....

[djeickme75]

I think if someone is looking for flat frequency response for 20hz to 20khz they wouldn't start with a speaker that is older than they are. If that is what you seek I don't see why you wouldn't simply stick to a more standard formula modern speaker. There are plenty of sonic standouts at all price points that you could pursue instead of putting down a product that simply has a different set of priorities than you have. Many (if not most vintage) products are bought by people who simply value different things in the reproduction of recorded music. I think rhythmic drive, mid range purity, micro and macro dynamics, cohesiveness and ultimately humanity are what most vintage speaker enthusiasts seek. My home theatre is flat from 15hz to 20knz but I can't stand listening to it with music. It has no life. It is the little imperfections that make things interesting. There is even evolutionary biology that confirms this. There are studies that show that people are more attracted to other people that have a few minor facial imperfections than those who are "perfect". I choose real and therefore often "flawed" products over perfect but unrealistic products all the time. My residence is a 115 year old church, I have solid wood doors because hollow core or MDF makes me cringe, my saxophone is a vintage 80 year old alto, I even make my own ketchup to have something that has more personality than Heinz. While I welcome and respect all opinions I will respectfully disagree with the last series of posts and say that the imperfections of the speaker are part of what makes it worth pursuing not the ultimate quest for perfection.

[djeickme75]

It should be noted that although the Heathkits are based on the Valencia there are a number of differences as well. It seems that the venting arrangement is different (I read this on another forums but have not yet had a chance to confirm).

The crossover is also different than the stock Valencia. I don't have the schematic in electronic form so bare with me as I describe it. The 806 is crossover with a 1.75mH inductor and a 8uF cap. There is also an LPad to adjust its level relative to the woofer. The woofer is crossed over using another 1.75mH inductor and a pair of caps that total 21uF connected in parallel. According to the schematic/manual that I have the design goal is a crossover that functions at 800hz. I would say that the crossover is pretty textbook and in general very simple as opposed to something with a lot of compensation in it.

[ZilchLab]

http://audioheritage.org/vbulletin/s...31&#post160631

Double the Cs and halve the Ls for your 8-Ohm version, N800-8K.

Then ask yourself why the C in the highpass is 8 uF.

[ZilchLab]

SPICE sim of voltage drives:

[kevinkr]

Clever..... Avoids the need for a step network I suspect..

I too like (horn based) vintage speaker systems that often don't measure that well relative to more modern designs. (viz:my Onkens with JBL 4333 and 2402 horn drivers.)

[ZilchLab]

You can add a step network as well, but you run out of headroom with 806As pretty quickly, since they begin with a 3 dB deficit in comparison to the larger-magnet Altec drivers.

I suspect what we most like about them derives from compression driver technology, which has itself moved well ahead since these early implementations. For home use, it almost has to be DIY; there's not much in the way of complete product available but at prohibitive cost.

In my experience, more performance can be squeezed out of 806As in Valencia and similar, but what I've shown above actively biamped with CD compensation is about the limit. To get beyond that, it's driver swap or augmentation.

With respect to compensation, I try to stay within what Altec themselves determined was acceptable for use with their drivers, as is evidenced in their own subsequent design upgrades, e.g., tangerine phase-plug drivers and Model 19 compensation would be a rational upgrade path for Valencias within the Altec lineage. Run the box models and look to Model 19 for woofer upgrade, as well....

[DonM]

Zilch,

When you measured that 806A on 811B (with and without CD compensation) and the N800F crossover do you remember the variable value of the parallel resistance (HF side) used in the test?

Thanks,

Don M

[ZilchLab]

Hi, Don,

Looking at it now, since the VHF is virtually superimposed, there was no attenuation.

Other than as attributable to impedance anomalies introduced by the N800-F attenuation potentiometer, or the L-pad used in its stead in N800-8K, the compensation established by the filter elements may be viewed as constant.

Some minor variabilities in frequency response may be seen in the curve families generated at various attenuation settings using the actual crossovers.

I never figured that out, in detail, but it did ultimately prompt me to conduct an investigation of just how fixed the reflected impedance of typical L-pads actually is. Even using fixed resistors as load, they rather suck....