Audio Signal Path - I'm confused

[Wil_Bloodworth]

I'm trying to understand the Main Amp / Power Output schematic of the Pioneer SX-990 and am confused as to how the audio signal travels through the circuit.

Below is the left channel. In BLUE, I've highlighted the route "I" think the audio signal path is taking to get to the output.  The RED and PINK lines have me baffled. YELLOW and GREEN are B+ and Ground/Return respectively.

NOTES:  At the base of the second transistor (2SC497), there is a 30K variable resistor (VR1).  I "think" this is the main bias adjustment. Attached to the collector of the same transistor is another variable resistor (VR3) which is 50 Ohms and I think this is used to set the idle current.

What have I assumed incorrectly?

What is the purpose of the PINK tracks?

How does the entire RED section work with the two bias pots and the stavistor?



Thank you!

- Wil

[Benta]

Somehow your color perception is different to mine.
I see a gray trace and a mustard-colored one. Nothing in pink or red. Sorry.

[Manul]

Somehow your color perception is different to mine.
I see a gray trace and a mustard-colored one. Nothing in pink or red. Sorry.

Welcome to the club (from someone who is also partialy colorblind).

[3roomlab]

i tried to look up stabistor
it is like unobtanium 

[IanB]

Somehow your color perception is different to mine.
I see a gray trace and a mustard-colored one. Nothing in pink or red. Sorry.

What colors can you tell apart (and also see as different from gray)?

[fourfathom]

I can't see the red, it looks gray, mostly.  I've got one of the typical red/green deficiencies.

About the circuit, the bottom blue path is high-frequency negative feedback, for stabilization.
That stabistor (haven't heard that term in a *long* time!) looks like essentially three diode-drops, so that plus the series resistor establishes the class-B bias current of that four-transistor output section.  The upper pair is a Darlington configuration, and the lower is a Sziklai pair.  These have different gains, but the negative feedback into the base of the second transistor linearizes the effect of that.  The Szilaki is used since I suppose they couldn't find a decent PNP complement to the upper output transistor.
The VR1 bias pot sets the DC level of the output, typically at half-supply voltage.

[Jeff eelcr]

VR1 is offset, VR3 is bias (idle current).
Jeff

[boB]

One thing important that I have learned over the years from others that really matters...

Do NOT use 4 way ties in a schematic.  Up to 3 only.

boB

[fourfathom]

One thing important that I have learned over the years from others that really matters...

Do NOT use 4 way ties in a schematic.  Up to 3 only.

boB

Agree strongly -- but is there such a thing as a 2-way tie?  Anyway, I was guessing a bit when I reviewed this schematic: "Is that a connection?  No, probably just a fat cross..."

[Calvin]

Hi,

since this amp is a single-supply design, the bias voltages of Q1(SC2372) and Q3(SC497) are generated by dedicated voltage dividers R3(2M2)/R3(100k) and R??(39k/VR1(33k)/R17(39k).
The supply voltage to the input stage is reduced and filtered by R7(18k)/C3(100µ/50V) ... which is advantageous or may even be neccessary due to the lowish PSRR of a singleended input stage.
C1(33µ/10V) keeps the amp´s input dc-free, resp. blocks dc from the biasing network of Q1.
C5(3µ3/25V) serves the same purpose to Q3.
R15 is the collector load for the transconductance input stage Q1, defining it´s gain.
Q3 is the VAS-stage (voltage amplification stage) with a 100pF Millercap, defining the openloop bandwidth of the amp.
Since the output stage is of pushpull type, D1(STV-3) together with VR3(50R) generate the required bias voltage to set the idle current for the output stage transistors.
D1 as a diode chain features a negative tempco, ideally countering the transistors positive tempco.
The aim is to keep the idle current constant over temperature variations of the output transistors.
As such D1 should be mounted with close thermal contact ... either on the same heatsink or directly on one of the transistors.
Q5/Q9 form a classical Darlington power transistor, Q7/Q11 form a complementary Darlington (Sziklai).
C7(100µ/50V) bootstraps the output voltage, keeping the voltage over it´s pins almost constant and as such the biasing of the output transistors.
VR1 sets the output voltage level to roughly 1/2 of the power supply voltage.
When trimming the amp VR1 should be tuned so that the output signal clips symmetrically.
R35(15K)/R11(150R) form a dc feedback loop defining the closedloop dc gain.
For the ac gain R39(8k2)||C15(47p) and C13(1000µ/35V) add to the feedback equation.
C13 acts as a dc-blocking cap towards the speaker connector.
The ac-feedback path could also bleed away C13´s charge at switch-off (if the speakers are connected through a -then open- protection relay).
I´m not exactly shure about C19(100p) bridging Q1´s BE-junction.
It might either bootstrap to increase the input impedance at HF, and/or forms a soft HF-emv-filter in cooperation with R1(2k2)
The RC output filter (10R/22nF) is a classical Zobel, providing for a defined HF-load to the amp.

This kind of amp topology was quite common in the ´early years´ up to the mid-80s.
Currently I´m servicing a Marantz 2230 Receiver (1975 iIrc) that features almost identical amps.

regards
Calvin

[EPAIII]

And that is exactly why bridges and connecting dots should both be ALWAYS used. And NEVER, EVER just two lines crossing without one or the other.

That's my opinion and I AM sticking to it. It prevents ALL chances of confusion.

One thing important that I have learned over the years from others that really matters...

Do NOT use 4 way ties in a schematic.  Up to 3 only.

boB

Agree strongly -- but is there such a thing as a 2-way tie?  Anyway, I was guessing a bit when I reviewed this schematic: "Is that a connection?  No, probably just a fat cross..."

[Wil_Bloodworth]

Hi,

since this amp is a single-supply design, the bias voltages of Q1(SC2372) and Q3(SC497) are generated by dedicated voltage dividers R3(2M2)/R3(100k) and R??(39k/VR1(33k)/R17(39k).
The supply voltage to the input stage is reduced and filtered by R7(18k)/C3(100µ/50V) ... which is advantageous or may even be neccessary due to the lowish PSRR of a singleended input stage.
C1(33µ/10V) keeps the amp´s input dc-free, resp. blocks dc from the biasing network of Q1.
C5(3µ3/25V) serves the same purpose to Q3.
R15 is the collector load for the transconductance input stage Q1, defining it´s gain.
Q3 is the VAS-stage (voltage amplification stage) with a 100pF Millercap, defining the openloop bandwidth of the amp.
Since the output stage is of pushpull type, D1(STV-3) together with VR3(50R) generate the required bias voltage to set the idle current for the output stage transistors.
D1 as a diode chain features a negative tempco, ideally countering the transistors positive tempco.
The aim is to keep the idle current constant over temperature variations of the output transistors.
As such D1 should be mounted with close thermal contact ... either on the same heatsink or directly on one of the transistors.
Q5/Q9 form a classical Darlington power transistor, Q7/Q11 form a complementary Darlington (Sziklai).
C7(100µ/50V) bootstraps the output voltage, keeping the voltage over it´s pins almost constant and as such the biasing of the output transistors.
VR1 sets the output voltage level to roughly 1/2 of the power supply voltage.
When trimming the amp VR1 should be tuned so that the output signal clips symmetrically.
R35(15K)/R11(150R) form a dc feedback loop defining the closedloop dc gain.
For the ac gain R39(8k2)||C15(47p) and C13(1000µ/35V) add to the feedback equation.
C13 acts as a dc-blocking cap towards the speaker connector.
The ac-feedback path could also bleed away C13´s charge at switch-off (if the speakers are connected through a -then open- protection relay).
I´m not exactly shure about C19(100p) bridging Q1´s BE-junction.
It might either bootstrap to increase the input impedance at HF, and/or forms a soft HF-emv-filter in cooperation with R1(2k2)
The RC output filter (10R/22nF) is a classical Zobel, providing for a defined HF-load to the amp.

This kind of amp topology was quite common in the ´early years´ up to the mid-80s.
Currently I´m servicing a Marantz 2230 Receiver (1975 iIrc) that features almost identical amps.

regards
Calvin

Hi Calvin,

SO. MUCH. AWESOMENESS... in that!  Thank you sir!

- Wil

[fourfathom]

And that is exactly why bridges and connecting dots should both be ALWAYS used. And NEVER, EVER just two lines crossing without one or the other.

That's my opinion and I AM sticking to it. It prevents ALL chances of confusion.

And here's my strongly-held opinion:
Never have a crossing schematic connection.  Offset one of the wires so you instead have two connecting points, and use two connecting dots.
Bridges or gaps are hard to do with most schematic tools, and are completely unnecessary with the above rule.  Using the offset connections makes it much easier to at a glance see the topology of the circuit, and there's never any ambiguity.  You can even omit the dots and it's still clear (but do use the dots).