For me it is an Allison with a bootstraped power supply.
I am a generalist or i get crazy.
MosFets, BJT, JFet, Tubes, Opamps, just building blocks for me.
We talked about 4 pole theory before.
2 inputs, 2 outputs.
They hammered that into my head at Paderborn in the 70th.
Taking about 9 dimensional matrix is another matter.
When i read my old calculations i do not understand a word or number.
Was that me that solved that equations ?
I was channeled, so high was the preasure.
I am a generalist or i get crazy.
MosFets, BJT, JFet, Tubes, Opamps, just building blocks for me.
We talked about 4 pole theory before.
2 inputs, 2 outputs.
They hammered that into my head at Paderborn in the 70th.
Taking about 9 dimensional matrix is another matter.
When i read my old calculations i do not understand a word or number.
Was that me that solved that equations ?
I was channeled, so high was the preasure.
Last edited:
And this is (I think) the cool part, I think it is both. Common base DC, Allison AC, ... a mix of kinds actually
(most probably Joachim is the one that has the answer, but it is a funny (possibly new) circuit).
This circuit does not show the same thing.
Tr at Vin drives base of output darlington.
Since collector of a transistor can never be the input,
Tr1 and Tr2 don't act as input drivers. They are driven
from output. I would call this driven bias.
Have to look for the articke by M. Hawksford (Bias
feedforward), it seeems to be similar.
Here is another case of a common base stage proceeding the output stage :http://www.keith-snook.info/wireles.../New Approach to class B Amplifier Design.pdf
A different view at the circuit of post 9983 http://www.diyaudio.com/forums/analogue-source/154210-mpp-999.html#post4006041 :
If you tell Tr at Vin to be a current source and move Vin
to the emitters of Tr1, Tr2 you admittedly arrive at a
structure which is very similar to the "final" circuit of post
10013 http://www.diyaudio.com/forums/analogue-source/154210-mpp-1002.html#post4007835 and with about the same parts count, but the
"Allison", if this is it, has gain and is inverting, the "final"
has non inverting unity gain. It is no longer "Allison" as
we knew her..
If you tell Tr at Vin to be a current source and move Vin
to the emitters of Tr1, Tr2 you admittedly arrive at a
structure which is very similar to the "final" circuit of post
10013 http://www.diyaudio.com/forums/analogue-source/154210-mpp-1002.html#post4007835 and with about the same parts count, but the
"Allison", if this is it, has gain and is inverting, the "final"
has non inverting unity gain. It is no longer "Allison" as
we knew her..
That is a really interesting article and I haven't seen it before. Thanks for posting. That is well-worth simulating with good models. Unfortunately my computer is crashed.
Steve Dunlap's smooth-switching output stage is one of the extremely few output stages I know of that actually achieves smooth switching in the crossover region. THD isn't remarkable, and the nonlinearity is shifted out into the excursions, but in terms of sonics it should score much better on the Gedlee metric you posted about.
Although the Gedlee metric is missing the frequency-dependent terms so it only operates on a static transfer curve. So if the transfer curve is distorted by feedback in it's -6db/oct frequency-dependent region, the Gedlee metric doesn't know about it.
Steve Dunlap's smooth-switching output stage is one of the extremely few output stages I know of that actually achieves smooth switching in the crossover region. THD isn't remarkable, and the nonlinearity is shifted out into the excursions, but in terms of sonics it should score much better on the Gedlee metric you posted about.
Although the Gedlee metric is missing the frequency-dependent terms so it only operates on a static transfer curve. So if the transfer curve is distorted by feedback in it's -6db/oct frequency-dependent region, the Gedlee metric doesn't know about it.
as_audio has explained the MosFet buffer in detail.
One could say that the diamond buffer is a real buffer and the Frans designed MosFet buffer is an amplifier with 100% feedback so that the amplification factor is x1.
That allows to setup the MosFet buffer with gain.
as-audio has send me a schematic that i will try.
One could say that the diamond buffer is a real buffer and the Frans designed MosFet buffer is an amplifier with 100% feedback so that the amplification factor is x1.
That allows to setup the MosFet buffer with gain.
as-audio has send me a schematic that i will try.
