MPP

For my taste the Cohen is rather complex. Also we want to beat the 1nV/qHz barrier, therefore the discrete input transistors. Some time ago Frans and me have shown what we coined the Dolfin.
Here up to 8 Opamps are in parallel at the input. That brakes the barrier too but is less fun.

The complexity of Cohen is purely the 'double balanced' OP stage. It has the
simplest front end using a pair of discrete BJT's biased with opamps FB resistor.
Just a/ add your fav ZTX IP BJT b/ insert passive RIAA c/ use whatever OP stage you desire.

TCD
 
For my taste the Cohen is rather complex. Also we want to beat the 1nV/qHz barrier, therefore the discrete input transistors. Some time ago Frans and me have shown what we coined the Dolfin.
Here up to 8 Opamps are in parallel at the input. That brakes the barrier too but is less fun.

Cohen is with discrete input transitors.
With the ZTX 851/951 you can go as low as 0.3nV/rtHz.
And with SE output all you need are 2 trannies and 3 op-amps :D

@Terry
Thx for the original Cohen paper, nice to see a face and get some insight in his background

Hans
 
Last edited:
Member
Joined 2006
Paid Member
I have used AD8429 in this balanced input phono: Chipomatic balanced input RIAA
Its included in the LTspice also because Analog Devices acquired Linear Technology.

Not a known type for audio applications but its a low voltage noise INA and you could use it for your simulations.

Hello Salas
Long time no talk :)

Thank you for the tip.

In this sim I am using the AD8429 with a servo that I picked from the Chipomatic :):)

The previous sim with the That1510 did not include the servo as it seems to output near 3mV offset with ref to GND....

Now I need to simplify the circuit as U4 and U5 seem intrusive
 

Attachments

  • DOUBLE BALANCED MCPREAMP ANTIRIAA RESPONSE AD8429 EVO6.asc
    8.4 KB · Views: 40
  • DBAL AD8429.pdf
    105.4 KB · Views: 113
I have used AD8429 in this balanced input phono: Chipomatic balanced input RIAA
Its included in the LTspice also because Analog Devices acquired Linear Technology.

Not a known type for audio applications but its a low voltage noise INA and you could use it for your simulations.

This chip looks pretty impressive. Very low distortion especially at high gains
and HF.

Do you *really need much less than 1nV/rt Hz?

TCD
 
diyAudio Chief Moderator
Joined 2002
Paid Member
This chip looks pretty impressive. Very low distortion especially at high gains
and HF.

Do you *really need much less than 1nV/rt Hz?

TCD

That 0.0073% measured THD+N figure translates to 83dB SNR. No vinyl surface quietness or TT wiring hum rejection can challenge such capability in a 62dB gain MC phono. Regardless of it offering balanced input impedance interconnection ability, hum never fully cancels with turntables.
So no, I don't need less nVrtHz INA chip for this design combination. SUT's noise dominates. Thus I have looked to marry the specific SUT's Zo with the specific INA's input current noise well instead.
 
Look here for a similar Cohen design, using THAT1240 and very low noise input transistors.
A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851 - Page 49 - Pro Audio Design Forum

Hans

Thanks for posting the link Hans.

There's more complete documentation on the MC preamp here: Flat Balanced Input Moving Coil Phono Preamp Construction Information - Pro Audio Design Forum

I'm not sure why the Cohen topology's double-balanced output is perceived as complex. It's only two diff amps with the inputs connected in anti-polarity. In addition to reducing noise they provide common mode rejection at DC eliminating coupling capacitors to strip off the first stage's DC.

This is not a "head amp" it's designed for direct conversion or "line level" RIAA EQ. Many of my clients mount this on their cutting lathe and then drive long cables with the balanced outputs.

The third stage in the schematic, not attributable to Cohen, is a "Differential DeBoo" servo. It's single film cap is the only capacitor in the signal path other than compensation and bypass. Being subtractive the servo forms a 6 dB/octave HP response.

The ZTX-series allow sub nV√Hz performance quite easily.

Flat_Balanced_Input_Moving_Coil_Phono_Preamp_ZTX851_FLIP_22_Schematic.png
 
Last edited:
So the signal @ the input must allready be equalized as per riaa curve right ?

RIAA EQ is applied after the preamp either in analog (with a companion board) or DSP. I originally designed it and an MM version for RAW capture to de-click pre-RIAA and then apply RIAA in DSP (using a .vst) to render the file. Some of my clients apply DSP in realtime.

Also the cap in the servo acts as rumble filter as it is a high pass filter...

Correct. It's about 10-12 Hz and due to the servo control point does not vary with gain.

Thank you very much for the insight regarding the opamps connection.

You're welcome!

The Differential DeBoo servo that I derived from the single-ended version I've only seen used in one other place: Deboo Differential DC Servo for AD851 HPA: アナログ回路のおもちゃ箱

I think I started using it around 2012 or so in a mic preamp and then found it had been used in a headphone amp. It's a useful little circuit because it's essentially an INA and has a common mode gain of "1." The jumper shunting the cap is to turn it off to check input offset.