Unusual amp from 1987

Jump to Latest
minek123

minek123

Member
Joined 2014
minek.com/
Paid Member
>Beautiful amplifier turns out. With mirror

Told you! :)
I didn't sim your mirror, but I have a feeling my mirror (in my sim post #665) was running at smaller current..
I also changed CCS at the bottom.
I might be mistaken, but I thought optimal current for LTP transistors is 1mA .. 1.2mA (each).
 
Last edited:
S

stanislav1957

Member
Joined 2018
And faster and more stable.

I might be mistaken, but I thought optimal current for LTP transistors is 1mA .. 1.2mA (each).
Click to expand...
I try to make do with fewer resistor ratings.I counted the mirrors to 300. the output constant is less than.At low current, the distortion increases . When high current diffused ,the noise growing . Compromise!
 
Last edited:
S

stanislav1957

Member
Joined 2018
answered from above. We haven't optimized the diagram yet, but only the sketch
 

Attachments

  • Усь Клецова1ток&#1.PNG
    Усь Клецова1ток&#1.PNG
    109.3 KB · Views: 142
Last edited:
S

stanislav1957

Member
Joined 2018
But this is not essential.It's just support.Try changing the voltage of the Zener diode.There only the power dissipation will change on the transistors.You can choose a mode in which the first two transistors in the cascode can be low-power,and the lower, third, more powerful.
 
Last edited:
minek123

minek123

Member
Joined 2014
minek.com/
Paid Member
Maxim,
1) Can you show Thd and FFT for your sim?
2) What is power dissipation on Q2 ?

With 45V rails:

It looks like over 200mW.
That's kind of unusual for LTP CCS...
I think Q2 has be TO-126 transistor, or CCS need to be changed to allow 2N5551

Q1/Q3 also run over 100mW. Too much...
 
Last edited:
minek123

minek123

Member
Joined 2014
minek.com/
Paid Member
>Classic. 2-component feedback.

With this kind of feedback (33k + 33k + (5p .. 33p)), can't get square waves to look good.

With 'normal' feedback (33k/22pF) square waves look perfect.

With 2-component feedback, C must be at least 33p, but SR is getting lower...

============

Tried few sims, it looks like with 2-component feedback SR is lower.
Maybe I'm wrong, still testing..
 
Last edited:
S

stanislav1957

Member
Joined 2018
minek123 said:
Maxim,
1) Can you show Thd and FFT for your sim?
2) What is power dissipation on Q2 ?

With 45V rails:

It looks like over 200mA.
That's kind of unusual for LTP CCS...
I think Q2 has be TO-126 transistor, or CCS need to be changed to allow 2N5551

Q1/Q3 also run over 100mW. Too much...
Click to expand...
1 Distortion at rated power and low power
2IMD
3 I'll show you the DC analysis .in two variants.With equal power across the cascode transistors and an alternative
 

Attachments

  • Усь Клецова dttstortion.PNG
    Усь Клецова dttstortion.PNG
    24.3 KB · Views: 121
  • Усь Клецова1 ОБР.PNG
    Усь Клецова1 ОБР.PNG
    20.3 KB · Views: 123
  • Усь Клецова1 IMD.PNG
    Усь Клецова1 IMD.PNG
    46.4 KB · Views: 121
S

stanislav1957

Member
Joined 2018
In the presence of high-quality capacitors, a 2-component OOS is not necessary.But I'm used to it
The IMD graph is not accurate, and the analysis time was set to be short.It should be even lower.
R13 increase the nominal value to 750-1kom and the cascode will not warm up
 
Last edited:
minek123

minek123

Member
Joined 2014
minek.com/
Paid Member
So soon??? :)

We need higher slew rate. The 1st sim I published yesterday, had SR=50V/us.
Now my latest sim is back to 15V/us

We can't have all small transistor in input LTP running hot at 100mW-250mW, on the other hand - with lower
dissipation, RS goes down...
 

Attachments

  • kletzov_1983.asc
    10.4 KB · Views: 55
Last edited:
You must log in or register to reply here.
Top Bottom