Hello DIYs,
I'm considering building a single-stator electrostatic headphone. I have noticed, however, that some say that such a design is inherently unlinear because of the differences in the electrostatic field force as the membrane moves back and forth. Apparently this is because the force acting on the membrane varies by the distance to the stator.
This surprises me, though, as I would guess that the variation of the electric field will be quite small given the sub-millimeter movements of the membrane. My reason for saying this is that the electric field emanated by the membrane - or the stator - to me appears to be more comparable to infinite flat surfaces than to a point source, the latter having square variations in force on the membrane (i.e. my guess).
I would assume that a single stator (constant charge) would exhibit very low electric field variations since the field is more or less the same. Or have I missed something?
I have attached a drawing that hopefully illustrates my thoughts.
Any of you have ideas on this?
Greetings 😉
Jesper
I'm considering building a single-stator electrostatic headphone. I have noticed, however, that some say that such a design is inherently unlinear because of the differences in the electrostatic field force as the membrane moves back and forth. Apparently this is because the force acting on the membrane varies by the distance to the stator.
This surprises me, though, as I would guess that the variation of the electric field will be quite small given the sub-millimeter movements of the membrane. My reason for saying this is that the electric field emanated by the membrane - or the stator - to me appears to be more comparable to infinite flat surfaces than to a point source, the latter having square variations in force on the membrane (i.e. my guess).
I would assume that a single stator (constant charge) would exhibit very low electric field variations since the field is more or less the same. Or have I missed something?
I have attached a drawing that hopefully illustrates my thoughts.
Any of you have ideas on this?
Greetings 😉
Jesper
Hi Jesper
For a single-stator ESL, you can't have constant charge on either the membrane or the stator.
You only get constant charge with constant voltage, but then there's no output.
There's a nice explanation of the fields and forces here: http://dept.physics.upenn.edu/~uglabs/lab_manual/electric_forces.pdf
The important bit (at top of page 2) is that the force on the membrane is proportional to the square of the voltage between it and the stator, and inversely proportional to the distance between them.
So e.g. with a bias of 500V and a signal of +-50V, you get about 2.5% 2'nd harmonic distortion. Movement of the membrane will add a bit more distortion too.
It might sound good anyway (a bit of low-order distortion isn't a disaster).
There's another problem: if the membrane's stretched too tight you get no bass, but if it's too loose it gets sucked onto the stator unless the bias voltage is low, in which case sensitivity is also low.
Why not just build a push-pull version? It's no harder, maybe even easier than single-sided.
Perhaps you want to drive the headphone with a single-ended signal e.g. the output from a single triode?
This can still be done with a normal two-stator headphone by only applying the audio to one stator and earthing the other stator.
Regards - Godfrey
For a single-stator ESL, you can't have constant charge on either the membrane or the stator.
You only get constant charge with constant voltage, but then there's no output.
There's a nice explanation of the fields and forces here: http://dept.physics.upenn.edu/~uglabs/lab_manual/electric_forces.pdf
The important bit (at top of page 2) is that the force on the membrane is proportional to the square of the voltage between it and the stator, and inversely proportional to the distance between them.
So e.g. with a bias of 500V and a signal of +-50V, you get about 2.5% 2'nd harmonic distortion. Movement of the membrane will add a bit more distortion too.
It might sound good anyway (a bit of low-order distortion isn't a disaster).
There's another problem: if the membrane's stretched too tight you get no bass, but if it's too loose it gets sucked onto the stator unless the bias voltage is low, in which case sensitivity is also low.
Why not just build a push-pull version? It's no harder, maybe even easier than single-sided.
Perhaps you want to drive the headphone with a single-ended signal e.g. the output from a single triode?
This can still be done with a normal two-stator headphone by only applying the audio to one stator and earthing the other stator.
Regards - Godfrey
Hi Godfrey,
- and thanks for your reply. Hmmm ... First it made my head spin ("why could it be that a single stator had no output with constant charge ...?") but I guess the reason is that the output power available to the electrostatic system is dependent upon the resistor in series with the constant charge membrane (right?) And if this is about 1 Gohm, the power is virtually zero, thus no output.
My reason for not wishing to use two stators is that I'd like to have as few "obstructions" between my ears and the membrane. But I'd like to ponder your idea about two stators, one of the grounded. And, yes, I would like to drive it from a single-ended source.
Thanks again for your reply - very clarifying (supposing my comments above is correct) 😉
Jesper
- and thanks for your reply. Hmmm ... First it made my head spin ("why could it be that a single stator had no output with constant charge ...?") but I guess the reason is that the output power available to the electrostatic system is dependent upon the resistor in series with the constant charge membrane (right?) And if this is about 1 Gohm, the power is virtually zero, thus no output.
My reason for not wishing to use two stators is that I'd like to have as few "obstructions" between my ears and the membrane. But I'd like to ponder your idea about two stators, one of the grounded. And, yes, I would like to drive it from a single-ended source.
Thanks again for your reply - very clarifying (supposing my comments above is correct) 😉
Jesper
Last edited:
electrical low cut-off frequency?
Hmmm...
BTW - can I ask you, or maybe another DIY'er reads this, how the electrical low cut-off frequency in an electrostatic headphone with two stators, and constant charge is calculated?
Would it be reasonably correct to say that it's a highpass filter formed by the combination of the capacitance between either of the stators and the membrane (parallel capacitance), and then the resistance of the charge resistor?
That is, if the capacitance between either of the stators and the membrane is e.g. 180 pf and and the membrane's charge capacitance is 2 Gohms, then the low cut-off frequency is: 1/(2*pi*2 Gohms*180pf/2) = 0,88 Hz ?
Best regards,
Jesper
Hmmm...
BTW - can I ask you, or maybe another DIY'er reads this, how the electrical low cut-off frequency in an electrostatic headphone with two stators, and constant charge is calculated?
Would it be reasonably correct to say that it's a highpass filter formed by the combination of the capacitance between either of the stators and the membrane (parallel capacitance), and then the resistance of the charge resistor?
That is, if the capacitance between either of the stators and the membrane is e.g. 180 pf and and the membrane's charge capacitance is 2 Gohms, then the low cut-off frequency is: 1/(2*pi*2 Gohms*180pf/2) = 0,88 Hz ?
Best regards,
Jesper
You'd have to be very careful not to let the membrane touch your ear, otherwise the membrane is obstructed and your ear gets zapped 😱
I've been thinking of building two-stator headphones too. I'm starting to think the idea of single-ended drive with two stators might be a good one.
Partly for safety but also because with the earthed plate against your ear, the ear-to-plate capacitance won't affect the signal.
Cheers - Godfrey
Hey Godfrey - appreciate that you consider my health 😀 . I was intending to put something in between the membrane and my ears, though. Having 500 volts just next to a very sensitive part of me might not be that conducive to a relaxed listening experience 😱
Cheers
Jesper
Cheers
Jesper
Firstly that should be: 1/(2*pi*2 Gohms*180pf*2) = 0,22 Hz because the membrane sees both capacitances in parallel i.e. 180 + 180 = 360pF.
However that is not a low-frequency cut-off in terms of frequency response. It is the frequency below which the membrane no longer operates at constant charge. So above that frequency, distortion reduces with increasing frequency.
There is no low-frequency electrical cut-off in an electrostatic headphone, only a mechanical one.
There will most likely be a low-frequency electrical roll-off in the drive circuit, though. For example, the stator is normally connected to earth with a resistor and the signal is applied to it through a capacitor. So with a 100K resistor and 100nF capacitor, the roll-off will be about 16Hz.
Regards - Godfrey
... and as long as you're putting something there, it may as well be another stator. 😉
oh - yes parallel capacitors ... C1 + C2 ... it slipped - thanks.
F
Why is this - to protect from electrostatic discharge if the membrane touches the stator?
F
Why is this - to protect from electrostatic discharge if the membrane touches the stator?
No, I was thinking of the DC decoupling at the output of the amplifier.
For example if you want to drive the headphone from the plate of a valve, the valve plate will have a few hundred volts DC on it, and you'd want to block that from the headphone.
With transformer coupling it would not be needed, but then the transformer itself would have a low-frequency roll-off.
In general, there is likely to be an electrical low-frequency roll-off somewhere - in the amplifier and/or the coupling between it and the headphone.
The electrical roll-off is easy to control though e.g. by choice of components.
The mechanical roll-off due to membrane tension is normally the one that determines the final result
For example if you want to drive the headphone from the plate of a valve, the valve plate will have a few hundred volts DC on it, and you'd want to block that from the headphone.
With transformer coupling it would not be needed, but then the transformer itself would have a low-frequency roll-off.
In general, there is likely to be an electrical low-frequency roll-off somewhere - in the amplifier and/or the coupling between it and the headphone.
The electrical roll-off is easy to control though e.g. by choice of components.
The mechanical roll-off due to membrane tension is normally the one that determines the final result
Just noticed that you are in Cape Town. Right now Denmark is snow-clad, sunny and white.
Yesterday it snowed most of the day and I would say that added about 10-15 centimeters of snow to the layer that was already there. I find it to be beautiful, however, imagine there can be quite a contrast to Cape Town by now?
My sister will be going to SA in a few weeks. Apparently, there's some education offered by a South African institution (she's a lawyer) and that means she gets the opportunity to travel to SA. Would like to go myself but tend to be a bit wary about it due to CO2 emissions ...
Anyway, thanks for clarifying again, Godfrey 😉
Greetings,
Jesper
Yesterday it snowed most of the day and I would say that added about 10-15 centimeters of snow to the layer that was already there. I find it to be beautiful, however, imagine there can be quite a contrast to Cape Town by now?
My sister will be going to SA in a few weeks. Apparently, there's some education offered by a South African institution (she's a lawyer) and that means she gets the opportunity to travel to SA. Would like to go myself but tend to be a bit wary about it due to CO2 emissions ...
Anyway, thanks for clarifying again, Godfrey 😉
Greetings,
Jesper
P.S.: Since you are also considering making headphones. Have you seen the comments made by phil47 on rectangular stator wire headphones here:
http://www.diyaudio.com/forums/headphones/132573-has-anybody-made-els-headphone-5.html
and then starting on page 5.
regards,
Jesper
http://www.diyaudio.com/forums/headphones/132573-has-anybody-made-els-headphone-5.html
and then starting on page 5.
regards,
Jesper
Thank you for the link, Jesper.
There is some very nice work in that thread! I doubt anything I build will be as pretty.
I plan to use PCB (printed circuit board) material for the stators and spacers.
The stators will have copper only on the inside surface, not the outside where it can be touched. The spacers will also have the copper on the inside to make the electrical connection to the membrane.
The copper can easily be etched away where it is not wanted, for example around the outside of the stators where the spacers will be glued on.
Perhaps there should be narrow earthed copper tracks around the outside edge of the stators and spacers where they can be touched as well.
____________________________________
There is a lot of discussion on that thread about what surface coating to use on the membrane. I don't think that is necessary.
Many years ago there were several articles in Wireless World about electrostatic headphones. In one of these, the author claimed that good results were obtained using polythene "cling-wrap" film with no coating on it.
This is the stretchy plastic film normally used in the kitchen to cover leftover food. I used it for an experimental electrostatic loudspeaker and found it worked fine.
Also: after heat-shrinking, it has a suitable tension for small diaphragms such as in headphones, so no fancy stretching equipment is needed.
So I will try that first. It is free and hopefully it will work.
Regards - Godfrey
There is some very nice work in that thread! I doubt anything I build will be as pretty.
I plan to use PCB (printed circuit board) material for the stators and spacers.
The stators will have copper only on the inside surface, not the outside where it can be touched. The spacers will also have the copper on the inside to make the electrical connection to the membrane.
The copper can easily be etched away where it is not wanted, for example around the outside of the stators where the spacers will be glued on.
Perhaps there should be narrow earthed copper tracks around the outside edge of the stators and spacers where they can be touched as well.
____________________________________
There is a lot of discussion on that thread about what surface coating to use on the membrane. I don't think that is necessary.
Many years ago there were several articles in Wireless World about electrostatic headphones. In one of these, the author claimed that good results were obtained using polythene "cling-wrap" film with no coating on it.
This is the stretchy plastic film normally used in the kitchen to cover leftover food. I used it for an experimental electrostatic loudspeaker and found it worked fine.
Also: after heat-shrinking, it has a suitable tension for small diaphragms such as in headphones, so no fancy stretching equipment is needed.
So I will try that first. It is free and hopefully it will work.
Regards - Godfrey
Hello godfrey,
Thanks for sharing your ideas. I'm thinking - if you are interested I can ask my sister to bring some 0,9 um foil, which is what I initially will be using, to SA when she flies (have to ask her first, though), and then she can post it while she is in SA?
Let me know if that is interesting to you.
Best,
Jesper
Thanks for sharing your ideas. I'm thinking - if you are interested I can ask my sister to bring some 0,9 um foil, which is what I initially will be using, to SA when she flies (have to ask her first, though), and then she can post it while she is in SA?
Let me know if that is interesting to you.
Best,
Jesper
BTW - have you considered using paper as a spacer? It's readily available and breakdown voltage is significantly higher than air....
best,
Jesper
best,
Jesper
Hi Jesper
Thank you for the kind offer, but please don't bother your sister on my account.
I'm sure I'll be able to find what I need locally.
Regards - Godfrey
Thank you for the kind offer, but please don't bother your sister on my account.
I'm sure I'll be able to find what I need locally.
Regards - Godfrey
I suppose paper could be used. It would need a number of layers to get the desired thickness, but that also makes it easy to change the spacing a little if needed.
Perhaps thin plastic sheet would be better than tree-paper? It would insulate better and won't absorb moisture.
PCB is easiest for me though as I already have a sheet of it.
Perhaps thin plastic sheet would be better than tree-paper? It would insulate better and won't absorb moisture.
PCB is easiest for me though as I already have a sheet of it.
- Status
- Not open for further replies.