Over the past decades various manufacturers have briefly introduced
curved transducers in order to widen the high frequency sweet spot of
planar transducers. As a solution this has never seemed to last long.
Why?

On 10 Apr 2005 09:41:04 -0700, "andy" wrote:

Over the past decades various manufacturers have briefly introduced
curved transducers in order to widen the high frequency sweet spot of
planar transducers. As a solution this has never seemed to last long.
Why?

Curved diaphragms don't flap nicely - and of course it is only
possible to curve a diaphragm in one plane. Quad achieve the 3D
curvature effect with flat diaphragms and delay lines.

d

Pearce Consulting
http://www.pearce.uk.com

"andy" wrote in news:1113151264.247544.279820
@g14g2000cwa.googlegroups.com:

Over the past decades various manufacturers have briefly introduced
curved transducers in order to widen the high frequency sweet spot of
planar transducers. As a solution this has never seemed to last long.

Martin Logan would be appalled to hear you call their product lifetime as
"brief".

Thanks for the interesting pointer. Perhaps an impossible request these
days but are you aware an objective review of one of their speakers
with the curved panels?

On 11 Apr 2005 01:13:39 -0700, "andy" wrote:

Thanks for the interesting pointer. Perhaps an impossible request these
days but are you aware an objective review of one of their speakers
with the curved panels?

No, just the usual HFN stuff.

--

Stewart Pinkerton | Music is Art - Audio is Engineering

"andy" wrote in news:1113207219.658971.246890
@f14g2000cwb.googlegroups.com:

Thanks for the interesting pointer. Perhaps an impossible request these
days but are you aware an objective review of one of their speakers
with the curved panels?

I don't know of any reviews that I would promote as being "objective". All
the ones I know about are rather subjective in nature -- at best they throw
in a few measurments that seem to have little bearing on the review itself.

That being said, I do like their speakers (but I like almost every
electrostatic speaker that I have heard).

Murray Peterson wrote:
"andy" wrote in news:1113207219.658971.246890
@f14g2000cwb.googlegroups.com:

Thanks for the interesting pointer. Perhaps an impossible request these
days but are you aware an objective review of one of their speakers
with the curved panels?

I don't know of any reviews that I would promote as being "objective". All
the ones I know about are rather subjective in nature -- at best they throw
in a few measurments that seem to have little bearing on the review itself.

That being said, I do like their speakers (but I like almost every
electrostatic speaker that I have heard).

Here's my subjective review.

Martin Logans are breathtaking. I love 'em, and consider them at least
a match for B&Ws at the same price point. Unfortunately, that's only
if you're sitting in the EXTREMELY NARROW sweet spot! Off-axis listening
is really badly compromised vs. almost any other decent speakers.

So the curve might help some, but not much.
Colin

The curvature helps to readuce "beaming" (on frontal side) of
frequencies of wavelengths smaller than the speaker panel width.
Whether this enhances the sweet spot is another issue.

There is a series of articles in SpeakerBulder on a "focussed array"
electrostat, using quite the inverted idea. I believe the author was
not impressed.

I woudl expect, but do not know this, that as Pierce perhaps suggests,
that such may lead to some non-linear behavior at larger diaphragm
excursions. Based on the assumption that the compliance is not similar
for same deflections in opposing directions. A flat diaphragm would
not have this problem.
I am not aware of issues of lifespan of such MLogan etc. panels.

Further, the Logans use some foam to separate the diaphragm from the
stators. Such may reduce LF efficiency, by reducing volume velocity.

Quad uses delay lines, Audiostatic a set o parallel stators (but I'm
not convinced this actually works).

Your premise though, is not correct.



On 10 Apr 2005 09:41:04 -0700, "andy" wrote:

Over the past decades various manufacturers have briefly introduced
curved transducers in order to widen the high frequency sweet spot of
planar transducers. As a solution this has never seemed to last long.
Why?

The B&W DM 70 is one of the examples I was referring to. This was a
hybrid like many of the Martin Logan designs with a cone bass unit and
an electrostatic mid/tweeter. The difference is that the electrostatic
unit is not a line source but closer to a more traditional point
source: it is wide but not tall. My original question concerned why
this type of approach has not been successfull.

I woudl expect, but do not know this, that as Pierce perhaps
suggests,
that such may lead to some non-linear behavior at larger diaphragm
excursions. Based on the assumption that the compliance is not
similar
for same deflections in opposing directions. A flat diaphragm would
not have this problem.

Stiffness is only relevant at low frequencies and, unfortunately,
standard flat electrostatics have problems here anyway. Curvature is
not going to make things better but my guess would be that it does
little harm. However, I have not done the sums and am happy to be
corrected by someone who has.

What is clear is that people like Quad who invested considerable effort
in the area came up with more complicated solutions.

Further, the Logans use some foam to separate the diaphragm from the
stators. Such may reduce LF efficiency, by reducing volume velocity.

Do you have a reference for this? Thanks.

andy wrote:
The B&W DM 70 is one of the examples I was referring to.
This was a hybrid like many of the Martin Logan designs
with a cone bass unit and an electrostatic mid/tweeter.
The difference is that the electrostatic unit is not a
line source but closer to a more traditional point
source: it is wide but not tall. My original question
concerned why this type of approach has not been successfull.

The DM-70 is also implementd in a very different fashion
than any of the other examples raised thus far. While the
HF radiator is quite wide, one must realize that it is made
up of a number of individual radiating element (I don't recall
the exact number but 13 comes to mind). Each element, though
physically part of the HF drivers, is indvidually edge-
restrained. This neatly solves the mechanical problem mentioned
earlier in this thread:

"Curved diaphragms don't flap nicely - and of course
it is only possible to curve a diaphragm in one plane."

by simply not dealing with the curved diaphragm problem: every one
of the diaphragms is flat: the array of diaphragms itself describes
a segment of a circle.

By the way, the notion implicit in the statement "Curved
diaphragms don't flap nicely" is that flat diaphragms do.
In fact, NO diaphragm "flaps nicely." If you think that
somehow that electrostatic diaphragms are these well-
behaved planes that behave nicely and cooperate to the applied
electrostatic forces by mobving uniformly, do yourself a
favor and NEVER EVER look at the actual motions of such
diaphragms: they are almost chaotic in their mis-behavior
at almost all frequencies with the posisble exception of the
very lowest in the frequency range for the unit.

But, to the original question: the reason they are not
successful? Well, consider the cost of implementing a robust,
reliable, consistent electrostatic system of ANY kind. Now,
factor in the potential size of the market and you start
running quickly into a serious mismatch. TO do it right is
not cheap, but the market is so small, that you can't afford
to do it right. You're stuck.

Things like the Martin Logan, the Quad and the DM-70 represent
a tiny portion of a small segment of a limited market.

Thus, I might suggest that the lack of success is as much or
more due to simply economics as it is to any notion of
technological issues. If there is not enough money to be made
in solving the problem, the problem is not going to get solved.
Or, conversely, unless there's money in solving a problem, the
problem might as well not even exist.

On 12 Apr 2005 04:34:23 -0700, "andy" wrote:

The B&W DM 70 is one of the examples I was referring to. This was a
hybrid like many of the Martin Logan designs with a cone bass unit and
an electrostatic mid/tweeter. The difference is that the electrostatic
unit is not a line source but closer to a more traditional point
source: it is wide but not tall. My original question concerned why
this type of approach has not been successfull.

Possibly because it doesn't work so well as a line source? And of
course it *has* been successful in the most famous implementation of
all, the Quad '63.

--

Stewart Pinkerton | Music is Art - Audio is Engineering

Hey Andy,

Yeah, bass is not their strength typically, though Audiostatics are
not bad (with their doble transformer invention).

A ref? Well, see RS Sander's excellent book for example. The section
on segmentation (p21 onwards). This is the way to make a
well-radiating tweeter for example: segement the large panel into two
panels with one being ~1/2" wide.

I would imagine, though that stiffnes is relevant at higher
frequencies at increasingly higher SPLs (i.e. when deflection remains
constant)...

And wrt to onlinear behavior of curved designs, see fig10-38 there
(p133)

Aah..the Dm70...loved them, but had to tear em apart....their reconed
woofers now serve as my subs...


On 12 Apr 2005 05:01:45 -0700, "andy" wrote:

I woudl expect, but do not know this, that as Pierce perhaps
suggests,
that such may lead to some non-linear behavior at larger diaphragm
excursions. Based on the assumption that the compliance is not
similar
for same deflections in opposing directions. A flat diaphragm would
not have this problem.

Stiffness is only relevant at low frequencies and, unfortunately,
standard flat electrostatics have problems here anyway. Curvature is
not going to make things better but my guess would be that it does
little harm. However, I have not done the sums and am happy to be
corrected by someone who has.

What is clear is that people like Quad who invested considerable effort
in the area came up with more complicated solutions.

Further, the Logans use some foam to separate the diaphragm from the
stators. Such may reduce LF efficiency, by reducing volume velocity.

Do you have a reference for this? Thanks.