If you have a esl panel with 2.8 nF capacitance between stators, and you drive this panel with a 1:175 step up transformer, then the minimum impedance that the amplifier will see is: (1/6.28x20000x0.0000000028) divided by (175x175) = 0.092 ohm!!
So if you will keep your amplifier in the comfort zone, the max step up must not be higher than 1:53 with your big panels....
Of course, You can put a resistance in series with the transformer for taming the high
resonance peak, but the value of this usally lies around 0.6-0.9 Ohm resulting in a good total minimum impedance of around 2 ohms when using 1:53 step up.
One more thing:There's no reason for building very big panels, as the acoustic impedance load on the membrane will be higher and dampen it's movements.
So if you will keep your amplifier in the comfort zone, the max step up must not be higher than 1:53 with your big panels....
Of course, You can put a resistance in series with the transformer for taming the high
resonance peak, but the value of this usally lies around 0.6-0.9 Ohm resulting in a good total minimum impedance of around 2 ohms when using 1:53 step up.
One more thing:There's no reason for building very big panels, as the acoustic impedance load on the membrane will be higher and dampen it's movements.
Hi, thanks for the clarification. Yes the impedance will be quite small, the driving amps are luckily quite big (24 power transistors, 120,000µF, 1,5kV trafo and two massive convection elements per mono).
Big stats sound quite good
. A line source is ideal when it's from floor to the ceiling. Also the width of the panel has to be something, if one wants to reproduce lower registers. Bass actually sounds very good from a panel, mine go to ~37Hz in my room if driven full range but I will limit the low end response because of the transformers, and handle the middle and low bass with dynamic woofers in dipole line arrays.
Very Good,Legis, it sound just about right I don't for see an issues if the primary is bifilar.
But if it is all possible see if you can have them double up on the insulation layer to the secondary base as this will help tremendously,or as much as they can.
If they use 2 mil material then 10 layers should be good, as an example.
AndrewT, the thicker the insulation the less parasitic capacitance the winding will have.
Even though the core is not connected to anything it still acts like one of the plates of a capacitor and couples each individual turn to form the total residual capacitance of the winding.
I wanted to undo one of my cores and add a thicker layer of insulation to prove this,But I am just not up to dealing with trying to straighten out that much wire and then rewind it again.
At least until I can devise a simple torrid winding hoop of some sorts.
jer
But if it is all possible see if you can have them double up on the insulation layer to the secondary base as this will help tremendously,or as much as they can.
If they use 2 mil material then 10 layers should be good, as an example.
AndrewT, the thicker the insulation the less parasitic capacitance the winding will have.
Even though the core is not connected to anything it still acts like one of the plates of a capacitor and couples each individual turn to form the total residual capacitance of the winding.
I wanted to undo one of my cores and add a thicker layer of insulation to prove this,But I am just not up to dealing with trying to straighten out that much wire and then rewind it again.
At least until I can devise a simple torrid winding hoop of some sorts.
jer
This maybe redundant, But here is a core that I have been considering on using,
Toroidal Core PN# 140
the also have some bigger ones as well !!!
But very affordable.
jer
Toroidal Core PN# 140
the also have some bigger ones as well !!!
But very affordable.
jer
Yes, big panels certainly sounds good, but there seems to be an optimum panel widht for high output at low freqs, 8-10 inch for a single membrane, two stator panel and
12-13 inch for a double membrane, three stator panel.
This is my own findings from a long life of building electrostats (and still counting).....
12-13 inch for a double membrane, three stator panel.
This is my own findings from a long life of building electrostats (and still counting).....
If it were me, I would consider using using 1 or even 3 of the center sections separated from the two end sections as just for the highs.
This would help greatly to reduce the load on the amp.
My little panels are just 3.25" wide and I get at least 45 to 60 degrees of horizontal dispersion for the highest frequency's.
I believe those large panels are already physically sectioned as such by the photos I have seen,so my comment may be irrelevant.
If the stator sections are physically separated from each other, This could easily be done with just a handful of resistors using a segmentation equalizing technique.
This would help with any lobing patterns that you might hear while walking around the room as well.
One of my next builds will be like that as I understand it works very well !!
I sure would like to have a listen to them !!!
Jer
This would help greatly to reduce the load on the amp.
My little panels are just 3.25" wide and I get at least 45 to 60 degrees of horizontal dispersion for the highest frequency's.
I believe those large panels are already physically sectioned as such by the photos I have seen,so my comment may be irrelevant.
If the stator sections are physically separated from each other, This could easily be done with just a handful of resistors using a segmentation equalizing technique.
This would help with any lobing patterns that you might hear while walking around the room as well.
One of my next builds will be like that as I understand it works very well !!
I sure would like to have a listen to them !!!
Jer
Last edited:
Can I use regular toroidal transformer, with 1:8 step up ratio to drive basic dynamic tweeter load (16ohm compression driver + 200R series resistor). Is the bandwidth enough to reach ultra frequencies without much phase shift? Lets say the primary inductance is for example 166mH. If that kind of trafo is a low pass filter with that load, which kind of transformer should it be? Bifilar wound pri and sec, minimal primary turns...? Should I use a electrostatic shield or not between the pri and sec?
The reason if anyone is wondering, is that I will be current driving the compression drivers, the step up trafo is just to boost the output level back to almost normal. Just a side hobby besides electrostatic speakers.
Thanks!
The reason if anyone is wondering, is that I will be current driving the compression drivers, the step up trafo is just to boost the output level back to almost normal. Just a side hobby besides electrostatic speakers
.Thanks!
Last edited:
Actually when I measure the two primaries in series, the inductance reads 166mH, but when the primaries are in parallel the inductance is vanishingly low 0,03mH. Secondary is unshorted in both cases. I quess this kind of primary inductance gives wide bandwidth?
The trafo has 2 x 50 primary turns on a 100VA core (~15-18cm2). If I remember correctly, the primaries are bifilarly wound.
The trafo has 2 x 50 primary turns on a 100VA core (~15-18cm2). If I remember correctly, the primaries are bifilarly wound.
Last edited:
Should be doable to keep phase < 30deg at 20Khz.
How low in frequency will the transformer have to operate? and what Vrms will it need to handle?
The current drive is approximated by the use of the 200R series resistance?
Any chance you need to swap polarity on one of the primaries when wired in parallel?
Hi Steve, thank for answering. I will be using 1st order line level low pass between 300 and 600Hz. 11Vrms is the maximum continuous drive voltage, the power dissipation in the two 100R resistors gets too high after that. They can handle 2-3x the power at short peaks, but the SPL at 11V is quite high already (it translates to ~8Vrms across the 110dB/2,83V compression driver I think).
Yes the series resistor acts as the main "V/I converter" so the non-linear V/I conversion of the voice coil gets smaller. The amount that the nonlinearities (caused by electric motor forces) are reduced at maximum, can be estimated by (200R + 16R [+ secondary R]) / 16R = ~13,5 times = ~22,6dB. I did not take the R of the secondary in the account, but it should be pretty low.
I connected the center wire of pri A with the side wire of pri B. If I connect the center wires of both pri A and B, the inductance goes to 0,01mH. And actually the measuring frequency is ~65kHz, this ebay meter changes it's frequency depending on the load.
What kind of design would be best? 50 turn on that size core is quite oversized regarding saturation. So maybe 25 turn pri? Bifilar would pri and sec (at least pri)? How about the electrostatic shield, good or not? Should I connect it to the chassis of the driving amplifier for fully balanced operation, or leave it floating?
Last edited:
I think that when you are measuring the paralleled inductance, you must change the polarity on one of the coils so that you get a correct reading of 166/2 = 83 mH.
Driving a dynamic speaker with a 8 times stepup trafo is not an good idea due to
the enormous impedance mismatch...it will give 16/(8*8) impedance = 0.25 ohm......
Driving a dynamic speaker with a 8 times stepup trafo is not an good idea due to
the enormous impedance mismatch...it will give 16/(8*8) impedance = 0.25 ohm......
Hi, each individual primary measures 0,04mH when measured separately. For some reason both primaries in series show much higher inductance. If I connect them in series "the wrong way", then the inductance also reads below 0,1mH. Could it be from the bifilar wounding that's causing this... Maybe the "correct" series connection negates the bifilar wounding's inductance cancelling effect.
Like was mentioned, there will be 200R resistor (or more likely 2 x 100R) in series with the 16ohm driver, so the load that the amp sees will be closer to ~3,38ohms.
Last edited:
If you don't have any inductance in a coil then you don't have a coil...
How are yoy measuring the inductance?
I cant see the reason for first stepping up the voltage 8 times, then reducing it 200/16=12.5 times with a resistor? All that you will achieve is a reduced sensitivity
and a impedance at 3.4 Ohm instead of 16.....
How are yoy measuring the inductance?
I cant see the reason for first stepping up the voltage 8 times, then reducing it 200/16=12.5 times with a resistor? All that you will achieve is a reduced sensitivity
and a impedance at 3.4 Ohm instead of 16.....
I use this kind of meter: http://www.ebay.com/itm/L-C-F-Induc...269?pt=LH_DefaultDomain_0&hash=item3f0e13fb9d
You can read more about the current driving here: Current-Drive - The Natural Way of Loudspeaker Operation
I have used 33R over a month in series with the same drivers and have been liking what I hear. I want to use more current drive, and still get good sensitivity, that's why the bigger series res and the step-up trafo.
Last edited:
The transformers' specifications that I ordered:
- Core's area ~11,5cm2, material is grain oriented silicon steel, Bmin = 1.7T
- 2 x 30 turn bifilar wound primaries (2,0mm2 wire)
- 2 x 120 turn bifilar wound secondaries (0,44mm2 wire)
- Primay is wound close to the core, secondary on top.
- Electrostatic shield
- Electromagnetic shield
What do you think, will it have high enough bandwidth when operated at 1:8 and 200R series resistance prior the driver? There is also possibility to use just ~100R series resistance and connect the trafo at 1:4 step-up by connecting the secondaries in parallel.
- Core's area ~11,5cm2, material is grain oriented silicon steel, Bmin = 1.7T
- 2 x 30 turn bifilar wound primaries (2,0mm2 wire)
- 2 x 120 turn bifilar wound secondaries (0,44mm2 wire)
- Primay is wound close to the core, secondary on top.
- Electrostatic shield
- Electromagnetic shield
What do you think, will it have high enough bandwidth when operated at 1:8 and 200R series resistance prior the driver? There is also possibility to use just ~100R series resistance and connect the trafo at 1:4 step-up by connecting the secondaries in parallel.
Last edited:
Based on measurement of other small toroidal power transformers, if transformer is wound with layers taking up full circumerence of toroidal core, I think it will. The only thing I am not sure about is how much space the electrostatic shield takes up, as I have not yet measured any with it. Any added space between primary and secondary increases leakage inductance. Primary of 30 turns should provide plenty of core capability for the frequency range you plan to drive it in.
Concerning measurement of primary inductance:
My guess is that your meter is trying to determine inductance at too high a frequency and is getting confused by other parasitics like winding capacitance. You need to have a meter you can set the measurement frequency to something low(< 200 Hz) where the primary inductance dominates the impedance.
Attached below is a simple test setup to calculate primary inductance based on voltage measurement across a series resistor(VR) and across the primary(VL). I'd recommend input voltage of 1Vrms @ 60Hz using series resistor R = 10 ohm.
Attachments
<- 2 x 120 turn bifilar wound secondaries (0,44mm2 wire)>
Never do bifilar winding for HV windings, doing so you create too much capacitance. Look at Tesla's patent Patent US512340 - NIKOLA TESLA - Google Patents
Never do bifilar winding for HV windings, doing so you create too much capacitance. Look at Tesla's patent Patent US512340 - NIKOLA TESLA - Google Patents
What is the negative side effect of extra capacitance in this application (dynamic driver and non-inductive resistor as load)? 120 turns is still quite low, ESL output transformer with several thousand turns would be a different story...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.