Such as Circlophone servo?
Perhaps this is good moment to ask - is there a post somewhere that explains principles of how Circlophone controls Iq? I've been unable to find it..
Elvee, perhaps you can explain as briefly as possible, how your servo works,
or direct me to the existing post?
Edit: I'm assuming that even Iq servo will need some kind of 'emitter resistors' (Circlophone has Shottky diodes in parallel with resistors)
to measure current, right? Se we can't get rid of them completely.
Last edited:
It is a possible example among others, maybe not the best
IIRC, I have given such a detailed description, it must hide somewhere in the hundreds of posts. I'll try to locate it
Yes, some means of current sensing remains necessary; one could use Hall sensors to avoid resistors completely, but it is not necessary to go that far: you are talking of emitter resistors, but resistors can be located elsewhere, in the collectors (or the equivalent in a composite) for example, where they almost don't interact with the signal, output impedance, X-over characteristics, etc
Another OPS with no emitter resistors
Hi minek123,
LKA (Ladislav) has started a thread for a bipolar current driven output stage here Low distortion current driven Class-B output stage
It can be safely used without emitter resistors including paralleling (see post 4). It does not need current sensing for the idle current and idle current is very small qualifying for "Class-B" (as Douglas Self defines it). It does still need Vbe temperature sensing to keep the idle current low when the heatsink gets hot.
It would be nice if idle current (bias voltage) could be closer to nothing (like the original Quad 405 current dumper) then there's no need for the Vbe temperature sensing ... but I found that if the idle current goes much below 1mA (about 400mV bias voltage at room temperature) then unacceptable levels of high order harmonics mess up the sound quality, and high order harmonics are hard to suppress with negative feedback.
Interestingly, in this LKA circuit I noticed some feedforward current flow from the input stage to the output (plot current through V7 to see it). Current flows if the output transistors gain falls below normal (approximately unity) such as in the crossover region. Maybe this circuit could be modified to use Quad current dumping? Notice the missing current due to delayed turn on of one power transistor is only around 20mA which is within the range of an input stage. Any thoughts anyone?
Hi minek123,
LKA (Ladislav) has started a thread for a bipolar current driven output stage here Low distortion current driven Class-B output stage
It can be safely used without emitter resistors including paralleling (see post 4). It does not need current sensing for the idle current and idle current is very small qualifying for "Class-B" (as Douglas Self defines it). It does still need Vbe temperature sensing to keep the idle current low when the heatsink gets hot.
It would be nice if idle current (bias voltage) could be closer to nothing (like the original Quad 405 current dumper) then there's no need for the Vbe temperature sensing ... but I found that if the idle current goes much below 1mA (about 400mV bias voltage at room temperature) then unacceptable levels of high order harmonics mess up the sound quality, and high order harmonics are hard to suppress with negative feedback.
Interestingly, in this LKA circuit I noticed some feedforward current flow from the input stage to the output (plot current through V7 to see it). Current flows if the output transistors gain falls below normal (approximately unity) such as in the crossover region. Maybe this circuit could be modified to use Quad current dumping? Notice the missing current due to delayed turn on of one power transistor is only around 20mA which is within the range of an input stage. Any thoughts anyone?