PDA

View Full Version : Requesting help - Zobel for 2.5 way? (Using 20mH superQ) Measuring Inductance



TW14MB
01-29-2019, 02:06 PM
Inspired by this thread: http://www.audioheritage.org/vbulletin/showthread.php?18913-Altec-9844-8B/

I want to try the 2.5 way setup on my Altec Malibu's [806a(16ohm) CD, 811 Horn, 2x 414Z(16ohm each in parallel=8ohm) 12" woofers per enclosure, N-800F crossover].

1. With a 2.5 way system where there is a main woofer and "helper woofer" in parallel, with a 20mH inductor in series with the second woofer, for calculating Zobel values, assuming if each woofer was 2mH each, would I say my total inductance was 1mh, ignoring the 20mh inductor?

2. Since there is not a published inductance for my woofer (that I could find for Altec 414z) is it ok to use the "standard" measurement at 1khz (like with a B&K 875B), considering my crossover point is 800hz and operating range much lower?

3. When measuring woofer inductance at 1khz or any other frequency, is it necessary to keep the cone stationary? I have read that the test should be done with the cone glued or in some kind of jig to keep it from moving.

Any help is appreciated!

Thank you.

Earl K
01-29-2019, 04:07 PM
2.5's don't necessarily need any more circuitry than the 20mH coil.

Adding impedance correction simply means one will get a true 6db slope on the .5 woofer.

The problem is; does one want less or more than a .5 ? One can only tell by ( Simming or ) building it and then trying out different circuit arrangements.

Don't bother trying to calculate a Zobel, just create one within XSim ( watching the impedance trace ) while finagling the RC values to get the flattest impedance curve.

:)

TW14MB
01-29-2019, 05:52 PM
Thanks Earl! Maybe you can tell that I'm trying to build on existing knowledge without going quite that far :).

In the 9844-8B thread the Zobel was discussed but not implemented. Ignoring the 2.5 way for now, Zilch, quoting you, for parallel 16ohm woofers suggested 18 uF and 7.5 Ohms for starters. I thought this was to help the 414z/N-800F combo behave correctly for the hf rolloff.

Now, considering the 2.5 setup, I am thinking those values will change because of the 20mH coil added. Yes? No? If that is no longer a simple math question, maybe I should try this?

1. Calculate (or sim) Zobel for a single 414z and connect this as the "main" woofer, through Zobel, to the N-800F outputs
2. Connect the "helper" woofer with 20mH in-series coil, directly to the same terminals as the N-800F input

Thoughts?

Thanks!

Earl K
01-29-2019, 09:50 PM
Thanks Earl! Maybe you can tell that I'm trying to build on existing knowledge without going quite that far :).

In the 9844-8B thread the Zobel was discussed but not implemented. Ignoring the 2.5 way for now, Zilch, quoting you, for parallel 16ohm woofers suggested 18 uF and 7.5 Ohms for starters. I thought this was to help the 414z/N-800F combo behave correctly for the hf rolloff.

Now, considering the 2.5 setup, I am thinking those values will change because of the 20mH coil added. Yes? No? If that is no longer a simple math question, maybe I should try this?

1. Calculate (or sim) Zobel for a single 414z and connect this as the "main" woofer, through Zobel, to the N-800F outputs
2. Connect the "helper" woofer with 20mH in-series coil, directly to the same terminals as the N-800F input

Thoughts?

Thanks!

A Zobel of 15R + 18uF will work just fine for one individual ".5" 414z .

SkywayRider didn't use a Zobel so why do you think you need one.

What do you hope to achieve by adding one?

The presence of a Zobel will not produce more bass .

Skywave never retuned his boxes ( as far as I remember ) which might have extended the response below 50hz.

:)

TW14MB
01-30-2019, 12:06 PM
Hi Earl,

Maybe the long version will help for background.

I have tube amps. MC30's and a few other EL34 and El84 based amps, with the smallest being a strange 6AV6 6AR5 5WPC amp. I thought it would be "fun" to try some high efficiency speakers.

I bought Altec Malibu's. Unhappy with the hf/mid sound I did some research. Inspired by the 9844-B thread and other information I set up a testing configuration with 4 channels solid state and an active crossover (Behringer CX3400). Using the "CD Horn" compensation button, I was able to acheive a sound that I like. Clean highs/mids. This gave me some confidence that with modifications I could improve the sound to my taste.

The goal is to get back to passive crossover. I have the specs to build a 30904 for 16ohms HF, and also wanted to use the LF Zobel as I wanted to see if that would help the mid response and make the load a little more predictable for the tubes. I was never intending to make a flat, or even a good, speaker out of the Malibu. Simply to add a few dollars worth of components in the form of a 30904 and see how it sounds.

After seeing the graphs on the 2.5 system I want to try that as well. I haven't bought the coils yet. I am assuming that adding the 20mH coil ruins my plans (or at least changes the values) for using the simple 30904 LF Zobel.

When you say "A Zobel of 15R + 18uF will work just fine for one separate ".5" 414z." do you mean:

1. Calculate (or sim) Zobel for a single 414z and connect this as the "main" woofer, through Zobel, to the N-800F outputs.
2. Connect the "helper" woofer with 20mH in-series coil, to the same terminals as the N-800F input, with its own 15R + 18uf Zobel.

I would rather have one Zobel for both woofers in parallel, but again I am stuck because I don't understand how the additional 20mH coil affects the Zobel values.

Thanks!

Earl K
01-30-2019, 02:20 PM
A classic Zobel put across your 2 woofers will adjust the mid-range response of the lowpass ( from the N800F ).

Woofers:

If that's a goal then put the standard ( for that era ) Altec Zobel consisting of a 18uF in series with a 7.5R resistor ( wired across the load wires leading to the paralleled woofers ).

Add a 20mH coil to one of your 414z woofers immediately before the woofer ( ie; after the low-pass portion of the N800F and after the Zobel ).

Adding a single Zobel working on both woofers will adjust the mids downwards ( something like this > taken from MrT's experience );

FROM:

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=30489&stc=1&d=1200428729

TO:

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=30490&stc=1&d=1200428866

That would help breakup a buildup of mids ( in the crossover region > if your speakers are suffering from that malaise ):
- as seen here in Skywave Riders measurement ( of his 9844 ) using the M19 crossover .
83319

You would wire up the lowpass section like the following pic;

83320

Horn Circuit:

Build a 16 ohm version of the 30923 filter ( ie; the mid-range padding circuit ) and install it into the horn circuit ( after the HP's cap & coil ).

I'll be surprised if after all the above you aren't still wanting more UHF >> & if it pans out that way, buy a new pair of GPA 902-8a's and swap them in for those anemic 806's .

:)

PS; TUBE AMP LOADS:

As far as dealing with creating an impedance equalizer ( to wire across the output terminals of your tube amp ) that's a completely different animal.

It's a series LCR circuit ( wired across the amp outputs ) that's "tuned" to the worst offending impedance peak created by the passive network ( usually in the crossover area ).

IOW, once everything else is happily accomplished from executing the above mods, then you can capture an impedance trace from one Malibu ( so that a special LCR circuit can be created to level out the load impedance ).

The effect of an LCR impedance EQ on the whole load ( that the amplifier sees ) will be something like the following ( designed to take out the large impedance peak around the crossover area of the N800F ).

Impedance is changed from Grey to Green;

83322

TW14MB
02-01-2019, 11:33 AM
Thanks Earl, this is definitive information for the ages. I will start a new Malibu thread to report back with results.