What's the function of the inverting amp in this mic pre?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
It appears to be part of the feedback loop, providing an inverted copy of the output signal to feed back to the second half of the input differential pair.
 
This is a variant "Cohen" topology. Peavey was already using this approach back when I started working there in the mid 80s.

The inverting op amp providing symmetrical feedback to the other input device delivers good balance using less active devices than the typical 3 op amp approach, with no performance compromise.

JR

Note : those 2sd786 very low noise input transistors went end of life obsolete near the turn of the century a couple decades ago.
 
This is a variant "Cohen" topology. Peavey was already using this approach back when I started working there in the mid 80s.

The inverting op amp providing symmetrical feedback to the other input device delivers good balance using less active devices than the typical 3 op amp approach, with no performance compromise.

JR

Note : those 2sd786 very low noise input transistors went end of life obsolete near the turn of the century a couple decades ago.
Thanks John.

So it's for improved CMRR?
 
I love this site, but how to get permission to post on certain pages? Sorry to hijack your thread, I just don't know what to do, I feel stuck.
 
I love this site, but how to get permission to post on certain pages? Sorry to hijack your thread, I just don't know what to do, I feel stuck.
you posted OK on this page...

Early new member posts get reviewed by the moderators for approval... if your post looked like spam (promoting products or a website) that might not meet approval.

JR
 
Actually, I was asking specifically about the function of the inverting amp, not the overall topology.
it is not working in isolation, it is part of the overall combined negative feedback loop. It mimics the other op amp with a symmetrical opposite polarity feedback signal. For good balance both inputs need to look the same to the outside world.
===
When I first saw that topology almost 40 years ago I didn't think it would be as good as the more common 3 op amp topology, but after review I became convinced that it was fully competent and one op amp cheaper.

My favorite kind of design. (y)

JR
 
I am familiar with the variant of that "the best capacitor is no capacitor"....

JR
I was very surprised to find inside the M-Audio DMP3, that there was no interstage caps beyond the one after the INA163. Not even on the final outputs. I didn't completely trace out the circuit, but I assume that 1/2 of each NJM4560 was configured as a DC servo?

I've never been able to find a schem for the DMP3.
 
Actually, I was asking specifically about the function of the inverting amp, not the overall topology.
As a complement to what John already said. You can think of it this way. Negative feedback has the property that it opposes the input signal. That is, a copy of the output signal is feed back in such a way that it opposes the input signal. So, the output of the differential amplifier goes into one op-amp, as it is, the output of that op-amp has the same polarity or is in-phase with the input at the base of transistor Q11, so by feeding the output of U3 (non-inverting) to the emmiter of Q11, when the input at the base of Q11 raises it voltage, the signal feed back at the emmiter also raises its voltage opposing any difference between the base-emmiter of Q11. If you only use that op-amp without the added inverting op-amp, then the 'cold' input at transistor Q10 does not get this same treatment, since the input of Q10 has the opposite polarity than the signal at the base of Q11, thus, the feedback is not balanced or symmetrical, that is, the output of the non-inverting op-amp will not oppose the signal at the base of Q10, this will increase distortion. By adding the inverting op-amp, it feeds a signal at the emmiter of Q10 with the same polarity of the cold signal at the input of Q10, hence opposing it. This way you have balanced feedback, that is, both transistors are getting the same treatment.
 
The SSM2016 (for example) used driven feedback to a single emitter. The opposing emitter was grounded through an identical value resistor. This maintains resistive balance (for CMR) but lacks signal balance. The opposing emitter, when driven by the inverted compliment, has both resistive and, as Duraflip points out, signal balance. This reduces even-order distortion.

The three op amp Cohen topology also requires a fourth stage if you want balanced output. The single or two op amp circuit realizes common mode rejection in the op amp stage (U3B in your posted circuit) which saves one op amp compared to Cohen. The output of U3A, which drives the opposing emitter, can also be used to provide a balanced output saving yet another op amp. In effect you get the benefit of four op amps using only two.
 
Last edited:
If anyone cares, I was a bit bored and decided to do some quick calculations:

The bias current of each transistor is very roughly 500uA.

The open-loop gain of each transistor is around100 V/V when the pot is at its minimum value, and around 1 V/V at its max value. This means that the differential output gain is 200 V/V and 2 V/V, respectively. This also means that the noise factor of the op-amps will be lowered by 200 when the gain is maximum.

The maximum closed-loop gain including the op-amps is roughly 1000 V/V or 60 dB, and the minimum gain is around 1.8 V/V or 4.8 dB. This is without counting the voltage divider effect due to the input impedance of the preamp and the output impedance of the mic. Considering these losses, the gain drops to around 930 V/V or 59 dB and 1.6 V/V or 4 dB.
 
Last edited:
Back
Top