schematic question - spectra sonics 110A

[QUEEF BAG]

not a repair question...just curious.

i pulled up this schemo just for something to look at while i ate a snack.
after a couple minutes, i am thinking WTF?

i do not under stand this input.  i don't remember ever seeing input going to an emitter.

it's not exactly like i'm a noob,   i've had years of experience not knowing much

what gives with this?

[Brian Roth]

From Ye Olde Daze, that is somewhat similar to a grounded-grid vacuum tube circuit.

low Z input impedance

Bri

[abbey road d enfer]

Yes, as Bri says. It has some advantages in terms of noise for low source impedance.
The designer must have been a radio designer. 
A typical common-base stage has very low input Z (about base impedance divided by beta), but since it operates under NFB, the emitter input impedance seen by the source is high enough, the 2.7k emitter resistor is probably the governing element here, so if a 1:3 input xfmr is used, the input impedance would be around 300r, which is adequate by broadcast standards (remember they are used to impedance matching, impedance bridging is a waste of power for them).

[JohnRoberts]

While I am not familiar with this topology, from the notes about gain, the feedback resistor forms a divider with the fixed 100 ohm and that gets cap coupled into the input transistor base. The transistor can be thought of as having a + and - input. A positive voltage at the base, causes more current to be pulled down the collector so this is the negative input. Alternately, moving the emitter voltage positive with respect to the base reduces the collector current so acts like a + input. The rest of the circuit inverts the signal polarity so that the overall feedback is negative. 

We are far more familiar with the transistor base used as the input in circuits because it is high impedance compared to the emitter, but this circuit demonstrates it can be used the other way around too.

It works, but i can't say that I see much advantage other than "hey guys, I bet you didn't think I could maker this work"....   

When I was in the steeper part of my learning curve (still learning) I would read schematics voraciously. These little circuit tricks are tools that we put in our circuit design library for possible future use. Who knows, some day you may want to use a transistor this way.

Keep studying.

JR

[QUEEF BAG]

thanks for the responses.

i now have more info sitting on my brain that will slowly seep in.

[Twenty Log]

Some logic gate families use this similar approach (some TTL, I'm also thinking ECL/PECL/MECL et cetera)... logic gates have "high gain" (or maybe just conditionally (un)stable is another term?)   

One can use "analog" style feedback with some logic gates and get them to be a little more "linear"...  Some families, can allow direct "shorts" from output back to input to make (un)latches, pulses, oscillators or whatnot...  caveat logicae... YMMV....

[abbey road d enfer]

Common-base (or common-gate for FET or common-grid for tubes) stages have become second-nature for RF engineers. The main reason is that they do not suffer the Miller effect (the output and input are in-phase) and their input impedance is easier matched to LC circuits.
I don't think there's any definite advantage using them at audio frequencies.

[Brian Roth]

For some additional info, see:

http://www.google.com/patents/US3376515

The circuit shown is pretty much the original Spectra Sonics 101 amplifier.  The later 110 is still similar.

Bri

[QUEEF BAG]

good info in the patent docs.  thanks!
thanks for all the "feedback"

in the early eighties i worked at a studio that had a couple racks
full of spectrasonics amp cards.  mostly for phones and CR monitors.

a lot of them blew up.     several crown power amps wandered in.

[Brian Roth]

Bill Dilley also used the same basic design to make a summing amp (I don't recall the Spectra Sonics card number):

http://www.google.com/patents/US3517270

Spectra Sonics used these various building blocks to create some sophisticated desks, beginning back in the early 1970's:

http://brianroth.com/library/spectra.pdf

Bri

[JohnRoberts]

FWIW and not the exact same circuit, there is a variant virtual earth summer where the - input (bus) is fed into the emitter of a common base transistor amp. The low noise Transamp mic pre had a had a popular summing amp application used that way.

JR

[Brian Roth]

I wonder if Dilley was the pioneer for using a virtual ground summing amp in a pro desk?  His patent was filed in 1969.

Bri

[abbey road d enfer]

I wonder if Dilley was the pioneer for using a virtual ground summing amp in a pro desk?  His patent was filed in 1969.

Bri

As usual, in these patent files, it's very difficult to find out exactly what is new and what is just current state-of-the-art, but it seems it is more about the arrangement of transistors than the virtual earth summing.
In fact, it is often easier to patent something that does something than a conceptual notion such as vitual mixing.
I think the concept of mathematic summing with operational amps was around for some time, with the current-then interest in analog computers. Audio guys may have just seized the opportunity.
If you read carefully the text, you may notice a curious claim that "Q2 and Q3 are connected to operate in common-base mode"; although it is true for Q3, which receives signal on its emitter, it is not the case for Q2, which, being part of a diff pair, receives signal on its base and outputs on both its emitter and collector.

[JohnRoberts]

+1 virtual earth summing was probably predicted by analog computing.

The Burr Brown Operational Amplifier book c.1971 describes the virtual earth sum amp. I suspect it was also in app notes from back then.

JR

[dcollins]

+1 virtual earth summing was probably predicted by analog computing.

The Burr Brown Operational Amplifier book c.1971 describes the virtual earth sum amp. I suspect it was also in app notes from back then.

About 20 years earlier actually.

http://www.philbrickarchive.org/k2-w_refurbished.pdf


DC

[JohnRoberts]

Once again the ancients are stealing our ideas...

JR

[Brian Roth]

While I realize the Philbrick opamp existed well before Dilley's patent, can anyone name an AUDIO desk/mixer which used virtual ground summing prior to, say, 1970?  That was my original question.

Bri

[JohnRoberts]

That's before my time. I didn't get involved in designing consoles until the late 70s, but can't image not using virtual earth inverting summers back then...

Since then I imagine all kinds of alternate approaches.

JR

[Brian Roth]

John....1970 was "before my time" as well!  I was still in High Screwel!!   

All I know from that era was passive summing followed by a "make up" gain amplifier.  Well into  the 1970's, Neve continued to use that same topology....and those same Neve desks (such as an 8068) command six-figure prices today.

I am NOT suggesting passive summing is better/worse than active summing....ain't gonna go there!  lol

FWIW, by 1977 I was familiar with using opamps as virtual ground summers....hence that was in the desk I designed/built.  I have NO idea where I learned of the concept, however.

I was merely curious when virtual ground summing first arrived on the scene for audio desks.

Bri

[JohnRoberts]

I do not recall where I first saw it but suspect it was probably in opamp applications notes or perhaps Nat Semi analog applications handouts.

The passive sum with make up gain is a carry over from early days when gain stages were common standardized circuit modules.

My recollection was that virtual earth was the "obvious" way to accomplish that task, but I truly do not recall seeing it somewhere first, presented as new or novel. Just another popular opamp trick.

JR