about what scheme?
R27 why?
R19 is linear, it may not be enough for the thermal stability of the output stage.
so far, in your version of the VT14 VT15 circuits are of little use.
The presence of C11 in the circuit indicates that phase lag compensation is needed due to the input capacitance of the input stage, which also reduces the speed in the NFB loop.
Last edited:
this is addressed by spreading the gain more evenly between the two gain stages. Your main amplification is done by the second cascade, to which a progressive correction is attached, which slows down the speed in the NFB loop.
modeled this circuit.
What input amplitude voltage did you expect the circuit to take, having a gain of 31.7 dB with a supply voltage of only +/-40 volts?
according to the AX-20 topology, I couldn’t get less than 210ns ...
Without deep modification, I think it will not work out less?
high-speed distortion can be reduced by increasing the initial current of the output transistors to 180mA per pair, and to expand the bandwidth, you can reduce the gain of the input stage by increasing the resistors in the emitters to 68 ohms (R5 R9 - see attach).
And slightly modified the clumsy topology with the use of a zener diode for other purposes - now the bias is stable during fluctuations both in positive and negative supply voltages.
Attachments
considered well? ok... what problems do these settings? if we take the Self amplifier as an example, then such parameters are achieved with a total current of the input differential stage of 20 mA, and which corresponds to an output power of 141 watts at a load of 4 ohms at a frequency of 468 kHz.
No replacement of the zener diode occurs, it is simply removed from the place where its main parameters do not correspond to its datasheets. I have seen this mess many times, so I suggested a simple solution. And it shouldn't affect the slew rate value.
P.S. you incorrectly attached my name AX20 to the file name, because I did not post the full version of modification. And I certainly would not overclock the gain over 30 dB, having a significant non-linearity of the frequency response to cover NFB the open loop. And not a very relatively high supply voltage.
However APEX schemes are classically correct and can be modified to very good basic characteristics!!
Gennady, give at least some parameters of the modification you proposed
Otherwise, the proposal looks like fortune telling by coffee grounds
Petya, (probably that's your name? My name begins with the letter H...), this modification does not have super-important parameters, for example, super low delay time and ultra-high signal slew rate, tBy matching the gain between the stages, the non-linearity of the frequency response is made smaller until the coverage of negative feedback. Due to this, the depth of negative feedback at a frequency of 20 kHz was increased and it became possible to use a circuit with an input amplitude voltage of up to 2.5 volts, as was written earlier....
As for fortune telling, I don’t know, everything is taken from an authentic sample. Judging by the type of transient response of which, there are no outliers, as in where you so famously show on the forum ...
Such modifications are based purely on the personal practical experience of listening to many modified amplifiers with various topologies, and such a correction scheme is chosen as the main one, Because this equalization results in a more open sound that many audiophiles love, reminiscent of the comfortable sound of a tube amplifier.
As for fortune telling, I don’t know, everything is taken from an authentic sample. Judging by the type of transient response of which, there are no outliers, as in where you so famously show on the forum ...
Such modifications are based purely on the personal practical experience of listening to many modified amplifiers with various topologies, and such a correction scheme is chosen as the main one, Because this equalization results in a more open sound that many audiophiles love, reminiscent of the comfortable sound of a tube amplifier.
Last edited:
P.S. there's an error here. offer option with a maximum input amplitude voltage of 1.8 volts.
OK.
Absolutely not, because in the case of a progressive correction of an important task, is the linear rates of the frequency response in the sound frequency range,it, or name differently a high pole - this very affects the comfort of sound in practice. Such a correction has specialty in the fact that the phase reversal is necessary only in the area of single amplification and this does not affect the sound.
In the case of an amplifier Cordell correction is carried out in a feedback loop - this can be seen from the linearity of the phase of the open loop.
the phase doesn't cross the 180-degree turn before the unity gain frequency, so don't worry.
