Apex A40 fundamental improvement. (Sandy)

[SandyTodorov1]

For not so disgusting. (APEX A40)

 

[SandyTodorov1]

For not so disgusting. (APEX A40)

It has not been completed.

 

[NMOS]

Very much compnents

whats the benefit ?

if its working please show it

 

[SandyTodorov1]

Which components can be removed without compromising parameters.

 

[U130421]

Do you have the open loop response of the original A40 to compare. Of which version is the measurement?

 

[SandyTodorov1]

Comparison of APEX-A40 and improved version.
APEX-A40 is an example of bad electronics.

 

[sajti]

Comparison of APEX-A40 and improved version.
APEX-A40 is an example of bad electronics.

Not necessarily bad, but there are many wrong part selection inside.

Sajti

 

[SandyTodorov1]

To use MJE340 / 350 for such small currents talk about the extraordinary incompetence of the author of the APEX A40.
Especially T25 / 26 which need not be high voltage.

 

[sajti]

To use MJE340 / 350 for such small currents talk about the extraordinary incompetence of the author of the APEX A40.
Especially T25 / 26 which need not be high voltage.

Sure. Another wrong decision is to use BC547/557 in the current mirror, because of the quasi saturation. BC337/327 is much better. BC550/560 would be better in the input stage because of the low Vce.
Another problem is the 2x1N4148 in the VAS, which gives not enough voltage shift. Much better to use LED instead of the 2 diodes.
Sajti

 

[SandyTodorov1]

Be used to Ic 1-2mA transistor with a maximum current 0,8A is a very bad idea.
Ft transistor for such small current is considerably less when the transistor is made for large Ic

 

[SandyTodorov1]

Comparison of the growth rate.
There is also a test with reduced C1 (10pF) to remove the influence of the R2C1 circuit.
APEX A40 (+87V/uS -53v/uS)
Sandy C1 = 100pF (+534V/uS -400v/uS)
Sandy C1 = 10pF (+829V/uS -920v/uS)

 

[sajti]

Be used to Ic 1-2mA transistor with a maximum current 0,8A is a very bad idea.
Ft transistor for such small current is considerably less when the transistor is made for large Ic

But using BC547 as T17, and T18 is even worse. That transistors are close to the saturation, and BC547 is one of the worst candidate. Check this topic:Looking for a small signal NPN with no Quasi Saturation: MEASURED DATA
Sajti

 

[U130421]

You are using c3503/a1381 for 3-5 ma have Ft of 80 to 100Mhz where its max 200Mhz Ft is at 30ma . Ok, c3423/a1360 are no more manufactured but widely available for DIY for 3 -10 ma has Ft of 200Mhz.

 

[SandyTodorov1]

But using BC547 as T17, and T18 is even worse. That transistors are close to the saturation, and BC547 is one of the worst candidate. Check this topic:Looking for a small signal NPN with no Quasi Saturation: MEASURED DATA
Sajti

So usually use a current mirror rather than current amplifier x5 and I secured Uce 4V of T15

 

[U130421]

You are forgetting that apex copied the front end of Marantz.

 

[SandyTodorov1]

I know that APEX-A40 is borrowed from this amp but is made worse.
What I suggest is significantly better both Marantz and the APEX-A40.

 

[sajti]

So usually use a current mirror rather than current amplifier x5 and I secured Uce 4V of T15

Your solution is better in this case.

Sajti

 

[SandyTodorov1]

Here gain without feedback.
And remove protection to understand easily scheme.
The use of T25/26 and T21 special tethering Т21/24 compensating h22b and Ccb of T16/17 allows to obtain amplification of 160db without the need for amplification in the T13/15 T14/18

 

[U130421]

To get the open loop response, on the original circuits just add 1G Henry in series with the feedback resistor.

 

[SandyTodorov1]

There is no difference between suggestions from you and my method.
Why put two items instead of just one.