Page 1 of 1

Gain block question

Posted: Sat May 11, 2019 8:34 am
by CeaSaR
Here I go again, another wierd discrete component audio query.

Still in the audio reproduction mode (it's my main electronic interest ;) ) and this time I'm trying to understand how the "volume control" on the PSAudio amps work. Paul (the P in PSAudio) says that they use gain blocks for their volume controls so that they have a constant impedance presented to the source, thereby not altering the source signal integrity usually caused by impedance mismatch.

I've put together a few sims using the most ubiquitous popcorn bjt, the 2n3904.

The first uses variable base voltage, which in turn produces variable low base current. There's a very miniscule range where the signal goes from a few nanovolts to full input. Setup is source into the collector, emitter to resistor that leads to gnd, and a variable resistor from +V to gnd, wiper to base.

The second uses a variable resistor in 1/2 of the gain resistor string to change the gain. Always amplifies somewhat from a few mv to way past the point of distortion. Setup is a classic Class A VAS with a same value VR paralleling the collector resistor, wiper tied to the collector end of the pair. Then I take the resultant signal off either the collector or emitter (inverted or non-inverted, respectively) into a resistor tied to gnd.

Both times I am reading the results from between the transistor and the load resistor.

Obviously I am having trouble getting it to work. AM I barking up the wrong tree with a bjt and I should use a MOSFET instead, or is there another way that I am missing.


(I believe I am looking for a voltage controlled amplifier with serious negative gain [0 signal passage] to unity gain or just slightly over unity. Probably will require MOSFET instead of bjt. Would still like to understand how to do with bjt as I enjoy a good challenge.:D )

Re: Gain block question

Posted: Mon May 13, 2019 3:38 pm
by haklesup
you'll never get High impedance using BJT and not good range if you are controlling the base with voltage (since it is a current controlled terminal, you need series resistance to turn that into a controllable current and it still will not be linear because a diode junction is not).

I think you will have better experience making a MOSFET amplifier Class AB type. The inputs can have very high impedance and are generally linear all the way out to the point you run out of power to drive. The new D class amplifier modules (cheap on ebay) are also very interesting