AoE x-Chapters - 2x.11 BJT bandwidth, fT

[Winfield Hill]

Another x-Chapters section to read and think about.
Here's 2x.11 where we take up BJT high-frequency
bandwidth limits, Ccb, fT, etc. I've also included
the two preceding sections, 2x.9 and 2x.10, but we
didn't have the time (or maybe also the inclination)
to do much with those. I'm primarily interested in
your comments on section 2x.11, it's a useful topic,
and I hope we properly understand it, and are clear.
My interests peak a little ways below the real true
RF region, so no s-parameters, or any of that stuff.

https://www.dropbox.com/s/br9caixbtderl44/2x.9-11_BJT_beta_fT_DRAFT.pdf?dl=1


--
Thanks,
- Win

 

[John Larkin]

On 4 Aug 2019 12:00:22 -0700, Winfield Hill <winfieldhill@yahoo.com>
wrote:

Another x-Chapters section to read and think about.
Here's 2x.11 where we take up BJT high-frequency
bandwidth limits, Ccb, fT, etc. I've also included
the two preceding sections, 2x.9 and 2x.10, but we
didn't have the time (or maybe also the inclination)
to do much with those. I'm primarily interested in
your comments on section 2x.11, it's a useful topic,
and I hope we properly understand it, and are clear.
My interests peak a little ways below the real true
RF region, so no s-parameters, or any of that stuff.

https://www.dropbox.com/s/br9caixbtderl44/2x.9-11_BJT_beta_fT_DRAFT.pdf?dl=1

P 135, I've seen lots of transistors ocillate with a well-bypassed
collector. That's not always a fix. A base resistor apparently is.

Wire bonds are evil.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[bitrex]

On 8/4/19 3:00 PM, Winfield Hill wrote:

Another x-Chapters section to read and think about.
Here's 2x.11 where we take up BJT high-frequency
bandwidth limits, Ccb, fT, etc. I've also included
the two preceding sections, 2x.9 and 2x.10, but we
didn't have the time (or maybe also the inclination)
to do much with those. I'm primarily interested in
your comments on section 2x.11, it's a useful topic,
and I hope we properly understand it, and are clear.
My interests peak a little ways below the real true
RF region, so no s-parameters, or any of that stuff.

https://www.dropbox.com/s/br9caixbtderl44/2x.9-11_BJT_beta_fT_DRAFT.pdf?dl=1

The 2N3904 looks like Mom's meatloaf or the Cleveland Brown's
aspirations as a team, in transistor form. It's not fancy or nothin' but
when filet mignon isn't available or too expensive it won't do ya wrong

 

[Phil Hobbs]

On 8/4/19 3:00 PM, Winfield Hill wrote:

Another x-Chapters section to read and think about.
Here's 2x.11 where we take up BJT high-frequency
bandwidth limits, Ccb, fT, etc. I've also included
the two preceding sections, 2x.9 and 2x.10, but we
didn't have the time (or maybe also the inclination)
to do much with those. I'm primarily interested in
your comments on section 2x.11, it's a useful topic,
and I hope we properly understand it, and are clear.
My interests peak a little ways below the real true
RF region, so no s-parameters, or any of that stuff.

https://www.dropbox.com/s/br9caixbtderl44/2x.9-11_BJT_beta_fT_DRAFT.pdf?dl=1

More interesting stuff, thanks.

You might want to add another sentence to the discussion of high-level
injection effects on beta--the origin of the base current is
recombination, which goes as the product of the electron and hole
densities. HLE is when injected carrier density >~= doping density,
resulting in enhanced recombination and so lower beta.

A Hartley oscillator is a Colpitts made by tapping the tank inductor
instead of the capacitor, and there's no tapped element in your circuit.
The emitter-follower oscillator is a negative-resistance oscillator,
which is why adding base resistance fixes it.

You might mention Carl Battjes' fT doubler, in which each side of a diff
pair has two transistors with their bases wired in series and their
collectors in parallel. (Battjes has a chapter in one of Jim Williams's
books, entitled "Who wakes the bugler?".)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Chris Jones]

On 05/08/2019 06:36, John Larkin wrote:

On 4 Aug 2019 12:00:22 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

Another x-Chapters section to read and think about.
Here's 2x.11 where we take up BJT high-frequency
bandwidth limits, Ccb, fT, etc. I've also included
the two preceding sections, 2x.9 and 2x.10, but we
didn't have the time (or maybe also the inclination)
to do much with those. I'm primarily interested in
your comments on section 2x.11, it's a useful topic,
and I hope we properly understand it, and are clear.
My interests peak a little ways below the real true
RF region, so no s-parameters, or any of that stuff.

https://www.dropbox.com/s/br9caixbtderl44/2x.9-11_BJT_beta_fT_DRAFT.pdf?dl=1

P 135, I've seen lots of transistors ocillate with a well-bypassed
collector. That's not always a fix. A base resistor apparently is.

Yes, I am not at all convinced that a high impedance at the collector is
necessary for an emitter follower to oscillate. If there is a capacitive
load on the output (emitter) then I think that it can make the input
impedance have a negative real part.

I found this document interesting, though I cannot claim to have
understood it fully yet:
http://audioworkshop.org/downloads/AMPLIFIERS_OSCILLATION_BJT_CIRCUITS.pdf

 

[John Larkin]

On Mon, 5 Aug 2019 23:00:59 +1000, Chris Jones
<lugnut808@spam.yahoo.com> wrote:

On 05/08/2019 06:36, John Larkin wrote:
On 4 Aug 2019 12:00:22 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

Another x-Chapters section to read and think about.
Here's 2x.11 where we take up BJT high-frequency
bandwidth limits, Ccb, fT, etc. I've also included
the two preceding sections, 2x.9 and 2x.10, but we
didn't have the time (or maybe also the inclination)
to do much with those. I'm primarily interested in
your comments on section 2x.11, it's a useful topic,
and I hope we properly understand it, and are clear.
My interests peak a little ways below the real true
RF region, so no s-parameters, or any of that stuff.

https://www.dropbox.com/s/br9caixbtderl44/2x.9-11_BJT_beta_fT_DRAFT.pdf?dl=1

P 135, I've seen lots of transistors ocillate with a well-bypassed
collector. That's not always a fix. A base resistor apparently is.

Yes, I am not at all convinced that a high impedance at the collector is
necessary for an emitter follower to oscillate. If there is a capacitive
load on the output (emitter) then I think that it can make the input
impedance have a negative real part.

An emitter follower with well bypassed base and collector, and just a
resistive pulldown on the emitter, will often oscillate.

I found this document interesting, though I cannot claim to have
understood it fully yet:
http://audioworkshop.org/downloads/AMPLIFIERS_OSCILLATION_BJT_CIRCUITS.pdf

That guy does a lot of impressive-looking math and ignores the leads
and wire bonds.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[whit3rd]

On Monday, August 5, 2019 at 8:10:19 AM UTC-7, John Larkin wrote:

An emitter follower with well bypassed base and collector, and just a
resistive pulldown on the emitter, will often oscillate.

Well, with some inductive feedback allowed between emitter or collector current and the base
wiring, you'll get over-unity gain sometimes. It's usually voltage slew capacitor coupled
that gets 'bypassed'. A resistor in the base lead is the R-L filter component you need, sometimes.

Either that, or careful wire placement.