I recently made ESL headphones, just to test my materials - I am in a (long) process of building big electrostatic panel speakers (160cm x 25cm), and I wanted to get a taste of the ESL sound.
I had a piece of 1mm aluminum, and I fashioned round stators from it. 5cm in diameter. I made the spacers from a cover of a plastic bucket - also 1mm in thickness. I don't know what the material is. Spacers are about 5mm in width, so the area for the diaphgram to move is 4cm in diameter. The diaphgram is cheap wrapping plastic (of the Elmu variety for those in Finland) - it's thickness is stated as 10µm on the box, material is LD-polyetylene.
I glued all the parts together with superthin soldering strips lodged in the right places for the electric contacts. Then I applied a bit of graphite powder on the diaphgram with cotton pieces. I measured the surface resistance to be in the order of 22 to 42 kilo-ohms per centimeter.
After the glue had dried, I soldered all the wires, put the drivers in their case (full sized headphones previously closed, now drilled open, with relatively tight ear cushions). For driving the ESLs I have two Hammond 125D push pull transformers and a high voltage source. Signal comes from my 1W Darling amp with approximately 250v signal swing, 5k / 4R OT.
I also put a lot of cotton on the backs of the drivers.
I thought I was good to go, and started experimenting. On the first try I got some sound, but it was very very weak even with the source and Darling amps volumes at max. Barely able to hear anything. At this point the bias was 405v and transformers at 4R / 27k. So I tried progressively bigger transformer ratios, and got a tiny bit more sound. Then I made a new voltage source, with 742v bias and transformers at 1.5R / 27k.
With this I got an ok listening volume, maybe 3/4 of what I usually use (I don't listen super loud), but even now I have Darling volume at 100% and source volume at 100%. What is more puzzling is that I have absolutely no bass. None at all. When comparing with my AKG601 headphones and with EQ (my source is a computer) I found I have no action at all pretty much below 700Hz. The response starts dropping around 1 - 2 kHz.
The sound is quite good on the highs - excellent imaging and presence. I find ESLs to be quite promising. However my first experiment is clearly not done right.
So, no volume or bass. Where did I go wrong?
Is the diaphgram area too small, with only 4cm diameter?
Did I use too much graphite - I think I did, but could this contribute to these problems?
Is the LD-polyetylene completely unsuited for ESLs? Could I have used too much tension?
I'm pretty much at a loss here for explanations. Any help would be appreciated.
I had a piece of 1mm aluminum, and I fashioned round stators from it. 5cm in diameter. I made the spacers from a cover of a plastic bucket - also 1mm in thickness. I don't know what the material is. Spacers are about 5mm in width, so the area for the diaphgram to move is 4cm in diameter. The diaphgram is cheap wrapping plastic (of the Elmu variety for those in Finland) - it's thickness is stated as 10µm on the box, material is LD-polyetylene.
I glued all the parts together with superthin soldering strips lodged in the right places for the electric contacts. Then I applied a bit of graphite powder on the diaphgram with cotton pieces. I measured the surface resistance to be in the order of 22 to 42 kilo-ohms per centimeter.
After the glue had dried, I soldered all the wires, put the drivers in their case (full sized headphones previously closed, now drilled open, with relatively tight ear cushions). For driving the ESLs I have two Hammond 125D push pull transformers and a high voltage source. Signal comes from my 1W Darling amp with approximately 250v signal swing, 5k / 4R OT.
I also put a lot of cotton on the backs of the drivers.
I thought I was good to go, and started experimenting. On the first try I got some sound, but it was very very weak even with the source and Darling amps volumes at max. Barely able to hear anything. At this point the bias was 405v and transformers at 4R / 27k. So I tried progressively bigger transformer ratios, and got a tiny bit more sound. Then I made a new voltage source, with 742v bias and transformers at 1.5R / 27k.
With this I got an ok listening volume, maybe 3/4 of what I usually use (I don't listen super loud), but even now I have Darling volume at 100% and source volume at 100%. What is more puzzling is that I have absolutely no bass. None at all. When comparing with my AKG601 headphones and with EQ (my source is a computer) I found I have no action at all pretty much below 700Hz. The response starts dropping around 1 - 2 kHz.
The sound is quite good on the highs - excellent imaging and presence. I find ESLs to be quite promising. However my first experiment is clearly not done right.
So, no volume or bass. Where did I go wrong?
Is the diaphgram area too small, with only 4cm diameter?
Did I use too much graphite - I think I did, but could this contribute to these problems?
Is the LD-polyetylene completely unsuited for ESLs? Could I have used too much tension?
I'm pretty much at a loss here for explanations. Any help would be appreciated.
Hi,
there are several mechanisms that spoil effeiciency with your HP-panel.
5cm diameter is a quite small area, of which just 4cm are effectively usable.
This means that only 64% of the HPs capacitance can be used to generate sound. The rest is wasted in the ´dead´ capacitance formed by area the spacers cover. Since the dielectric constant of the spacer material is larger than 1 (probabely betwen 3 and 6) the ´dead´ capacitance needs to be multiplied with the value of the dielectric constant compared to air as dielectric. This means that in this case the wasted or ´dead´ capacitance is even larger than the effective capacitance. This capacitance needs to be driven by the signal voltage too. The spacers should cover as few as possible of the active area of the panel. A solution could be to have a spacer ring of larger diameter, so that just a thin ring of the stator rim rests on the spacer ring.
1mm of spacer thickness is far too much. Think of 0.5mm or even less.
Thinner spacers increase the capacitance of the panel. Its impedance comes down, hence the transformation factor of the Audio transformer may be lower.
If there´s a mismatch between primary and secondary impedances, the driving amp can´t deliver its specified power. In this case the impedance the amplifier sees, is probabely much higher than 8Ohms.
Don´t use graphite as coating. Its usually quite low in impedance value and it shows a strong tendency to get everywhere where it shouldn´t and form leakage paths. With the impedance values You measured there´s a great chance that the Bias supply is overloaded and doesn´t deliver its specified voltage. A 1mm air gap may survive voltages of up to 2kV in theory, and 1.5kV in praxis (depending very much on the mechanical tension of the membrane).
The problem of no bass is probabely due to the wrong transformer parameters. The lower bandwidth limit is a direct function of the trannies primary inductance. The higher the inductance value the lower the botteom end frequency.
Mechanical tension of the membrane affects bass response too, but for a HP (and only for a HP where the diaphragm is in close proximity to the ear) a resonance of ~80-100Hz still provides for enough bass. This is usually achievable by simple thermal treatment of the membrane, which makes mounting of the HP easier.
jauu
Calvin
there are several mechanisms that spoil effeiciency with your HP-panel.
5cm diameter is a quite small area, of which just 4cm are effectively usable.
This means that only 64% of the HPs capacitance can be used to generate sound. The rest is wasted in the ´dead´ capacitance formed by area the spacers cover. Since the dielectric constant of the spacer material is larger than 1 (probabely betwen 3 and 6) the ´dead´ capacitance needs to be multiplied with the value of the dielectric constant compared to air as dielectric. This means that in this case the wasted or ´dead´ capacitance is even larger than the effective capacitance. This capacitance needs to be driven by the signal voltage too. The spacers should cover as few as possible of the active area of the panel. A solution could be to have a spacer ring of larger diameter, so that just a thin ring of the stator rim rests on the spacer ring.
1mm of spacer thickness is far too much. Think of 0.5mm or even less.
Thinner spacers increase the capacitance of the panel. Its impedance comes down, hence the transformation factor of the Audio transformer may be lower.
If there´s a mismatch between primary and secondary impedances, the driving amp can´t deliver its specified power. In this case the impedance the amplifier sees, is probabely much higher than 8Ohms.
Don´t use graphite as coating. Its usually quite low in impedance value and it shows a strong tendency to get everywhere where it shouldn´t and form leakage paths. With the impedance values You measured there´s a great chance that the Bias supply is overloaded and doesn´t deliver its specified voltage. A 1mm air gap may survive voltages of up to 2kV in theory, and 1.5kV in praxis (depending very much on the mechanical tension of the membrane).
The problem of no bass is probabely due to the wrong transformer parameters. The lower bandwidth limit is a direct function of the trannies primary inductance. The higher the inductance value the lower the botteom end frequency.
Mechanical tension of the membrane affects bass response too, but for a HP (and only for a HP where the diaphragm is in close proximity to the ear) a resonance of ~80-100Hz still provides for enough bass. This is usually achievable by simple thermal treatment of the membrane, which makes mounting of the HP easier.
jauu
Calvin
It sounds like too much graphite and the diaphragm size is way too small. I tried something like that before and got zero bass. Also, a tight coupling to your ears is essential for tight controlled bass.
Here's some additional info from my site:
DIY Electrostatic Headphones - Wen stuff
Note I don't recommend graphite at all anymore. Too many issues. Go with what the ES speaker guys use.
Good Luck!
Here's some additional info from my site:
DIY Electrostatic Headphones - Wen stuff
Note I don't recommend graphite at all anymore. Too many issues. Go with what the ES speaker guys use.
Good Luck!
Hi MrCurwen,
I think you can do a few improvements to your design.
1. Use a thinner spacer. I think 0.5 mm is more suitable for you.
2. Increase the bias voltage to around 600V.
3. Don't use graphite powder. Even though it works, it takes a lot of effort to really rub the graphite onto your diaphragm. I find a permanent antistatic solution much easier to apply. A very tiny drop is all you need. It will give you a resistance of around 1-100 Mega ohm per square. I use Staticide 6300. It works wonderfully.
Wachara C.
I think you can do a few improvements to your design.
1. Use a thinner spacer. I think 0.5 mm is more suitable for you.
2. Increase the bias voltage to around 600V.
3. Don't use graphite powder. Even though it works, it takes a lot of effort to really rub the graphite onto your diaphragm. I find a permanent antistatic solution much easier to apply. A very tiny drop is all you need. It will give you a resistance of around 1-100 Mega ohm per square. I use Staticide 6300. It works wonderfully.
Wachara C.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.