With the top cover on, the regulator heat sinks run quite a bit hotter. I'd estimate close to 60C, whereas they were only warm without. Since the dissipation per device is only 1.5W, thermal drop from junction to heatsink should be minimal, so I'm not too worried about it. Overall, the chassis remains only slightly warm and electrolytics aren't getting hot at all. I don't think there's any problem, but it's a reminder that there are tradeoffs packaging in a small chassis.
I had made some relevant simulations few years ago I found in my files. If you mean normal BAL connections as at the picture's left vs those at the picture's right (BAL_RR). There is also a summation amp on both types outputs to aid those sims so ignore it. Don't know, maybe the charts are helpful to you, although theoretical.
Attachments
Theoretical results look the same, don't know if there is some benefit with practical circuit halves never being 100% the same due to parts little tolerances. BAL_RR was on how to configure for best connecting an SE source to a BAL amp by the way.
...On the other hand linking the two phases circuits feedback return points with a resistor involves widened loop area vs local resistors to ground. Since its about two sizeable discrete preamps, not like with smd op-amp chips to could make a tiny block. Maybe increased chances of picking interference. You will see how it will work and what you will prefer in practice.
Tying the two R7 resistors together of each channel instead of their going to ground produced a hum in my balanced build. At the time Salas had mentioned to cut the common ground between the two halves of the board to see whether that solved the issue but I did not bother to try and reverted back to R7 to ground.
Is your shunt attenuator a 4 section unit or a two section with the shunt between the phases?
nash
