My Balanced Line Receiver

Unfortunately I have no boards left from the group buy. However, I just designed a new board for the same circuit:
IMG_0584 small.png
After building one and verifying its performance, I have four spare boards of this design. PM me if interested.

The differences from the earlier board are:
  • Input is now protected by a quad of diodes as suggested above by @Mark Tillotson
  • "Normal" size (D=2.5mm L=7mm axial) resistors are used throughout
  • More room for input, output and power connectors - Molex KK254 or similar-sized connectors can be used
  • Slightly larger board, 50×50mm size (one channel)
The performance is the same as before. The distortion is determined by the opamp used and is low. Here are some measurements with a pair of LM4562. Note the test signal is attenuated after the board and before the ADC to reduce the ADC's own distortion and reveal low level distortion products:
1kHz THD 600 ohm load.png
19+2-kHz IMD 600 ohm load.png

The common mode rejection is very much defined by the matching of the resistors. The results shown below are for 0.1% CMF55 resistors (the brown ones in the photo above):
Common Mode 10ohm Mismatched Impedances.png
Common Mode 600ohm Mismatched Impedances.png
Common Mode Matched Impedances.png

The ground sense reduces the difference between the local ground and that of the downstream device by about 40dB:
Ground ssnse.png

The board if relatively wideband:
Differential Mode.png
 
I was thinking a PCB that would mount by directly soldering to the PCB pins of the Neutrik jack, to save on wiring efforts as well as drilling/tapping/mounting on a sheet or backplate/baseplate. As far as universality goes, Neutrik jacks should be available pretty much anywhere in the world and I don't see any reason they couldn't be shopped for any application.

https://www.audiosciencereview.com/...d-measurements-of-hifiberry-dac-pro-xlr.8569/

https://shop.klotz-ais.com/mis-tx4mk-1.html

These examples above have multiple jacks per board but obviously not a requirement by any means.
 
In terms of THD, I got the best results with OPA2156 and/or OPA1656:
 
  • Thank You
Reactions: alexcp
The distance between the holes of the IC and those of the power connector is 150 mil. Your connector needs 125 mil one way and 100 mil, the other, so technically, it will fit. However, two things to keep in mind: (1) if you use an IC socket, that will take away 50 mil from those 150; and (2) the connector's plug may hang out beyond the 100 mil taken by header. The MT-100 IDC plugs look slim on the drawing, but I have no experience with these connectors and can't confirm positively.

The right-angle header should fit, but obviously would need some space outside of the board.

If you want slim polarised headers, something like Molex SL series may be a good fit,
 
That's a cute board, but it does not include a difference amplifier. The two halves of the dual opamp buffer and amplify the differential signal, but it remains differential. One would need to add an opamp with four matched resistors, or an INA134 or similar, at the output.

Also, in terms of layout, the area enclosed by each differential pair should be minimized. A mirror-image layout is exactly the wrong way to do it.