Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Inverting OP-AMP - how does it work ?

Status
Not open for further replies.

Nigel Goodwin

Super Moderator
Most Helpful Member
psevdonim said:
how does the 741 in this schematics https://www.electronics-lab.com/projects/audio/041/index.html work ?
If R10 would go to pin 6 of 741, I'd understand it. But here the C2 interrupts the connection.

The 741 feeds R11, the current feeding this resistor (as always) has to come from the supply pins, so the current to the supply pins is used for seperate positive and negative outputs feeding the output transistors. As normal the feedback comes from the output of the complete amplifier, and NOT just the output of the opamp.

It's rather an unusual and confusing technique - it's done simply because the 741 requires a lower voltage supply than the output stage - this technique not only limits the voltage to the 741, but also amplifies it's output to more than it's supply rails.
 

TheVictim

Member
C2 is passing audio frequencies on to the darlington pairs AND to the feedback resistor. C1 passes high frequency noise around R10 and back into the input where it can be eliminated.

Edit: Yes, R10/R1 for audio frequencies, so a gain ~ 10. As the frequency gets higher the reactance of C1 drops, removing more of it from the input, eliminating hiss.
 
Last edited:

psevdonim

New Member
I can't quite agree with the function of C2, because it's 5600pF, so it has little effect to audio range, and it's connected more to the speaker rather than to output transistors.
This schematics is about 25 years old. Does anyone know, who the original constructor is and where was it published for the first time ?
 

audioguru

Well-Known Member
Most Helpful Member
I think the old circuit will have high distortion at high frequencies because the 741 opamp is so bad.
I also think the output darlington transistors will have thermal runaway.
They get warm which causes them to conduct more, which causes more heat, which causes them to conduct more, which causes ... (meltdown).
 

TheVictim

Member
The circuit has complexity that just isn't needed. There is no reason to use darlington pairs in a push-pull here anymore. Even if you still used the 741 you can use a single supply voltage biasing the non-inverting input with 1/2Vcc, then use a power amp like a 386 (or something with a bit more performance, but it is a good match for the venerable 741).
 

Nigel Goodwin

Super Moderator
Most Helpful Member
TheVictim said:
The circuit has complexity that just isn't needed. There is no reason to use darlington pairs in a push-pull here anymore. Even if you still used the 741 you can use a single supply voltage biasing the non-inverting input with 1/2Vcc, then use a power amp like a 386 (or something with a bit more performance, but it is a good match for the venerable 741).

Whatever you're smoking, how about sharing it? :p

The circuit is probably as simple as it can be, how do you propose to get 80W RMS from an LM386?.
 

TheVictim

Member
lol, OK, I should probably pay more attention to what is going on, but there are still single package solutions that exist he could use.
 

TheVictim

Member
Digging through my junkbox I found a STK439. Quite dated, but it will deliver the necessary output, and in stereo, in one package.
 

Nigel Goodwin

Super Moderator
Most Helpful Member
TheVictim said:
Digging through my junkbox I found a STK439. Quite dated, but it will deliver the necessary output, and in stereo, in one package.

They tend to be fairly expensive though, require a LOT of support components, and quickly become obselete - plus any failure requires replacement of the entire expensive IC (if it's still available?, and they rarely are!).

The original link is a VERY old Motorola application note, and there are obviously many much more modern solutions - but the STK439 isn't really one of them.

The main 'advantage' of the circuit is that it's pretty well a 'junk box' design, it's easy to make from parts you probably already have to hand.
 

TheVictim

Member
I pretty much KNOW it isn't available anymore (it was salvaged from a water damaged Japanese stereo circa 1980), but it was just sitting in my junk box.

I think Audioguru is right about the possibility of thermal runaway on the darlington pairs. They need something to limit the E-C current flow.
 

Nigel Goodwin

Super Moderator
Most Helpful Member
I would certainly prefer to see emitter resistors in the output darlingtons, although it might be that the transistors aren't biased on at all without drive?, relying on the 'high' gain of the opamp to overcome crossover distortion.
 

audioguru

Well-Known Member
Most Helpful Member
An old 741 opamp cannot slew fast enough to reduce crossover distortion in output transistors. Darlingtons would be twice as bad.

I installed and fixed Korean telephone systems in 1980 and they used a 1458 dual 741 opamp driving complementary output transistors with their bases tied together. The cutomers complained about the horrible crossover distortion so I added a resistor which allowed the opamp to drive the 45 ohm speaker directly then the output transistors boosted it.
 

Attachments

  • crossover distortion.GIF
    crossover distortion.GIF
    6.3 KB · Views: 364

Hero999

Banned
Even with your modification it will still sound crap.
 

TheVictim

Member
When making impedance calculations for circuits using darlingtons, do you multiply the gain of each stage within the darlington together, or is another method used?
 

audioguru

Well-Known Member
Most Helpful Member
Hero999 said:
Even with your modification it will still sound crap.
No, it sounds fine (except for the poor quality from the 1458 dual 741 opamp).

The opamp has a max output current of 20ma which drives the 45 ohm speaker through a 39 ohm resistor until the voltage across the resistor is high enough for the output transistors to turn on.
The negative feedback keeps the gain at what it needs to be for a smooth output.
 

Shytalker

New Member
Unusual Amp Circuit

psevdonim said:
I can't quite agree with the function of C2, because it's 5600pF, so it has little effect to audio range, and it's connected more to the speaker rather than to output transistors.
This schematics is about 25 years old. Does anyone know, who the original constructor is and where was it published for the first time ?

I don't know who designed this but I have seen a variation of this technique in a top-rank Hi-fi amp but with better output driving and a much better op-amp. In fact the one I looked at had eight power MOSFETs on each side of the output totem pole with 0.25ohm source resistors on each one for thermal stability. It worked very well.

I can see a high-frequency feedback arrangement here. If you imagine for a moment that the output is going positive, this would be coupled back into the op-amp output via C2 (at higher frequencies) and reduce emitter current in T1. This would reduce base current in T3 and tend to cancel the original positive movement on the output. The overall effect would be to reduce open-loop gain at hf and give the circuit more stability (less likely to become an oscillator).

The closed-loop gain is set by the ratio of R10 to (R1 + source impedance) with C1 causing a drop off in the gain at higher frequencies. The input impedance of the amplifier is more or less determined by the value of R1.

(I'm late replying to this one as I've only just signed up.)
 

Nigel Goodwin

Super Moderator
Most Helpful Member
TheVictim said:
When making impedance calculations for circuits using darlingtons, do you multiply the gain of each stage within the darlington together, or is another method used?

You basically treat it as a single transistor with a VERY high gain, in the case of the circuit above it is physically just a single transistor.
 
Status
Not open for further replies.

EE World Online Articles

Loading
Top