Differential Amplifier question (probably very simple)

[realkillercults]

HELLO
In the past, I've built myself a stompbox that appropriated certain parts of a certain schematic of a certain old solid-state preamp. The original preamp had this odd "tube" voice emulator circuit that, in my limited understanding of EE, is nothing but an op-amp gain stage feeding a differential amplifier transistor network thing; the diff-amp outputs are fed into opposite sides of a transformer, giving a single output that then reaches an op-amp output amplifier stage.

the diff-amp/transformer stage here:



while I have successfully built a working unit using this piece of the schematic, I wonder if there is a different method by which I can pair the combined output of Q20/Q21 with something other than a transformer (I think this is called balancing the signal?). could I perhaps use an op-amp in some capacity? another transistor network? something? anything that's not a big dumb expensive transformer?

from what I've read and understand about differentials, Q20 and Q21 are producing a + and - phase signal(?). I am not sure which tranny produces which, and I'm not sure what exactly happens to the phasing within the transformer...

any help is appreciated. thank you much! this is my first post !!

[Rob Strand]

It's the IVP preamp for sure  .

The most direct way to combine the outputs is with an opamp based differential amplifier:



Probably choose
R1 = R2 + R4
R3/R1 = R4/R2

Equal R's will look like a 2:1 transformer.
And R4 = 2* R2, R3 = 2* R1  looks like a 1:1 transformer.


[Edit:
Forgot to add you will need some series caps in each arm to stop the added stage affecting the DC level
]

[realkillercults]

It's the IVP preamp for sure  .

SHUT UP.
you got me.

I suspected opamp was the way to go. thanks.
Which Q produces which signal though?

[Rob Strand]

SHUT UP.
you got me.

Ha!

Which Q produces which signal though?

Both.

[realkillercults]

Both.

lol
okay, smart guy. I'll assume you mean to say it doesn't matter which collector lead serves as V1 or V2.

[Phoenix]

Which Q produces which signal though?

Q20 produces an inverted output, Q21 produces a non-inverted output.

[Rob Strand]

lol
okay, smart guy. I'll assume you mean to say it doesn't matter which collector lead serves as V1 or V2.

I wasn't trying to be smart.  That's what it does (and independent of me!).

[PRR]

Welcome, realkillercults.

The transformer (flaws) may be VERY important to "tone".

[realkillercults]

Welcome, realkillercults.

The transformer (flaws) may be VERY important to "tone".

I've thought this may be the case. I'm up to experiment though of course.

outside of bandwidth constriction, any ideas what a transformer would really be contributing here?

[Rob Strand]

The load on the transformer can have an effect as well.

[Phoenix]

outside of bandwidth constriction, any ideas what a transformer would really be contributing here?

Potentially saturation at high signal amplitudes, which causes loss of inductance and therefore loss of bass response, giving a dynamic bandwidth shift.
Potentially non-flat bandwidth, which of course could be replicated with alternative passive circuitry, but it's probably simpler to just use a transformer.
Small amount of crossover distortion due to core hysteresis, which can be different in character from typical AB amplifier crossover distortion.
Various other transformer non-linearities.

Which of these is more important, or even really matter is going to depend a lot on the exact circuit configuration, how hard it's driven, and the transformer used.

[realkillercults]

outside of bandwidth constriction, any ideas what a transformer would really be contributing here?

Potentially saturation at high signal amplitudes, which causes loss of inductance and therefore loss of bass response, giving a dynamic bandwidth shift.

absolutely this. this happens. and it's the one tonal characteristic I'm not crazy about.

edit: If I ultimately end up preferring the transformer route, this effect would be subdued if a larger bandwidth transformer is used, yes?

[Rob Strand]

If I ultimately end up preferring the transformer route, this effect would be subdued if a larger bandwidth transformer

To some degree, the Transformer bandwidth "in circuit" depends on the drive impedance and the load impedance of the actual circuit.

For the transformer:

https://projectivp.wordpress.com/category/schematics-manuals-documents/

Apparently the original transformer was marked
"8OTM018
HONGKONG"

or
"80TM018
HONGKONG"

The surprising thing is mouser have transformers:
42TM018   Xicon Audio Signal Transformer TM018 10K CT to 10K CT    (obsolete)
42TM018-RC   Xicon Audio Signal Transformer TM018 10K CT to 10K CT
10k to 10k with a centre tap
550 ohm and 500 ohm DCR

The part number similarity is too good to be true and the physical size of the Xicon transformer is virtually identical to the original.

There's also this ebay pic,
https://www.ebay.com/itm/Xicon-Audio-Signal-Transformer-TM018-10K-CT-to-10K-CT-Ring-Modulator-OK-/311035323932?rmvSB=true

[realkillercults]

yep. that's the one I've used.