Shunt Current Feedback

[Stee]

is important to know that the differential stage introduces emphasis
the correct method to apply the subtraction is the potentiometer

so it takes an inverting amplifier
preferably with high input impedance

is also good solution
take the feedback current
able to equalize the performance of the speaker

this is my configuration happier
provides a dynamic scene (XLR source indispensable)
detailed and intensive

 

[DF96]

This senses the output current. It will therefore raise the output impedance, which will create an 'interesting' frequency response with almost all speakers (which are usually designed for a low output impedance). For example, you will get a bass boost at the bass resonance. It may sound like a cheap ghetto blaster.

 

[Ouroboros]

There are some good arguments for driving a speaker unit from a high impedance, made by Malcolm Hawksford among others. I am not wholly convinced though, except for increasing Qts in open-baffle loudspeakers.

 

[DF96]

But only when the speaker was designed to be driven that way. Most speakers are not. They require electrical damping of their mechanical resonance.

 

[Stee]

yes of course

my loudspeakers are unfiltered
I don't know what happens with cross-over

 

[analog_sa]

detailed and intensive

Have you actually built an M-9 clone?

Do you think the input fet buffer will actually work into a dead short?

 

[DF96]

my loudspeakers are unfiltered
I don't know what happens with cross-over

Not a matter of filtering or crossovers, but driver and cabinet design. What controls the resonances, electrical or mechanical damping?

 

[Stee]

actually there is a problem with the resonance
the system oscillates in the radio frequency
when the gain control is very close to returning from shunt

I have 4 pieces of M9 with Motorola power Bjt
but I wanted to edit an A100
the model that was born before
http://www.diygene.com/schemas/a-100sch.pdf

I deleted two input resistances and OP-Amp
to bring the next sequence M9
The PSU is made with 4X12V batteries with relais
and 2 switching PSU 24V

I have never heard better
a connection has been lucky
probably because it's rare to have an inverting amplifier

 

[forr]

Were the amp not inverting, using some precautions, it could give a negative output impedance to lower the electrical damping of a driver.

 

[analog_sa]

I have never heard better

I don't doubt this for a sec. Still, can you say into what impedance is the input buffer working? What kind of distortion can you expect from it alone?

 

[Stee]

high impedance jfet input buffer

The input buffer is the heart of the system
because it allows you to bring out the detail with great accuracy
extrapolated from a difference (the potentiometer) from very low impedance
so as to avoid a tube preamp

to test the limit of the resonance
I applied a resistance 10khom
the potentiometer output

Tomorrow I will try to apply a small capacitor
As is normally used for the feedback resistor
and I will try to lower the shunt resistance of non-inductive 0R22

the transformer is the ideal tool to perform addition (Hot and Cold)
as said before
the differential stage (for the valves too) is not suitable for audio

the best input transformer is a Tamura
600 to 300 ohm 2:1
This feature allows you to get a low susceptibility
of the source (DAC1 Little_Dot)
and also helps reduce the harmonics of even order
making the sound pleasant to listen to long

Bias is AB1 with a small fan on the radiator
not known at the time distortion
even with a full piano

still can not reach a minimum volume

 

[analog_sa]

The input buffer is the heart of the system

I see you avoid my simple question. May be a language issue, so let's get some numbers.

Originally, the fet buffer is loaded with a 470R resistor. At 1v input 1kHz second harmonic is around -80db and third -76db. Quite acceptable.

In your improved circuit fets are working into something closer to 30ohm load. Third harmonic is now at -41db.

Intensive indeed.

 

[Stee]

About external NFB

I do not know how do you calculate the load of the JFET
because it is floating
the zero point determined by the resistance of feedback does not exist anymore

 

[analog_sa]

The load are the common base input transistors. Why do you think they had to use an input buffer in the first place?

 

[Stee]

property of global feedback

because if you put the buffer in the feedback loop
After you have a low input impedance
that's why you need a transformer

Note that with the JFET into the ring
corrects the error of the buffer itself

to calculate the load buffer
must consider the current flowing on the common base input stage

 

[mhouston]

is important to know that the differential stage introduces emphasis
the correct method to apply the subtraction is the potentiometer

so it takes an inverting amplifier
preferably with high input impedance

is also good solution
take the feedback current
able to equalize the performance of the speaker

this is my configuration happier
provides a dynamic scene (XLR source indispensable)
detailed and intensive

I have seen ccts. using chips which introduce shunt FB. What I would call a transconductance amp. The author never really felt it was an absolute solution more of an interest to be investigated.

In the schematic show there are far too many active and passive parts. Every part buggers the music!

 

[Stee]

more simple amplifier

see this solution

 

[analog_sa]

Every part buggers the music!

It depends. The circuit above will generate more than 1% thd, mostly third and not counting the transformer. Apparently this contributes to an "intensive" sound.

 

[Stee]

optimized inverting amplifier

simple version

 

[Stee]

What about simple version?

It depends. The circuit above will generate more than 1% thd, mostly third and not counting the transformer. Apparently this contributes to an "intensive" sound.