Graphene-based ESL reported

It seems suitable perhaps for capacitive pressure transducers. An electrostatic transducer that is suitable for audio should have constant charge. This material has high conductivity according to the Wikipedia article, so it would work in constant voltage mode, which is inherently non-linear.
 
Member
Joined 2008
Paid Member
It seems suitable perhaps for capacitive pressure transducers. An electrostatic transducer that is suitable for audio should have constant charge. This material has high conductivity according to the Wikipedia article, so it would work in constant voltage mode, which is inherently non-linear.

Moves us closer to the acoustic ideal: a diaphragm as light as air.

About the constant-charge issue, wouldn't it be OK to have a large resistor in series with the diaphragm or must the diaphragm itself keep the charge from moving around on the diaphragm?

Room temperature super-conductive cone-driver voice coils are another ideal the graphene addresses.

Ben
 
Bill Gates pours cash into graphene condoms

I think this qualifies for this forum. Here's a link to a description of an electrostatic driver with a graphene diaphragm:

nanotechweb

I haven't seen any reports of sheets of graphene big enough to make room-filling sound, though.

(Here's a Wikipedia link for those new to graphene: Graphene)

Few

This may be the answer for large sheets of graphene.

Lead ONTO your pencil: Bill Gates pours cash into graphene condoms ? The Register
 
Member
Joined 2008
Paid Member
.. I have enough trouble already convincing my wife that my relations with the HiFi are strictly platonic

.. new cucumber-shaped ESL drivers have great dispersion

.. first it was using Halo Shampoo for coating the membrane, now its...

.... and a million more dumb jokes

Ben
 
About the constant-charge issue, wouldn't it be OK to have a large resistor in series with the diaphragm or must the diaphragm itself keep the charge from moving around on the diaphragm?

Hi. I don't think it would be OK, as the distribution of charge across the diaphragm would be unstable.

I believe that finding a diaphragm coating of the requisite (high) resistivity and durability is a major headache when designing ESLs. As is the ability to operate properly in the context of a wide range of ambient humidity.
 
Member
Joined 2008
Paid Member
Hi. I don't think it would be OK, as the distribution of charge across the diaphragm would be unstable.

I believe that finding a diaphragm coating of the requisite (high) resistivity and durability is a major headache when designing ESLs. As is the ability to operate properly in the context of a wide range of ambient humidity.
What if we had resistive plates instead of a resistive diaphragm?

Dayton-Wrights use very strong but open plastic structures for plates with a metallic conductive coating applied by spray-painting to the surfaces that face the diaphragm. But that coating could be more resistive rather than conductive. Likewise for the way the Quads are made, I think, with printed circuit boards.

Ben
 
What if we had resistive plates instead of a resistive diaphragm?

Dayton-Wrights use very strong but open plastic structures for plates with a metallic conductive coating applied by spray-painting to the surfaces that face the diaphragm. But that coating could be more resistive rather than conductive. Likewise for the way the Quads are made, I think, with printed circuit boards.

Ben

Resistive stators would result in frequency attenuation just like segmentation.
And it would not do what we like: constant charge on the moving diaphragm.

The diaphragm can be either constant charge (high resistance coating) or constant voltage (low resistance coating) or something in between depending on the resistance and speed of movement.

With constant voltage the charge on the film becomes a function of the position. This can be understood if we look at the film and stators as a capacitor, for which we know Q=C*V. C changes because of the change in position (which represents the gap between our capacitor plates), so now you can either keep Q or V constant, not both. With a low resistance coating, V is held constant by the power supply hence Q must change. Change of Q means change of driving force F = Q*E, which means F becomes a function of position.

We would like the driving force to be a function of only the audio signal (E). When it becomes a function of position we get distortion.
 
Member
Joined 2008
Paid Member
arend-jan -

Many thanks for that explanation. I understand now (and should have read that chapter in Hunt's book more carefully) but shouldn't the effect be the same whether on the diaphragm or stators?

Can't we supply the bias to the plates uniformly (or uniformly enough) so that we don't have frequency gradients (or at least keeping only the frequency gradients we like)? But still keep the resistance high?

There must be other advantages to fussing with the stators instead of trying to coat diaphragms with trick fluids.

Ben
 
Last edited:
Ah I see, you want to exchange roles between the diaphragm and the electrodes? So bias the plates and modulate the diaphragm? I think Final did something like that or at least they have a patent on 'inverted drive'.

In that case there would be no problem with high resistance on the plates. They would be constant charge so a constant electric field would be set up between the plates.

I have never analyzed such a system. I'd have to think about it.
 
Member
Joined 2008
Paid Member
Ah I see, you want to exchange roles between the diaphragm and the electrodes? So bias the plates and modulate the diaphragm? I think Final did something like that or at least they have a patent on 'inverted drive'.

In that case there would be no problem with high resistance on the plates. They would be constant charge so a constant electric field would be set up between the plates.

I have never analyzed such a system. I'd have to think about it.

I cant claim any expertise here, but I don't see how it matters whether the diaphragm or the stators get what signal so long as their relative values are the same. For sure, all kinds of variations in polarity of bias and other issues and more differences when you talk of direct drive amps* where the tube B+ voltage is in the mix.

Ben
*I think DD ("Sanders" amp) is hearably better. Pity these amps are so challenging to construct.
 
Member
Joined 2008
Paid Member
@bentoronto,

In your thoughts/questions about using resistive stators, are you considering the diaphragm to be highly conductive?
You mention wanting to get away from the need to coat the diaphragm with tricky fluids, so I am assuming that is the case.
Glad to see you jump in.

This discussion started when somebody pointed out that a graphene diaphragm, although vanishingly light in weight, would be highly conductive, thus contradicting a basic principle of ESL theory. So I asked about making the charge keep from migrating about using resistive coating on the stators (which are easier to work with too).

Ben
 
This discussion started when somebody pointed out that a graphene diaphragm, although vanishingly light in weight, would be highly conductive, thus contradicting a basic principle of ESL theory. So I asked about making the charge keep from migrating about using resistive coating on the stators

Applying a resistive coating to a highly conductive stator would have negligible affect since electrons will take the path of least resistance, moving about easily through the highly conductive stator rather than the coating. If the resistive coating was applied to a non-conductive stator material as you suggested in post#10 it would roll off the high frequency response and reduce sensitivity(if resistivity is high enough).

The standard push-pull ESL configuration is shown in Figure (1A). If there were no leakage paths, we could remove the bias supply and charging resistor leaving the diaphragm with a surplus or deficit of electrons depending on the polarity of the bias supply. The charged diaphragm will move in response to polarity and strength of the electric field between the stators. The field strength E = (voltage between the stators) / (distance between stators). In theory, if the diaphragm remained perfectly centered between the stators and there were no membrane modes, the ESL would behave as a CQ(constant charge) ESL even if the diaphragm was highly conductive. In practice, we need to keep the diaphragm as highly resistive as possible to achieve the theoretical(low distortion) advantages of CQ mode.
More details on CQ and CV(constant voltage) mode operation here:
http://www.diyaudio.com/forums/plan...ted-ask-why-bias-voltage-esl.html#post3589946

Turning our attention to stators, the current supplied by the step-up transformer makes a single loop from the transformer thru the ESL capacitance between the stators as shown in (1B). The ESL impedance Zesl is that of the ESL panel capacitance = 1/(2 x pi x frequenc x Cesl) and is thus inversely proportional to frequency(ie higher frequency = lower impedance). If any resistance is added to the series circuit as shown in (1C), there will be a voltage drop across these resistors. The remaining voltage will show up across the stators, producing a reduced electric field to drive the diaphragm. How much stator voltage is lost passing thru the resistances will depend on the relative magnitude of the resistance to the Zesl impedance. Since Zesl becomes smaller at higher frequencies, there will be a HF roll off starting at the frequency where Zesl becomes smaller than the resistor values.

What would happen if the stators were made from a resistive material or coating as is (1D)? Unlike the lumped resistance of in (1C) that caused a HF roll of the entire stator voltage, the resistive stator would result in HF roll off that starts lower and lower in frequency the further away the stator section was from the contact point. This is the concept used for segmented line sources to improve polar response and equalize the naturally rising response of ESLs. Think of it as a segmented stator with infinitely small segments. Note however, that if you wanted to increase the stator resistance to values needed to keep charge from moving around(>1e8) the HF roll off point would start <10Hz and you would thus lose substantial sensitivity.

Basically, making the stators highly resistive keeps us from being able to change the electric field at audio rates without substantial losses. So, what if we were to try driving the diaphragm from the transformer instead of the stators which we have made highly resistive? This would require a change to the “inverted” configuration. Since this post is getting pretty long, I’ll address that subject separately later tonight.
 

Attachments

  • CQstat_01.gif
    CQstat_01.gif
    44.3 KB · Views: 255
Member
Joined 2008
Paid Member
Thanks for good explanation of things.

But some frequency discrimination across the panel is trafficked by Quad and others to good effect.

And why have a single point of contact for the signal/bias into the high resistance stators. Easy to have the voltage applied to the whole periphery, center line, etc?

Ben
 
But some frequency discrimination across the panel is trafficked by Quad and others to good effect.
Right, this was why I made reference to the resistive stator acting like a segmented stator.
But instead of strips or rings of the stator connected by discrete ladder resistors, you have infinitely small sections fed from a distributed resistance.

However, use of a distributed resistance stator design for frequency response and polar shaping functions would use resistances 3 to 5 orders of magnitude lower than what would be required to keep charge stationary at mid and lower frequencies.

And why have a single point of contact for the signal/bias into the high resistance stators. Easy to have the voltage applied to the whole periphery, center line, etc?
The cross-section view of the ESL was only intended for illustrating the principles involved.
Certainly a center-line contact strip would be appropriate for a line source.
The attached patent shows a dozen or so other contact configurations for resistive stators that at least one person thought might be useful.
 

Attachments

  • ATC_6760455.pdf
    952.2 KB · Views: 75