I have dealt with blowing a tweeter and it ends up producing no output
usually because the voice coil gets fried by the too high output. I
currently have a stereo pair of speakers whose tweeter on one side is down
approximately 10 dB from the other tweeter, starting at rolloff and
continuing up through 20 kHz. I have switched the tweeters in question from
one cabinet to the other with the same results, therefore it's definitely
the tweeter and not the crossover or amp etc. My question then is if it is
possible to PARTIALLY blow a tweeter and if yes how could I confirm this?
The voice coil/diaphragm assembly is sold and is field replaceable but I
don't want to spend that money only to find it's still down 10 dB i.e. there
is something wrong with the magnet having shifted etc. Can I somehow test
the voice coil or physically inspect it for damage to confirm it's just that
prior to ordering a replacement or should I go ahead a get a whole new
tweeter. Morel MDT-29. Thanks and Happy Holidays! PS They are Ferro Fluid
cooled, can that spill out / does it need to be replaced?

On Tue, 21 Dec 2004 19:14:32 -0500, "Troll" wrote:

I have dealt with blowing a tweeter and it ends up producing no output
usually because the voice coil gets fried by the too high output. I
currently have a stereo pair of speakers whose tweeter on one side is down
approximately 10 dB from the other tweeter, starting at rolloff and
continuing up through 20 kHz. I have switched the tweeters in question from
one cabinet to the other with the same results, therefore it's definitely
the tweeter and not the crossover or amp etc. My question then is if it is
possible to PARTIALLY blow a tweeter and if yes how could I confirm this?
The voice coil/diaphragm assembly is sold and is field replaceable but I
don't want to spend that money only to find it's still down 10 dB i.e. there
is something wrong with the magnet having shifted etc. Can I somehow test
the voice coil or physically inspect it for damage to confirm it's just that
prior to ordering a replacement or should I go ahead a get a whole new
tweeter. Morel MDT-29. Thanks and Happy Holidays! PS They are Ferro Fluid
cooled, can that spill out / does it need to be replaced?


It is possible to have a short between the turns in the voice coil. That
would diminish the output.



Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs http://www3.sympatico.ca/borism/

"Boris Mohar" wrote because "Troll" wrote

I have dealt with blowing a tweeter and it ends up producing no output
usually because the voice coil gets fried by the too high output. I
currently have a stereo pair of speakers whose tweeter on one side is
down
approximately 10 dB from the other tweeter, starting at rolloff and
continuing up through 20 kHz. I have switched the tweeters in question
from
one cabinet to the other with the same results, therefore it's definitely
the tweeter and not the crossover or amp etc. My question then is if it
is
possible to PARTIALLY blow a tweeter and if yes how could I confirm this?
The voice coil/diaphragm assembly is sold and is field replaceable but I
don't want to spend that money only to find it's still down 10 dB i.e.
there
is something wrong with the magnet having shifted etc. Can I somehow
test
the voice coil or physically inspect it for damage to confirm it's just
that
prior to ordering a replacement or should I go ahead a get a whole new
tweeter. Morel MDT-29. Thanks and Happy Holidays! PS They are Ferro
Fluid
cooled, can that spill out / does it need to be replaced?


It is possible to have a short between the turns in the voice coil.
That
would diminish the output.

Hi Chaps,
As a shorted turn (or equally, lots of 'em) will have a major effect on
the AC impedance of the tweeter, a frequency sweep (e.g. on a test-tone CD)
with a non-inductive resistor in series with the speaker (I've used 27 Ohm 5
Watt carbons saved from old TV's!) will let you chart the overall speaker
impedance, by measuring the AC current through and and voltage across the
speaker. A shorted turn is pretty low-impedance at all frequencies, so it'll
show up as a rise in current/drop in voltage at the crossover freqency, if
you can be personally arsed to hook all the bits up ) A decent multimeter
is required, though.
A shifted magnet in a tweeter would surprise me, have you been using the
cabinet as a hammer, or maybe a clapperboard? Tweeter magnet assemblies are
pretty robust - I've had a voice-coil jarred out of position on a woofer
when drunken flat-mate knocked a really nice old Leak Sandwich off a shelf
onto the floor, face-down, which was unfortunate and terminal [1], but
that's the sort of impulse necessary with 50 square Inches of cone, not one!
I'm not aware of the composition of ferrofluid... is it something which
starts polymerising if it overheats? [2] If so, it could have thickened
enough to present more friction to the voice-coil in the magnetic gap?
If you're the gentle type, you could *extremely* carefully apply
pressure close to the voice-coil and judge whether the tweeter is binding in
the gap - ideally, apply pressure evenly around the coil, as the drive
signal would - an overheated voicecoil usually swells as it gently roasts,
and turns will rise and rub on the magnet assembly. Very occasionally, a
high turn will rub itself smooth on the magnet (after quite a few hours'
use), and the tweeter will seem ok for a while , but this rubbing thins the
winding, leaving a weak spot that pretends to be a fuse next time the volume
goes a bit too high.

HTH,
Dave H.
(The engineer formerly known as Homeless)


[1] for the speaker not the flatmate, regrettably.
[2] I'd guess at a mineral oil with magnetic particles in suspension, and
probably it would gum up if cooked.

Dave H. wrote:


A shorted turn is pretty low-impedance at all frequencies,

Including DC so why not just measure its resistance with a multimeter?

--
Eiron.

"Eiron" wrote 'cos
I wrote:


A shorted turn is pretty low-impedance at all frequencies,

Including DC so why not just measure its resistance with a multimeter?

Hi Eiron,
A single shorted turn on a coil of maybe 100 turns of fine wire in a
tweeter will make almost no difference to the DC resistance (perhaps
dropping it from 6 to 5.94 Ohms), a *very* high-precision test will be
necessary - a Wheatstone bridge with a high-quality galvanometer,
f'rinstance. The difference between the DC resistances of the two tweeters
won't really be noticeable, and will be within normal production tolerances.
A single shorted turn is, however, a short-circuit secondary transformer
coil to the changing magnetic flux in the voice-coil, and causes a much
larger drop in the voice-coil's AC impedance, and surprisingly large
currents low in the shorted turn. This difference is measurable with
hobbyist test equipment, whereas the difference at DC isn't.

HTH,
Dave H.
(The engineer formerly known as Homeless)

On Wed, 22 Dec 2004 16:04:15 GMT, "Dave H."
wrote:


"Eiron" wrote 'cos
I wrote:


A shorted turn is pretty low-impedance at all frequencies,

Including DC so why not just measure its resistance with a multimeter?

Hi Eiron,
A single shorted turn on a coil of maybe 100 turns of fine wire in a
tweeter will make almost no difference to the DC resistance (perhaps
dropping it from 6 to 5.94 Ohms), a *very* high-precision test will be
necessary - a Wheatstone bridge with a high-quality galvanometer,
f'rinstance. The difference between the DC resistances of the two tweeters
won't really be noticeable, and will be within normal production tolerances.
A single shorted turn is, however, a short-circuit secondary transformer
coil to the changing magnetic flux in the voice-coil, and causes a much
larger drop in the voice-coil's AC impedance, and surprisingly large
currents low in the shorted turn. This difference is measurable with
hobbyist test equipment, whereas the difference at DC isn't.

HTH,
Dave H.
(The engineer formerly known as Homeless)


Beat me to it Dave. Excellent explanation. I have nothing to add.



Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs http://www3.sympatico.ca/borism/

Dave H. wrote:
"Eiron" wrote 'cos
I wrote:


A shorted turn is pretty low-impedance at all frequencies,

Including DC so why not just measure its resistance with a
multimeter?

Hi Eiron,
A single shorted turn on a coil of maybe 100 turns of fine wire
in a
tweeter will make almost no difference to the DC resistance (perhaps
dropping it from 6 to 5.94 Ohms), a *very* high-precision test will
be
necessary - a Wheatstone bridge with a high-quality galvanometer,
f'rinstance. The difference between the DC resistances of the two
tweeters
won't really be noticeable, and will be within normal production
tolerances.

Wrong. Your assumption is that any such shorted turns MUST be
between adjacent turns on the same layer. The vast majority of
tweeter voice coils are, in fact, multi-layer, mostly two-layer.
A short is as likely, maybe even more so, to occur between layers
than between turns, because of the high pressure points where windings
cross over one another.

Given that scenario, it's possible to have a short which causes the
DC resistance to be ANYWHERE from slightly less than the proper
down to a near dead short.

A single shorted turn is, however, a short-circuit secondary
transformer
coil to the changing magnetic flux in the voice-coil, and causes a
much
larger drop in the voice-coil's AC impedance, and surprisingly large
currents low in the shorted turn.

No, it does not. The coupling between the rest of the coil and the
shorted turn, if it's a single turn, is tiny: if the voice coil has
100 turns, as you say, and one of them is shorted, then, assuming
the magnet assembly provides PERFECT flux coupling (ot doesn't
come anywhere near close), then the coupling to the shorted turn is
all of 1/99.

Now consider that the voice coil is, in NORMAL operations, surrounded
by one whopper of a shorted secondary turn, that being the front
metal plate of the magnet assembly, and consider this much lower-
resistance shorted secondary does NOT "cause a much larger drop
in AC impedance" as a result, it should become apparent why your
"shorted turn" theory doesn't work.

More likely causes of the problem include:

1. The voice coil was heated sufficiently that the adhesive failed
on the voice coil windings, and a number of voice coil windings
have now dropped off and are no longer in the gap. With fewer turns
in the gap, the Bl product goes down, and so does efficiency.

2. A voice coil winding has come loose and become lodged in the gap.

3. Heat has sufficiently deformed the suspension that the diaphragm
or voice coil is now mechanically constrained.

4. The connection points on the flex lead leading into the voice
coil are compromised and exhibit a high resistance.

5. While it takes a LOT of heat to do this, getting the magnet
assembly sufficiently hot can cause partial loss of magnetic
flux.

And these are just some examples of problems I have actually SEEN.
I could speculate on many more plausible causes.

But I do not recall ever finding a failure due to shorted windings.

(The engineer formerly known as Homeless)
Was Homeless a requisite step to becoming an engineer?

Save that it's wrong.

Boris Mohar wrote:
On Wed, 22 Dec 2004 16:04:15 GMT, "Dave H."
wrote:
A single shorted turn on a coil of maybe 100 turns
...

Beat me to it Dave. Excellent explanation. I have nothing to add.
Save that it's wrong.

In article ,
Dave H. wrote:

A shifted magnet in a tweeter would surprise me, have you been using the
cabinet as a hammer, or maybe a clapperboard? Tweeter magnet assemblies are
pretty robust - I've had a voice-coil jarred out of position on a woofer
when drunken flat-mate knocked a really nice old Leak Sandwich off a shelf
onto the floor, face-down, which was unfortunate and terminal [1], but
that's the sort of impulse necessary with 50 square Inches of cone, not one!

The only shifted magnet I've ever seen in a tweeter was in a Dynaudio
D-28 1" dome tweeter, used in one of my Fried B/2 satellite speaker
systems.

The speaker fell off of its stand, and landed upside-down on its top,
during the Loma Prieta earthquake in 1989. The force of the impact
broke the magnet looks from the driver frame. 'Twas a pretty big
magnet.

I glued it back, but wasn't able to get the alignment correct, and the
driver's output was many dB down from where it should have been. I
ended up replacing the tweeters in both speakers with the current
D-28/2 model, on Bud Fried's advice, and the speakers were restored to
proper operation.

I'm not aware of the composition of ferrofluid... is it something which
starts polymerising if it overheats? [2] If so, it could have thickened
enough to present more friction to the voice-coil in the magnetic gap?

It's a suspension of magnetic particles in a liquid carrier (usually a
mineral oil, I believe). I doubt that it's likely to thicken.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

On 22 Dec 2004 09:31:21 -0800, wrote:


No, it does not. The coupling between the rest of the coil and the
shorted turn, if it's a single turn, is tiny: if the voice coil has
100 turns, as you say, and one of them is shorted, then, assuming
the magnet assembly provides PERFECT flux coupling (ot doesn't
come anywhere near close), then the coupling to the shorted turn is
all of 1/99.

That is turns ratio of primary to secondary, the secondary being the
shorted turn. If the flux coupling was perfect it would be 100%. Like you
say, it is not.
Now consider that the voice coil is, in NORMAL operations, surrounded
by one whopper of a shorted secondary turn, that being the front
metal plate of the magnet assembly, and consider this much lower-
resistance shorted secondary does NOT "cause a much larger drop
in AC impedance" as a result, it should become apparent why your
"shorted turn" theory doesn't work.
That metal plate is way to far from the coil to couple to the voice coil.
Have you seen voice coils wound on an Aluminum mandrel. The mandrel does not
go the full circumference of the voice coil. There is a tiny gap. That is
to prevent the shorted turn.

More likely causes of the problem include:

1. The voice coil was heated sufficiently that the adhesive failed
on the voice coil windings, and a number of voice coil windings
have now dropped off and are no longer in the gap. With fewer turns
in the gap, the Bl product goes down, and so does efficiency.
Possible
2. A voice coil winding has come loose and become lodged in the gap.
Possible
3. Heat has sufficiently deformed the suspension that the diaphragm
or voice coil is now mechanically constrained.
Possible
4. The connection points on the flex lead leading into the voice
coil are compromised and exhibit a high resistance.
That one you should be able to hear.
5. While it takes a LOT of heat to do this, getting the magnet
assembly sufficiently hot can cause partial loss of magnetic
flux.
It would have to be on fire to go past the curie point
And these are just some examples of problems I have actually SEEN.
I could speculate on many more plausible causes.

But I do not recall ever finding a failure due to shorted windings.
Did you look?

(The engineer formerly known as Homeless)
Was Homeless a requisite step to becoming an engineer?
Are you planning a carrier move?
--
Boris Mohar

Boris Mohar wrote:
On 22 Dec 2004 09:31:21 -0800, wrote:


No, it does not. The coupling between the rest of the coil and the
shorted turn, if it's a single turn, is tiny: if the voice coil has
100 turns, as you say, and one of them is shorted, then, assuming
the magnet assembly provides PERFECT flux coupling (ot doesn't
come anywhere near close), then the coupling to the shorted turn is
all of 1/99.

That is turns ratio of primary to secondary, the secondary being
the
shorted turn. If the flux coupling was perfect it would be 100%.
Like you
say, it is not.
Now consider that the voice coil is, in NORMAL operations,
surrounded
by one whopper of a shorted secondary turn, that being the front
metal plate of the magnet assembly, and consider this much lower-
resistance shorted secondary does NOT "cause a much larger drop
in AC impedance" as a result, it should become apparent why your
"shorted turn" theory doesn't work.
That metal plate is way to far from the coil to couple to the voice
coil.

No, it is not at all. The distance between the coil and
the face of the cap is actually quite small compared to
extend of the voice coil.

Have you seen voice coils wound on an Aluminum mandrel. The mandrel
does not
go the full circumference of the voice coil. There is a tiny gap.
That is
to prevent the shorted turn.

Actually I have seen and specifically examined many thousands of
voice coils, thank you for asking.

More likely causes of the problem include:

1. The voice coil was heated sufficiently that the adhesive failed
on the voice coil windings, and a number of voice coil windings
have now dropped off and are no longer in the gap. With fewer turns
in the gap, the Bl product goes down, and so does efficiency.
Possible

For one client, a tweeter manufacturer, who presented me with
about 1600 failed or damaged tweeters to examine to determine
the potential cause, approximately 25% of them suffered precisely
this problem. So, yes, it's indeed "possible." (The tweeters all
were products of the US distributor of this brand of tweeters,
and all were sold to nationwide automotive sound installer, who
chose to use a brand of amps with significant DC on the amp output.
Since the tweeters were part of an electronicallt biamped system,
the installer assumed no DC protection was needed. Big mistake)

2. A voice coil winding has come loose and become lodged in the gap.
Possible

IN the above case, maybe 5% suffered from this.
3. Heat has sufficiently deformed the suspension that the diaphragm
or voice coil is now mechanically constrained.
Possible

THis was actually the case in the majority, maybe 60%

4. The connection points on the flex lead leading into the voice
coil are compromised and exhibit a high resistance.
That one you should be able to hear.

Well, the problem the guy was describing was a reduced level.
This one is pretty rare, I have seen it only in one specific
model of tweeter whose lead dress was such that there was a high
degree of fatigue right at the attachment mpoint of the lead-
in wire to the front plate. Instead of flexing evenly over its
length, all the flexig occured at the attachment point.

5. While it takes a LOT of heat to do this, getting the magnet
assembly sufficiently hot can cause partial loss of magnetic
flux.
It would have to be on fire to go past the curie point

No, it wouldn't. You can get quite a significant reduction
before you reach the Curie point, well before you reach it.
Since efficiency goes as the square of the Bl product, and since
most tweeter magnets are run right about at saturation of the
front plate gap edge (which puts it at a non-optimum point on
the B-H curve) it doesn't take a lot to kick it over and loose
a fair bit of energy. A 10% loss in flux density will give you
a 2 dB loss in efficiency, and you could do that by heating simple
thermal OR mechanical stress: banging the tweeter without dislodging
the magnet can result in a loss in flux density.

And these are just some examples of problems I have actually SEEN.
I could speculate on many more plausible causes.

But I do not recall ever finding a failure due to shorted windings.
Did you look?

Yes, in many, MANY thousands of tweeters. I was explicitly contracted
on a number of occasions to do just that.

(The engineer formerly known as Homeless)
Was Homeless a requisite step to becoming an engineer?
Are you planning a carrier move?

Nope, there are enough Homeless engineers and engineer wannabe's
as it is, I'll stay gainfully employed as a real engineer instead.

wrote:
Boris Mohar wrote:
On 22 Dec 2004 09:31:21 -0800, wrote:


snip
4. The connection points on the flex lead leading into the voice
coil are compromised and exhibit a high resistance.
That one you should be able to hear.

Well, the problem the guy was describing was a reduced level.
This one is pretty rare, I have seen it only in one specific
model of tweeter whose lead dress was such that there was a high
degree of fatigue right at the attachment mpoint of the lead-
in wire to the front plate. Instead of flexing evenly over its
length, all the flexig occured at the attachment point.

I had a set of tweeters which failed this way after a transient surge. I
was able to unwind one turn off the vc and reattach....

jak

all testing revealed no problems (thanks for the procedures would have liked
to find a problem without opening up the actual tweeter) disassembly
revealed that the pole piece had indeed shifted and was pressing up against
the inside of the voice coil thus physically damping the diaphragm. ends up
explaining why the output was still consistent across it's frequency range
just down 10dB. thanks for all the expert technical advice

"Troll" wrote in message
...
I have dealt with blowing a tweeter and it ends up producing no output
usually because the voice coil gets fried by the too high output. I
currently have a stereo pair of speakers whose tweeter on one side is down
approximately 10 dB from the other tweeter, starting at rolloff and
continuing up through 20 kHz. I have switched the tweeters in question
from one cabinet to the other with the same results, therefore it's
definitely the tweeter and not the crossover or amp etc. My question then
is if it is possible to PARTIALLY blow a tweeter and if yes how could I
confirm this? The voice coil/diaphragm assembly is sold and is field
replaceable but I don't want to spend that money only to find it's still
down 10 dB i.e. there is something wrong with the magnet having shifted
etc. Can I somehow test the voice coil or physically inspect it for damage
to confirm it's just that prior to ordering a replacement or should I go
ahead a get a whole new tweeter. Morel MDT-29. Thanks and Happy Holidays!
PS They are Ferro Fluid cooled, can that spill out / does it need to be
replaced?