some questions about electrostataic amp

[2b青年爱发烧]

planning to build an electrostataic amp for my SRM L700.

I have built several amps for my Moving coil earphones (HD650, K601, K702...), but never built an electrostataic one, I read some topics about it, also some diagrams of amps , there are still some questions need your help.

1. Never saw the full differential structure (for example, opa1632, lme49724) used in the electrostataic amp, why?

2. The electrostataic earspeakers have much larger impedance than moving coil ones, the only capacitive impedance is isolated by the output resistor (usually 5.1k), so...the electrostataic amp works like without loads?

3. Why the output resistor is usually 5.1k ?

4. The bias terminal also has a resistor, maybe it will affect the sounds ?

5. Is the emitter follower necessary ? I noticed that the stax lower price  amps (323, 353x) possess an emitter follower, it was said that "has enabled wide-range reproduction resulting in both the extended dynamic range at high frequency and the improved tone quality"....but the higher price products (717, 727, and other tube amps) DO NOT use emitter follower, so... I'm puzzled.

 

Temporarily above, many thanks~~~

[kevin gilmore]

1) go and find complementary parts that work at 900v that are not obsolete

2) the load is more than you think. up to 20ma at 20khz

3) protect the diaphram

4) takes minutes to charge, quality of resistor is meaningless

5) 717 and 727 definitely have emitter follower output stages with current sinks.

Edited February 24, 2019 by kevin gilmore

[2b青年爱发烧]

21 hours ago, kevin gilmore said:

1) go and find complementary parts that work at 900v that are not obsolete

2) the load is more than you think. up to 20ma at 20khz

3) protect the diaphram

4) takes minutes to charge, quality of resistor is meaningless

5) 717 and 727 definitely have emitter follower output stages with current sinks.

Thanks very much for your reply~

1) I checked out mouser and other sale sites, really there was no more high voltage P type devices, sad...but found a few 2sa1968 on taobao.com, maybe I can use them.

3) what if short the resistor? yeah...maybe it's dangerous.

5) sorry I remembered it wrong, is it better with a follower?

[joehpj]

1 hour ago, 2b青年爱发烧 said:

Thanks very much for your reply~

1) I checked out mouser and other sale sites, really there was no more high voltage P type devices, sad...but found a few 2sa1968 on taobao.com, maybe I can use them.

3) what if short the resistor? yeah...maybe it's dangerous.

5) sorry I remembered it wrong, is it better with a follower?

use modern SMD ones like STN9360 instead of 2sa1968 on taobao.

[Pars]

2SA1968->taobao = likely fake


Sent from my iPhone using Tapatalk

[2b青年爱发烧]

1 hour ago, joehpj said:

use modern SMD ones like STN9360 instead of 2sa1968 on taobao.

thanx, I'll check it out

56 minutes ago, Pars said:

2SA1968->taobao = likely fake


Sent from my iPhone using Tapatalk

yes there are quite many fake ones, have to pick carefully.

[spritzer]

2SA1968 now will pretty much mean fake.  I doubt many real ones are still in circulation as we have been dealing with fakes for about a decade now. 

[2b青年爱发烧]

Since the KGSSHV Carbon is famous as a reference of electrostatic amp, I think it would be better to do some spice simulations to check the baseline of its circuit.

Many components were not found in LTSPICE so I used similar ones instead. It works except the quiescent current at the output stage is large. (Anyone could please tell me what it is ?)

Now I meet a problem, that is, how to use the fourier transfer directive (.four) on the voltage between two output ? I tried many way but all failed:

Fourier components of V(out+)
DC component:-3.16154

Harmonic    Frequency     Fourier     Normalized     Phase      Normalized
 Number       [Hz]       Component     Component    [degree]    Phase [deg]
    1       1.000e+03    9.355e+01    1.000e+00       -0.93°        0.00°
    2       2.000e+03    2.356e-02    2.518e-04      -94.27°      -93.34°
    3       3.000e+03    2.172e-04    2.321e-06      -21.46°      -20.53°
    4       4.000e+03    5.108e-05    5.460e-07      128.27°      129.21°
    5       5.000e+03    3.615e-06    3.864e-08     -154.01°     -153.07°
    6       6.000e+03    5.703e-07    6.096e-09      -51.12°      -50.18°
    7       7.000e+03    1.475e-07    1.577e-09        7.65°        8.58°
    8       8.000e+03    1.874e-07    2.003e-09       16.11°       17.04°
    9       9.000e+03    8.659e-07    9.256e-09      164.63°      165.56°
Total Harmonic Distortion: 0.025185%

.fourier quantity "V(v(out+)-v(out-))" not pressent in data.

Thank you.

Yes I can check it in the FFT draw, but still want a text report.

[2b青年爱发烧]

find the solution, just input V(out+,out-) will be OK.

 

Fourier components of V(out+,out-)
DC component:-6.40787e-008

Harmonic    Frequency     Fourier     Normalized     Phase      Normalized
 Number       [Hz]       Component     Component    [degree]    Phase [deg]
    1       1.000e+03    1.441e+02    1.000e+00       -0.93°        0.00°
    2       2.000e+03    5.728e-08    3.976e-10       -3.00°       -2.06°
    3       3.000e+03    1.966e-04    1.365e-06      -22.41°      -21.48°
    4       4.000e+03    1.111e-07    7.712e-10       -0.34°        0.60°
    5       5.000e+03    2.646e-06    1.837e-08     -164.58°     -163.65°
    6       6.000e+03    1.705e-07    1.183e-09        2.85°        3.79°
    7       7.000e+03    2.017e-07    1.400e-09       20.00°       20.93°
    8       8.000e+03    2.243e-07    1.557e-09        1.29°        2.22°
    9       9.000e+03    1.429e-06    9.920e-09      163.50°      164.43°
Total Harmonic Distortion: 0.000136%

 

the THD is fairly good