Unusual amp from 1987

[stanislav1957]

That's it. C6 10pF and everything works.
Meander learned to look in LTspice

 

[minek123]

Maxim, you don't sleep?
Beautiful square waves

You also tweaked couple of other values (R,C)..

 

[stanislav1957]

Even better.
It is necessary to look at the meander signal and the base of the first transistor .The distortion graph at 20 kHz would be interesting to see.

 

[minek123]

From the square waves, it looks like Slew Rate is approx. 30V/us
Much better than before.

Still 0.0 % Thd
FFT for 1kHz sinus looks flatter (2nd, 3rd, 4th harmonics),
and everything is lifted up by 10 dB.

Looks like Q6 (in my sim) has to be low beta (A type). C type is more prone to trigger oscillations.
Are BC546/556 OK in there?
Q4 is running at over 80mW, little bit high for a small transistor?

 

[stanislav1957]

Better watch IHD

 

[stanislav1957]

From the square waves, it looks like Slew Rate is approx. 30V/us
Much better than before.

Q4, is running at over 80mW, little bit high for a small transistor?

Put a 150ohm resistor between emitters Q1 and Q4 .Your file

 

[minek123]

Thanks Maxim!

 

[stanislav1957]

Thanks Maxim!

Bad idea. a 100-200 Ohm resistor between the Q1 and Q2 ' on your file you need to put

 

[minek123]

Works like a champ. 40mW power.

Q1 and Q2 ?
Not Q1 and Q4 ?

 

[stanislav1957]

Works like a champ. 40mW power.

Q1 and Q2 ?
Not Q1 and Q4 ?

quite so. Otherwise the gain drops

 

[minek123]

1) Resistor in the collector of Q1 (between Q1 and Q2) lowers power consumption in Q1 only. Q4 stays at 80mW.

2) Resistor between emitters of Q1 and Q4, lowers power consumption in Q4, which is what we need, right?

 

[stanislav1957]

Quite true. Otherwise, the distortion increases greatly

 

[LKA]

Hi, I downloaded and run the sim, post #543. Is it OK ?

 

[minek123]

>Hi, I downloaded and run the sim, post #543. Is it OK ?



This is latest of mine, C7 increased to correct shape of square.
We are still not clear where the resistor should go Q1/Q2/Q4
so I left it out. The worst case - Q4 needs to be bigger.

Maxim is using Multisim, and from what I've seen, square waves looked better in his Multisim simulations...

I'm getting different shapes/small oscillations depending on the amplitude of the square signal.
It may look good for 2 Vpp, but worse for 1.6 Vpp


It's still "work in progress", but getting there...

 

[stanislav1957]

Here is my meander.The signal at the oramp input (summation point) is good
I use Microcap12 .

 

[metal]

Minek, by the looks of it, Maxim uses Microcap, not multisim.

 

[stanislav1957]

And the signal level of the square wave at the output of the generator can be reduced? I'm sure the level is high

 

[Nico Ras]

Minek, how does it sound, have you run the wave simulation and listened to it yet. One can easily spot sound signatures listening to it even with zero distortions.

 

[Nico Ras]

You should also use the spread in component tolerances, as is you are using identical components with zero tolerance. It is all there in the manual. One can actually make a real amplifier and listen to it before touching a soldering iron or ordering a component.

 

[stanislav1957]

Lowered the level to -0.5/0.5 and the picture became normal