Discussion:
Key to THX Quality?
(too old to reply)
Michael Rhodes
2008-06-08 13:17:24 UTC
Permalink
Reading up on THX, I found this in Wikipedia...

"THX also provides certified theaters with a special crossover
circuit whose use is part of the standard."

Now what could that be, and why haven't we been told about it?
Couldn't that be the Zobel ciircuit? I was wondering where the Zobel
might be found commercially; and there it is, in THX. You might want
to install it yourself, without buying THX theaters. I have, and it
was quite an improvement to my system.

Though simple in itself, the problem with adding the Zobel is that the
speakers will probably sound bassy afterward, and should be rebalanced
afterwards. If that is not possible then a graphics equalizer might
correct that problem. The improvement in sound quality should be
worth it.

Before reading this, know that the solution is more simpler, and less
expensive, than the explanation below. The solution is at the end of
this post.

You can find information about the Zobel in other places on the
internet, but I will try to explain it here. Audio companies will not
give you this. They prefer you buy equalizers or more speakers to
ATTEMPT to correct the problem. They don't make much money off the
little Zobel. They won't even sell it to us.


The Zobel: •-----/\/\/\-----| |------•
Rz Cz

The simple Zobel circuit corrects the impedance of coil drivers that
are connected to passive crossovers. Passive crossovers, which are
the majority of crossovers used, require constant impedance at their
output to the driver.

Active crossovers, as opposed to passive, connect the driver directly
to the amp. They do not require Zobel-corrected drivers. The
amplifier, in such cases, is in direct control of the voltage across
the driver. The voltage across the driver is what is important.

Why is voltage important? Crossovers are voltage divider circuits,
with the driver as part of that circuit. And driver volume is voltage
dependent. So, as impedance of the driver increases with frequency
(with its coil acting as an inductor), then the voltage across the
driver likewise increases. Higher voltages increase volume at higher
frequencies.

The crossover can only try to reduce the voltage, but high impedance
prevented it. We want drivers with constant impedance, but the audio
industry is refusing to sell them to us. They could. Speaker
builders (including Bose) refuse to add this circuit to their systems.
It is a conspiracy; for the sake of a bloated and problematic audio
industry. I seriously believe this circuit is THE key to THX quality.
It is what they want, but bundle it with the rest to hide it.

The driver's coil acting as an inductor will negate the filtering of
the crossover at higher frequencies. So the woofer puts out midrange
it shouldn't. And the midrange is acting as a tweeter when it
shouldn't.

Expected Midrance Response Actual Response
___ __
/ \ / ----___
/ \ /

It is impossible to balance the speakers to a flat response with a
drivers acting this way. Distortion will also increase. Except for
THX theaters (unless you know others), this is the typical design of
the sytem sold to the public!

The Zobel is simply a resistor and capacitor in series. It is
similar to a hi-pass filter that is connected in parallel with a
driver that is acting as a low-pass filter, to ensure the driver's
actual impedance remains near its rated impedance.

The THX circuit probably includes a resonance compensator along with
the Zobel. Most all systems which vibrate, such as drivers, will also
resonate to some extent. When a speaker resonates its impedance also
increases. Resonance occurs at the low end of the frequency spectrum
(usually below the driver's normal operating range), as opposed to
inductor effects, which are at the high end.

Unless one is an electronics engineer, resonance compensation is
difficult to calculate. But it might not even be required in
well-made drivers that are well-damped. My Speakerlab midranges are
claimed to be well-damped. I've check its resonance. It peaked at 12
ohms at 200 hz. This is minimal, compared to some charts I've seen on
the internet. My Klipschorn tweeters are damped using magnetic fluid,
so I don't worry about those either. I could easily blow my tweeter
just by testing it, unless I had professional equipment. I do not.

From an impedance perspective woofer resonance is ignorable, since at
those low frequencies the crossover is essentially a wire to the
woofer. There is no L-pad. So the amp is direct control of its
voltage. Woofer resonance is ignorable, but not inductance!

Back to the Zobel...

There are standard equations for determining the values of the Zobel
resistor and capacitor. The standard values are Rz = 1.25 * driver dc
resistance, and Cz = driver inductance/(Rz^2). Another source I've
found suggests that Rz should equal driver resistance. BUT ANYTHING
HELPS! Seriously.

------->

Lacking information about driver characteristics (the inductance)
should not prevent you from installing something. I would suggest...

Rz = 8 ohms (or 4 for 4 ohm speakers.)
Cz = 30 uf, for the woofer.
Cz = 10 uf for the midrange.

For 4-ohm systems, the capacitor values should probably also be
reduced for the smaller coils. My speakers are all 8-ohm, so I have
no experience with 4-ohm, but the electronics is the same. You might
reduce capacitors to half the suggested values and see how that works.

------->

I consider tweeter Zobels only as an option. I added a 3 uf Zobel to
my Klipschorn tweeters. They are very efficient horns. But the
improvement wasn't noticeable as it was to the woofer and midrange.
Some tweeters may actually depend on impedance variations to produce
the ultra highs. It may be best not to install a Zobel to the tweeter.

That being said, speakers should be balanced. The purpose of the
Zobel is for a constant impedance at the high end. The 3 uf Zobel
added to the tweeter may not provide an 8-ohm impedance, but, if
balanced, it should flatten impedance (and therefore, hopefully,
response) at higher frequencies.

Speakers are not perfect inductors. The capacitor of the Zobel will
actually over-correct the inductor nature of the driver. But a slight
over-correction is much preferred over not having it installed at all.
--
Michael
Michael Rhodes
2008-06-09 09:45:24 UTC
Permalink
Adding a Zobel to a driver will affect the phase of that driver, so
that it will change its phase with its adjacent driver(s). Driver
phase is important in the crossover portion of the spectrum. The
volumes should add together, and not cancel. Outside the crossover
regions the phase differences are ignorable.

Phase can be checked simply by reversing the leads to the midrange (or
tweeter of a 2-way), then applying a tone at the crossover
frequency(s). Leave the other speaker as is, and compare volume
between the two. The loudest volume is the correct phase. Use that
polarity of leads.

WinISD is a program run on Windows computers, and is free off the
internet at
http://www.linearteam.dk/default.aspx?pageid=winisd. It has a
built-in signal generator. That is the only function I've ever used.
Connect your computer to your audio system and you can use WinISD to
send tones to your speakers.

You check by switching the connecting wires to the midrange only. Then
switch only the woofer or tweeter after discovering which phase is
off, if any.

Technically...

My speakers have 1st-order crossovers (simplest), except for the
tweeter, which is 2nd order (an additional inductor). Reversing only
the midrange leads did seem to improve its phase with the woofer
somewhat. But it really helped its phase with the 2nd-order tweeter.
My speakers now have their midrange wires connected in reverse.

My understanding is that crossovers without the Zobel tend to show
little phase shift above the crossover frequency. But with impedance
correction, then the phase will resemble the typical phase shifts seen
in vibrations class at school.

If you're curious, (assuming my math model is correct), the woofer
phase shift of my Zobel-corrected crossover is 0 to -90º, being -45º
at the -3 db point.

The midrange (with reversed leads) is -90º to 90º; being -135º at -3
db, and 135º at the other -3 db point. (Not reversed, it would be 90
to -90, and 45 and -45 at the -3 db points.)

The tweeter phase shift is 180º to 0º, being 90º at the -3 db point.
(The tweeter would be 90º to 0º if it were a first order filter, and
45º at the -3 db point.)

Matching phases at the crossover points, the woofer/midrange is
-45/-135 (90º difference theoretical), and the midrange/tweeter is
135/90 (45º difference theoretical). The problem with theory is that
actual crossover points need not be at the -3 db point. Shifting that
location requires examination of the phase by sound testing as above,
to see if reversing leads will improve it. We also need to know what
the actual crossover points are. I will post a method do so.

Switching the mid's leads should normally do little to change
bass/mid phase (though it did with mine), but did bring mid and
tweeter to within 45º, instead of 135º.

(I think; well, fairly sure.)
--
Michael

On Sun, 08 Jun 2008 07:17:24 -0600, Michael Rhodes
Post by Michael Rhodes
Reading up on THX, I found this in Wikipedia...
"THX also provides certified theaters with a special crossover
circuit whose use is part of the standard."
Now what could that be, and why haven't we been told about it?
Couldn't that be the Zobel ciircuit? I was wondering where the Zobel
might be found commercially; and there it is, in THX. You might want
to install it yourself, without buying THX theaters. I have, and it
was quite an improvement to my system.
Though simple in itself, the problem with adding the Zobel is that the
speakers will probably sound bassy afterward, and should be rebalanced
afterwards. If that is not possible then a graphics equalizer might
correct that problem. The improvement in sound quality should be
worth it.
Before reading this, know that the solution is more simpler, and less
expensive, than the explanation below. The solution is at the end of
this post.
You can find information about the Zobel in other places on the
internet, but I will try to explain it here. Audio companies will not
give you this. They prefer you buy equalizers or more speakers to
ATTEMPT to correct the problem. They don't make much money off the
little Zobel. They won't even sell it to us.
The Zobel: •-----/\/\/\-----| |------•
Rz Cz
The simple Zobel circuit corrects the impedance of coil drivers that
are connected to passive crossovers. Passive crossovers, which are
the majority of crossovers used, require constant impedance at their
output to the driver.
Active crossovers, as opposed to passive, connect the driver directly
to the amp. They do not require Zobel-corrected drivers. The
amplifier, in such cases, is in direct control of the voltage across
the driver. The voltage across the driver is what is important.
Why is voltage important? Crossovers are voltage divider circuits,
with the driver as part of that circuit. And driver volume is voltage
dependent. So, as impedance of the driver increases with frequency
(with its coil acting as an inductor), then the voltage across the
driver likewise increases. Higher voltages increase volume at higher
frequencies.
The crossover can only try to reduce the voltage, but high impedance
prevented it. We want drivers with constant impedance, but the audio
industry is refusing to sell them to us. They could. Speaker
builders (including Bose) refuse to add this circuit to their systems.
It is a conspiracy; for the sake of a bloated and problematic audio
industry. I seriously believe this circuit is THE key to THX quality.
It is what they want, but bundle it with the rest to hide it.
The driver's coil acting as an inductor will negate the filtering of
the crossover at higher frequencies. So the woofer puts out midrange
it shouldn't. And the midrange is acting as a tweeter when it
shouldn't.
Expected Midrance Response Actual Response
___ __
/ \ / ----___
/ \ /
It is impossible to balance the speakers to a flat response with a
drivers acting this way. Distortion will also increase. Except for
THX theaters (unless you know others), this is the typical design of
the sytem sold to the public!
The Zobel is simply a resistor and capacitor in series. It is
similar to a hi-pass filter that is connected in parallel with a
driver that is acting as a low-pass filter, to ensure the driver's
actual impedance remains near its rated impedance.
The THX circuit probably includes a resonance compensator along with
the Zobel. Most all systems which vibrate, such as drivers, will also
resonate to some extent. When a speaker resonates its impedance also
increases. Resonance occurs at the low end of the frequency spectrum
(usually below the driver's normal operating range), as opposed to
inductor effects, which are at the high end.
Unless one is an electronics engineer, resonance compensation is
difficult to calculate. But it might not even be required in
well-made drivers that are well-damped. My Speakerlab midranges are
claimed to be well-damped. I've check its resonance. It peaked at 12
ohms at 200 hz. This is minimal, compared to some charts I've seen on
the internet. My Klipschorn tweeters are damped using magnetic fluid,
so I don't worry about those either. I could easily blow my tweeter
just by testing it, unless I had professional equipment. I do not.
From an impedance perspective woofer resonance is ignorable, since at
those low frequencies the crossover is essentially a wire to the
woofer. There is no L-pad. So the amp is direct control of its
voltage. Woofer resonance is ignorable, but not inductance!
Back to the Zobel...
There are standard equations for determining the values of the Zobel
resistor and capacitor. The standard values are Rz = 1.25 * driver dc
resistance, and Cz = driver inductance/(Rz^2). Another source I've
found suggests that Rz should equal driver resistance. BUT ANYTHING
HELPS! Seriously.
------->
Lacking information about driver characteristics (the inductance)
should not prevent you from installing something. I would suggest...
Rz = 8 ohms (or 4 for 4 ohm speakers.)
Cz = 30 uf, for the woofer.
Cz = 10 uf for the midrange.
For 4-ohm systems, the capacitor values should probably also be
reduced for the smaller coils. My speakers are all 8-ohm, so I have
no experience with 4-ohm, but the electronics is the same. You might
reduce capacitors to half the suggested values and see how that works.
------->
I consider tweeter Zobels only as an option. I added a 3 uf Zobel to
my Klipschorn tweeters. They are very efficient horns. But the
improvement wasn't noticeable as it was to the woofer and midrange.
Some tweeters may actually depend on impedance variations to produce
the ultra highs. It may be best not to install a Zobel to the tweeter.
That being said, speakers should be balanced. The purpose of the
Zobel is for a constant impedance at the high end. The 3 uf Zobel
added to the tweeter may not provide an 8-ohm impedance, but, if
balanced, it should flatten impedance (and therefore, hopefully,
response) at higher frequencies.
Speakers are not perfect inductors. The capacitor of the Zobel will
actually over-correct the inductor nature of the driver. But a slight
over-correction is much preferred over not having it installed at all.
Michael Rhodes
2008-07-07 05:35:52 UTC
Permalink
On Sun, 08 Jun 2008 07:17:24 -0600, Michael Rhodes
Post by Michael Rhodes
The simple Zobel circuit corrects the impedance of coil drivers that
are connected to passive crossovers. Passive crossovers, which are
the majority of crossovers used, require constant impedance at their
output to the driver.
Uh-oh, more technical stuff. Oh if there were only a news group
specific to speakers! But Bose won't let us have it, so I'll drop
this in here...

There is a method of determining the impedance of a driver at audio
frequencies; along with the Zobel circuit, if it is connected. The
results would be a series of data points plotted graphically, across
its useful frequency spectrum. (Or tables of data, as in the examples
I've provided below.)

For a Zobel-corrected driver, the results should approximate a flat
line somewhere near the driver's rated impedance. But without the
Zobel, then you can see just just high a driver will go above its 8
ohm rating, and it is many times over!

After checking the results, then you may fine-tune the Zobel for its
driver; whether by changing the resistor value, or the capacitor.
Changing the resistor shifts the entire plot up or down, while
changing the capacitor will only affect the high-frequency portion of
the plot. My drivers/Zobels test are mostly within a -1.5 ohms of
rated 8-ohm impedance. In the frequency ranges of filtering, lower
than 8 ohms is better than higher, as it increases filtering.
Post by Michael Rhodes
CAUTION must be taken. It is hazardous to test a driver much
below its operating frequency range. Low frequencies can produce
excessive cone movement and destroy the driver. The tweeter
especially could be easily destroyed by the voltages and frequencies
suggested in this post. Probably the same for domed midranges. Care
should be taken with any midrange which is not designed for full-range
use.

Zobel tests are conducted in or above the useful frequency range, and
as so then the tests should be safe. Resonance tests, however (if you
want to know), are in the range below its usual crossover frequency.
Those can be dangerous.

I would not touch my tweeter with any power source unless it passes
through its crossover. The crossover is usually sufficient to protect
most drivers from too-low frequencies, (but also prevents testing it
in the manner described here). That's why my tweeter filter is
2nd-order, for a sharper drop-off against those lows.

A simple voltage divider circuit is used for the tests, along with a
sine-wave source. Again, use WinISD.exe in your computer connected to
your amp...

Rb
•-----/\/\/\-------•------•
|
|
Vin Ra Vout
|
|
•------------------•------•

Where ...
Rb
Ra = -------------------- = Impedance of device
(Vin/Vout - 1) at a particular frequency

Rb = resistance of known value in range of that tested, ~10 ohms.
Vin = sine wave source of say 2.5 to 3 volts or so.

It's best to test the driver while in its enclosure. Also, for
accuracy, when measuring the resistance of Rb be sure to subtract the
resistance of the leads. My VOM measures 0.3 ohms across the leads.

The tests will be noisy, but a relatively high voltage is required for
accuracy. AC ranges in most VOMs are limited to powerline voltages,
but if you can get measurements to 0.1 volt then the suggested voltage
should be accurated enough.

My Woofer Test...
Rb = 7.9 ohms

Freq Vin Vout Ra (No Zobel)
-----------------------------------------------------------------
250 6.9 3.6 8.6
1000 7.35 5.4 22
2000 7.2 6.0 40
5000 7.0 6.25 66
10000 6.0 5.55 97
-----------------------------------------------------------------

Freq Vin Vout Ra (Z=8.3 ohm, 30uf)
-----------------------------------------------------------------
250 6.9 3.3 7.2
1000 7.2 3.5 7.5
2000 7.05 3.3 7.0
5000 6.75 3.1 6.7
10000 5.9 2.6 6.2
-----------------------------------------------------------------

Freq Vin Vout Rs (Z=8.3 ohm, 20 uf)
-----------------------------------------------------------------
250 6.9 3.45 7.9
1000 7.3 3.9 9.1
2000 7.1 3.5 7.7
5000 6.8 3.2 7.0
10000 5.9 2.6 6.2
-----------------------------------------------------------------

My Midrange Test...
Rb = 7.9 ohms

Freq Vin Vout Ra (no Zobel)
-----------------------------------------------------------------
250 6.9 3.25 7.0
1000 7.2 4.0 9.9
2000 7.1 4.5 13.7
5000 6.9 5.1 22.4
10000 6.0 4.8 31.6
-----------------------------------------------------------------

Freq Vin Vout Ra (Z=8.3 ohm, 7.2 uf)
-----------------------------------------------------------------
250 6.9 3.2 6.8
500 7.2 3.4 7.1
1000 7.25 3.75 8.5
2000 7.1 3.7 8.6
5000 6.8 3.15 6.8
10000 5.85 2.55 6.1
----------------------------------------------------------------

Freq Vin Vout Rs (8.3 ohms, 5 uf)
-----------------------------------------------------------------
250 6.9 3.2 6.8
500 7.2 3.4 7.1
1000 7.25 3.9 9.2
2000 7.1 4.0 10.2
5000 6.8 3.5 8.4
10000 5.9 2.7 6.7
----------------------------------------------------------------

My suggestions for capacitor values seem a bit high for my speakers.
But they're easy to remember, therefore suggest; and the results are
not so far off from each other. But maybe I should've suggested 20 uf
and 5 uf, instead of 30 and 10.


Resonance tests are made the same way, but at low frequencies. The
tests are made in search of the frequency where a peak impedance is
found. I cannot help you with the circuit required to correct the
resonance, though you can check to see if your midrange resonance is
substantial.

My Midrange Resonance Test...

Freq Vin Vout Ra (Rb=10 in this case)
-----------------------------------
200 2.3 0.9 6.4
190 2.3 0.9 6.4
180 2.3 1.0 7.7
170 2.3 1.05 8.4
160 2.3 1.1 9.2
150 2.35 1.2 10.4
140 2.4 1.25 10.9
130 2.4 1.3 11.8
120 2.4 1.3 11.8
110 2.4 1.3 11.8
100 2.4 1.2 10
90 2.45? 1.15 8.8
80 2.45 1.1 8.1
------------------------------------

The resonance peak of my Speakerlab M608R midrange was ~12 ohms at 120
hz. They claimed it as a damped midrange, and it apparently is. On
the internet, I've seen graphs of resonance peaks over 40 ohms, so
won't worry about this one -- though I've wondered if the test is
correct.
--
Michael
Michael Rhodes
2008-07-07 18:05:01 UTC
Permalink
On Sun, 06 Jul 2008 23:35:52 -0600, Michael Rhodes
Post by Michael Rhodes
Also, for
accuracy, when measuring the resistance of Rb be sure to subtract the
resistance of the leads. My VOM measures 0.3 ohms across the leads.
My VOM is self-zeroing for resistance (to 0.3 ohms anyway), and cannot
be adjusted. Maybe yours can, and subtracting the resistance of the
leads may sound silly to you.
(refering to Zobel cap size, close to the end of the post, before the
resonance test...)
Post by Michael Rhodes
My suggestions for capacitor values seem a bit high for my speakers.
But they're easy to remember, therefore suggest; and the results are
not so far off from each other. But maybe I should've suggested 20 uf
and 5 uf, instead of 30 and 10.
My woofers are 8-inch and low power rated (intended to be temporary).
Your woofer may be larger, 10 or 12-inch. I assumed larger woofers
might have larger inductance, therefore requiring larger Zobel
capacitors. But that was an assumption.
--
Michael
Michael Rhodes
2008-07-08 19:12:22 UTC
Permalink
On Sun, 06 Jul 2008 23:35:52 -0600, Michael Rhodes
Post by Michael Rhodes
I cannot help you with the circuit required to correct the
resonance, though you can check to see if your midrange resonance is
substantial.
One more post to this thread, having nothing better to do than pester
the audio industry. It is needed. I would be quite happy for anyone
to post corrections.

The anti-resonant circuit is a series RLC circuit (a notch filter), in
parallel with the driver. It is much more difficult to size properly
than the Zobel, while being less useful. The frequency range of a
resonant peak is much smaller than that corrected by the Zobel. A
driver with a coil will always need compensation for inductance, but a
driver can be constructed so its resonance is ignorable.

One procedure of determining the values of the components of the
resonance compensation circuit, near as I can figure from an
electronic’s text, though have not tested, is...

1) Find dc coil resistance.
2) Test for resistances in the range of the resonant frequency,
constructing a table of resistances vs frequency. Go low enough in
frequency to clearly determine the peak resistance, and past it enough
such that the resistance at the lowest tested frequency is at least
halfway between the lowest and the peak.
3) Subtract dc coil resistance from each test value above, to modify
that table.
4) Find the resonant frequency (searched for when making the table).
5) Find the ‘half-power’ (0.707) frequency points. There will be two,
one on either side of resonant frequency.
6) Find the bandwidth of the peak, in hertz = Freq1(.707) -
Freq2(.707).
7) Find Q = Resonant Freq / bandwidth

For a series RLC circuit, Q = Wo*L/R = 1/(Wo*C*R)
where Q = Resonant Freq/bandwidth, ‘Quality factor’
(the sharper the peak the higher the Q)
Wo = resonant frequency in radians/sec (2*pi*f)
L = desired coil inductance, in henries
C = desired capacitance, in farads.
R = the same as the dc resistance of the driver.
so...

From the earlier test of midrange.

Freq Volts Vspk Resistance
------------------------------------
200 2.3 0.9 6.4
190 2.3 0.9 6.4
180 2.3 1.0 7.7
170 2.3 1.05 8.4
160 2.3 1.1 9.2
150 2.35 1.2 10.4
140 2.4 1.25 10.9
130 2.4 1.3 11.8
120 2.4 1.3 11.8
110 2.4 1.3 11.8
100 2.4 1.2 10
90 2.45 1.15 8.8
80 2.45 1.1 8.1
------------------------------------

Rdc for my midrange was 6.7 ohms, as tested by my ohm-meter, which was
higher than the 6.4 ohms tested above; though not by a lot. The
variance, I think, is due to phase variations across the resistor. But
I will use the 6.4 ohms as the base dc resistance. If the peak were
considerable then the error would be small.

So modifying the table above...

Freq Volts Vspk Resistance
------------------------------------
200 2.3 0.9 0
190 2.3 0.9? 0
180 2.3 1.0 1.3
170 2.3 1.05 2.0
160 2.3 1.1 2.8
150 2.35 1.2 4.0
140 2.4 1.25 4.5
130 2.4 1.3 5.4
120 2.4 1.3 5.4
110 2.4 1.3 5.4
100 2.4 1.2 3.6
90 2.45? 1.15 2.4
80 2.45 1.1 1.7
------------------------------------

The resistance @ .707 half-power point = 5.4*.707 = 3.8 ohms. By
interpolation, the frequencies where that resistance is at are 152 hz
and 101 hz, giving a bandwith of 51 hz. The midpoint frequency
between those is 127 hz. It’s sort of close to 120 hz. The variation
is that the 0.707 freqs is probably based on a log (octave) scale.
(I'm trying to learn.)

Q = 120/(152-101) = 2.4

Wo = frequency in rads/sec = 120 * 2 * 3.14 = 754 rads/sec

So....
Rr = given as Rdc = 6.4 ohms
(found Rr info on the net - but not the rest of the procedure.)
Cr = 1/(Wo*Q*Rr) = 1/(754*2.35*6.4) = 88 uf
Lr = Q*R/Wo = 2.35*6.4/754 = 20 mh.

The 88 uf, and especially the 20 mh values, are very large. But they
appear proportioned for the frequency range of the peak. Good luck on
finding a 20 mh coil, if you actually need it. Then on testing it to
see it the thing actually works. A damped midrange may be a better
investment. If manufacturers aren't making them then they should be
punished severely.

---------

I bit the bullet and tested my irreplaceable, 30-year old Klipschorn
tweeter for resonance! Using an input voltage of ONLY 1.0 Vrms and a
10 ohm resistor divider. Yes, it was loud. The value of the results
is marginal for calculations, but I could detect no impedance peaks,
down to 1.3 khz. That was expected, as it is damped with ferrofluid.
And that tweeter still functions.

Used a 9.8 ohm resistor divider. Vout is the voltage across the
tweeter. The VOM is digital, but will flash alternating digits
occasionally. I guess at the time between digits, then assume a
voltage of, say, 0.28, if it stays on 0.3 much longer than 0.2.

Freq Vin Vout
-------------------------
5000 0.9 0.28
4000 0.95 0.3
3000 1.0 0.25
2500 1.0 0.25
2000 1.0 0.28
1500 1.07 0.28
1300 1.1 0.3

I hope my 'diary' has been of some interest to you. No other
ponderings are planned for posting.
--
Michael
Michael Rhodes
2008-07-08 23:02:37 UTC
Permalink
On Tue, 08 Jul 2008 13:12:22 -0600, Michael Rhodes
Good luck on finding a 20 mh coil, if you actually need it.
This coil is not a power-carrying component. And its resistance will
not interfere with its use, so a finer wire will work fine. Fine wire
coils are much smaller than those big things used with woofers. But
include coil resistance as part of the total resistance, reducing the
value of the resistor to accomodate.

Don't place it next to other coils on the mounting board, and
preferably at right angles to them. All coils should be mounted to
isolate them from others.
--
Michael

Loading...