As far as I know, there are billions of electret microphones of every quality right up to the highest out there. I'm pretty sure some headphones were electret (Koss?) and have always wondered why they didn't become as common as in mics. Or why they weren't used for stationary systems. Anybody know?
Of course, when you read about them, it sounds like a cross between a permanent magnet and a perpetual motion machine.
Of course, when you read about them, it sounds like a cross between a permanent magnet and a perpetual motion machine.
Not speaking with authority here, but my gut feeling would be cost -- I think electret films are very expensive per unit area, and considering that a speaker needs so much more area than a mic or a pvdf sensor...
Another guess (again, just guessing, would be nice if someone could shed more light on this): maybe the charge imbalance attainable with electret films is much less than what can be attained by coating & electrifying the film, leading to a much lower sensitivity for electret ESL speakers, all other factors being equal?
Kenneth
Hi,
I know of manufacturers trying to improve elecrets with regard to longevity and amount of charge. My latest infos were that they are testing a film thats charge would be equivalent to 2kV bias.
If that film is low enough in weight and features sufficient mechanical strength and endurance a large electret apeaker panel would hold a few advantages over the classical way of biasing. There woud be no flashovers between stator and membrane and charge distribution over membrane would be very even. There would also be no need for a wall conncection to power the ESL.
jauu
Calvin
I know of manufacturers trying to improve elecrets with regard to longevity and amount of charge. My latest infos were that they are testing a film thats charge would be equivalent to 2kV bias.
If that film is low enough in weight and features sufficient mechanical strength and endurance a large electret apeaker panel would hold a few advantages over the classical way of biasing. There woud be no flashovers between stator and membrane and charge distribution over membrane would be very even. There would also be no need for a wall conncection to power the ESL.
jauu
Calvin
Thank you for summary. 2kv is a useful bias voltage esp. if you can have closer stator spacing (in the absence of HV bias).
No small benefits to get rid of HV bias, instant turn-on, perhaps pre-fabricated electret force could be modulated across membrane for various purposes, etc.
I suppose it moves the construction away from DIYers.
No small benefits to get rid of HV bias, instant turn-on, perhaps pre-fabricated electret force could be modulated across membrane for various purposes, etc.
I suppose it moves the construction away from DIYers.
Hi,
why should it dis-DIY? It´d be just a matter of sourcing the film. But that´s a problem with any other film too, nowadays that the production of certain films has ceased like some Hostaphan and Mylar types.
With regard to building things would become easier and safer, especially no handling with high voltage supplies -apart from the audio tranny..
But I´m afraid that weight-per-unit of the film would be too high. Also the mechanical strength probabely won´t be sufficient for hard mechanical tension, which would ask for different methods to achieve high values of stiffness with low tension.
jauu
Calvin
why should it dis-DIY? It´d be just a matter of sourcing the film. But that´s a problem with any other film too, nowadays that the production of certain films has ceased like some Hostaphan and Mylar types.
With regard to building things would become easier and safer, especially no handling with high voltage supplies -apart from the audio tranny..
But I´m afraid that weight-per-unit of the film would be too high. Also the mechanical strength probabely won´t be sufficient for hard mechanical tension, which would ask for different methods to achieve high values of stiffness with low tension.
jauu
Calvin
Last edited:
I have a question or two about this technique.
Let's assume we have a ESL with a 3kVpp signal at full swing.
When we attenuate the signal (we don't want the speaker att full full spl all the time) with a max attenuation of 60dB we get a 3V drive signal.
Now, sourcing the HV bias from 3kV is easy, what do you do when you only have 3V available?
Let's assume we have a ESL with a 3kVpp signal at full swing.
When we attenuate the signal (we don't want the speaker att full full spl all the time) with a max attenuation of 60dB we get a 3V drive signal.
Now, sourcing the HV bias from 3kV is easy, what do you do when you only have 3V available?
Koss used that technique in the ESP-6 headphones in the '70s. The ESP-9 can run this way in 'self energized' mode but they are definitely better running with external power. Electrets are permanently charged capacitors but you have to add mass to the membrane to be the electret. Not a good way to go.
IIRC Sony was big on electrets (though their high end condenser mics were not electrets) while AKG was externally energized condenser mics - IMO better than Sony electrets.
G²
IIRC Sony was big on electrets (though their high end condenser mics were not electrets) while AKG was externally energized condenser mics - IMO better than Sony electrets.
G²
Last edited:
What's readily avaiable is 10u PTFE or trifluorine based - cold flow, moreover it's bipolar i.e. different charges on the opposite sides.
Secondly during the construction the least problem appeared to be film, coating and bias supply altogether.
It might of been myself, not exactly mechanical engineer, but getting rigid enough rod stator appeared to be most complex.
Power can easily be derived from signal itself. Nobody is bothered by Li primary cell in most computers either.
Some wishful thinking flollows:
Can't the electret be applied atop of membrane like PTFE coated corona safe films from Dupont, Kapton if memory serves right.
Just get film and heat it up to the right remperature in strong electric field.
The latter's easily achievable due to really good electric properties of Kapton itself.
Are not we looking in the wrong place?
and off topic
Are not we helping not exactly DIYes with items like floor varnish after all...
BTW PET price is less then 2-4 eu per kilo and even at small roll of 50 kg you'll be on par when buying 30 m @ 3 eu per m.
All you need is usually a phone call away...
"Just get film and heat it up to the right remperature in strong electric field."
A friend at Bell Labs was a key person in electret development, Dr. James West (being an Afro-American scientist in 1967 was something special too)*. As I understand electrets, you need to charge them rather than buy a roll of pre-charged stuff. That is why microphone electrets charged in a little cabinet are feasible but not big, speaker-sized membranes.
Is that right?
*Jim was also "keeper" of the world's largest anechoic chamber (at that time) and really good guy to be friends with. The electret mic invention has something to do with the decreasing supply of proprietary carbon for making telephone mics. Fascinating.
http://web.mit.edu/invent/iow/westsessler.html
A friend at Bell Labs was a key person in electret development, Dr. James West (being an Afro-American scientist in 1967 was something special too)*. As I understand electrets, you need to charge them rather than buy a roll of pre-charged stuff. That is why microphone electrets charged in a little cabinet are feasible but not big, speaker-sized membranes.
Is that right?
*Jim was also "keeper" of the world's largest anechoic chamber (at that time) and really good guy to be friends with. The electret mic invention has something to do with the decreasing supply of proprietary carbon for making telephone mics. Fascinating.
http://web.mit.edu/invent/iow/westsessler.html
Last edited:
Hi,
its not as easy as Alex suggests.
There are materials which are good elektrets like FEP and PTFE and derivatives of these. Typically the mechanical parameters are not as well. PTFE e.g shows a quite high value of temperature coefficient which might lead to a serious loss of mechanical tension , hence a considerable drop in resonance frequency. So a compound membrane consisting of a base film with good mechanical properties like PET needs to be laminated to a film or layer of a good elektret. Such elektret films may be in the thickness range 12µm to 25µm, hence the weight of such a membrane is high enough that the HF-response is affected.
A possibility might by a so called Back-Elektret, where the Elektret material is laminated onto the stator surface and a metallized membrane is used. But this is difficult to achieve/manufacture because of temperature probs (~300°C) and evenness of the stator and elektret-layer, etc. etc.
A ´self energized´ Headphone circuit is described in Patent US 3,632,903 of Koss.
The circuit is connected to the audio transformer. A voltage doubler-rectifier feeds a capacitor with paralleled Zener-diodes.
The Cap is charged up as long as the voltage supplied by the doubler-rectifier exceeds its own voltage. The voltage over the cap remains constant as long as there is no load that requires currents. A such the cap works similar to a battery. The zener-diodes limit the voltage over the cap to a safe value for the ESL-panel. The charge the membrane needs is supplied by the cap. It only needs small current to recharge the losses caused by leakage. The doubler-rectifier on the other hand can supply larger currents to the cap as long as its voltage is higher than the cap´s own. Problems may occur when the loudness/signal-voltage is low over elongated periods and the caps voltage slowly sinks.
A later elaboration of this circuit was patented to Koss under US 3,992,585. Here to the earlier rectifier-cap circuit a oscillator-voltage-doubler circuit was added.
jauu
Calvin
its not as easy as Alex suggests.
There are materials which are good elektrets like FEP and PTFE and derivatives of these. Typically the mechanical parameters are not as well. PTFE e.g shows a quite high value of temperature coefficient which might lead to a serious loss of mechanical tension , hence a considerable drop in resonance frequency. So a compound membrane consisting of a base film with good mechanical properties like PET needs to be laminated to a film or layer of a good elektret. Such elektret films may be in the thickness range 12µm to 25µm, hence the weight of such a membrane is high enough that the HF-response is affected.
A possibility might by a so called Back-Elektret, where the Elektret material is laminated onto the stator surface and a metallized membrane is used. But this is difficult to achieve/manufacture because of temperature probs (~300°C) and evenness of the stator and elektret-layer, etc. etc.
A ´self energized´ Headphone circuit is described in Patent US 3,632,903 of Koss.
The circuit is connected to the audio transformer. A voltage doubler-rectifier feeds a capacitor with paralleled Zener-diodes.
The Cap is charged up as long as the voltage supplied by the doubler-rectifier exceeds its own voltage. The voltage over the cap remains constant as long as there is no load that requires currents. A such the cap works similar to a battery. The zener-diodes limit the voltage over the cap to a safe value for the ESL-panel. The charge the membrane needs is supplied by the cap. It only needs small current to recharge the losses caused by leakage. The doubler-rectifier on the other hand can supply larger currents to the cap as long as its voltage is higher than the cap´s own. Problems may occur when the loudness/signal-voltage is low over elongated periods and the caps voltage slowly sinks.
A later elaboration of this circuit was patented to Koss under US 3,992,585. Here to the earlier rectifier-cap circuit a oscillator-voltage-doubler circuit was added.
jauu
Calvin
Last edited:
Just 2 cents
Well my suggestion was to give it a try using off the shelf product and see if it fits.
Inverted ESL idea was also at the surface - stator not necessaralry is metal. @ main material/coating ratio higher than 10 coating in laminate gonna follow the base. With no temperature excursions it will stay forever...
Well my suggestion was to give it a try using off the shelf product and see if it fits.
Inverted ESL idea was also at the surface - stator not necessaralry is metal. @ main material/coating ratio higher than 10 coating in laminate gonna follow the base. With no temperature excursions it will stay forever...
Last edited:
Hi,
For a portable bias supply I would start with a "joule thief" circuit (wikipedia it) and replace the LED load with a step up winding and voltage multiplier. You could have it separate from the audio circuitry so you could turn it on and charge up the ESL's so they are ready to use before turning on the audio.
For a portable bias supply I would start with a "joule thief" circuit (wikipedia it) and replace the LED load with a step up winding and voltage multiplier. You could have it separate from the audio circuitry so you could turn it on and charge up the ESL's so they are ready to use before turning on the audio.
Last edited:
Hi,
the ´joule thief´ is a very basic oscillator circuit on its own, working with low supply voltages, generating only small supply voltages. It needs a dc-input voltage, so some kind of rectifier circuit and energy storage is required. Generating high voltages from low-voltage sources requires considerable supply currents easily in the range of hundreds of mA.
So, since the theme of the thread is a biasless ESL or an ESL not requiring an external power supply, it folllows that the bias generator circuit needs to be fed with already high supply voltages to keep current consumption low enough.
jauu
Calvin
the ´joule thief´ is a very basic oscillator circuit on its own, working with low supply voltages, generating only small supply voltages. It needs a dc-input voltage, so some kind of rectifier circuit and energy storage is required. Generating high voltages from low-voltage sources requires considerable supply currents easily in the range of hundreds of mA.
So, since the theme of the thread is a biasless ESL or an ESL not requiring an external power supply, it folllows that the bias generator circuit needs to be fed with already high supply voltages to keep current consumption low enough.
jauu
Calvin
My main issue with this idea I think I Already mentioned it earlier.
Getting bias from a high voltage source is easy.
What do you you with the attenuated signal?
A regular preamp can easily attenuate 60dB?
I'm speaking under the assumption that we want a fixed bias.
In the end it would seem it's both easier and better to go for a fixed separate bias supply fed with it's own 230V?
Getting bias from a high voltage source is easy.
What do you you with the attenuated signal?
A regular preamp can easily attenuate 60dB?
I'm speaking under the assumption that we want a fixed bias.
In the end it would seem it's both easier and better to go for a fixed separate bias supply fed with it's own 230V?
Last edited:
The best solution I found for powering the ESL bias from the audio signal is outlined in the Koss patent US 3,992,585 Calvin mentioned earlier.
The circuit has three parts:
1) A small battery charger powered directly from the amplifier output, not from the step-up transformer output.
2) High frequency oscillator powered by the battery used to drive a voltage multiplier to supply the bias voltage.
3) An electronic switch that senses the audio output from the amplifier and turns on the high frequency oscillator only when you are listening to music so as not to drain the battery over night.
This solves the problem of the bias voltage falling to low levels when you are listening quietly for extending lengths of time.
You would have to decide if the added complexity is worth the trouble.
Many years ago I built one set, but have not bothered with it since then.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.