A voltage-output VAS adds clarity but is not essential. A voltage-output VAS with the gain of 1000 in series with a 100ohm resistor can be replaced by its Norton equivalent: a current-output VAS with the transconductance of 10, loaded by the same 100 ohm resistor. The nulling principle remains the same. If you replace E2 in your simulation with a voltage dependent current source G with the value of 10, you'll get a nice null, too.
BTW my post #58 above mistakenly shows that the left end of the 2-ohm resistor R3 of the nulling network is connected to the output of "OPS" and not to the output of "VAS"; sorry about that. With a voltage-output VAS it makes no difference for the results of the simulation - that's why I didn't notice it until it was too late to correct. It does make all the difference with a current-output VAS though.
BTW my post #58 above mistakenly shows that the left end of the 2-ohm resistor R3 of the nulling network is connected to the output of "OPS" and not to the output of "VAS"; sorry about that. With a voltage-output VAS it makes no difference for the results of the simulation - that's why I didn't notice it until it was too late to correct. It does make all the difference with a current-output VAS though.
Last edited:
Note: I am posting this as I have a related question on the Class-A output stage (plus two other questions). If this should be in a separate thread, kindly remove my post and let me know.
I have a few questions on the Quad 405 that I am struggling to understand and that I can't find answers to. I have built two Chinese Quad 405 clones (LJM), of a Quad 405-1 revision. I have implemented some of Keith Snook's and Bernd Ludwig's suggested modifications. Some of which were successful (e.g. the non-inverting Opamp configuration lowers the noise) others not so (the 'ring of two' class-A output stage with the feedback capacitor C8 returning to the emitter of Tr2 not being HF stable). I have limited means to measure absolute distortion figures (PicoScope 2000), so with some modifications, I have to rely on what others have tested and published. I am also looking at the subsequent Quad 306 and 606 to see if there are any other modifications that could be beneficial to the 405.
My questions are the following:
I have a few questions on the Quad 405 that I am struggling to understand and that I can't find answers to. I have built two Chinese Quad 405 clones (LJM), of a Quad 405-1 revision. I have implemented some of Keith Snook's and Bernd Ludwig's suggested modifications. Some of which were successful (e.g. the non-inverting Opamp configuration lowers the noise) others not so (the 'ring of two' class-A output stage with the feedback capacitor C8 returning to the emitter of Tr2 not being HF stable). I have limited means to measure absolute distortion figures (PicoScope 2000), so with some modifications, I have to rely on what others have tested and published. I am also looking at the subsequent Quad 306 and 606 to see if there are any other modifications that could be beneficial to the 405.
My questions are the following:
- Walker's papers show a current source for the class-A driver of 50 mA. In the 405 these are resistors R30/31 setting the current to 50 V / 2 x 0.56 k = 45 mA. The closest you can get to 50 mA with the E-12 series. Fair enough. In the 306, these resistors are kept at 560 Ohm, while the supply voltage is lowered to -38 V, lowering the current to 34 mA. I am planning to run my "Quad's 405" on + and -42 V. Is there any benefit in lowering R30/R31 to 470 Ohm, as this would bring the current back up to around 45 mA?
- Snook shows a base stopper resistor of 10 Ohm for upper dumper Tr9. Quad introduced this in the 606 where there are triple dumpers (R7 in the 606 schematic). I suppose to promote better stability. However, for a single transistor output, 10 Ohm seems much too high to me as a single TO-3 output transistor suffers from significant beta-droop at higher currents. Wouldn't 2R2 be more sensible, if at all necessary?
- In the 306/606 and further revisions, Quad introduced another diode in the dumper section (D7 in the 606, D8 in the 306). It anti-connects the base/emitter of the upper dumper(s) (Tr9/10/11). What is the purpose of this diode. Would there be any benefit in implementing this in a 405?
Attachments
The diode protects the emitter junction from excessive reverse voltage. The base-emitter reverse breakdown voltage is low, typically 5V, and can easily be exceeded e.g. in case of clipping or shorted output.
Thank you for your answer. Pardon my ignorance, but would you know why it was only added to the upper dumper, and not the lower one?
Last edited:
It was added between them, not to either one or the other of them
And I have no belief in the protection theory advanced here. The Quad documents state that it was added to ‘correct the response at 20kHz’.
@ejp you are right about the extra diode to reduce the dead zone (D15 in the 405-2 schematic). My question was on D6 in the 306 schematic. Since I have never seen such a flyback diode on an emitter/base junction, I think I will not implement this.
As for my first question, I will test the amplifier at both 50 V and 42 V, with and without altering the current source R30/31 for the 42 V supply, and try to check for differences in harmonic distortion.
For my second question, I will not go with a base stopper for the upper dumper unless I see HF instabilities, and then I will max. it at 2R2.
Thank-you all for your help.
As for my first question, I will test the amplifier at both 50 V and 42 V, with and without altering the current source R30/31 for the 42 V supply, and try to check for differences in harmonic distortion.
For my second question, I will not go with a base stopper for the upper dumper unless I see HF instabilities, and then I will max. it at 2R2.
Thank-you all for your help.
