Wow, that's an old preamp initially released around 1977.
First, do you have anything connected to the preamp inputs? If so, disconnect them.
The Mark Levinson ML1 could be a difficult preamp to service because it looks like important electronic circuity are hidden inside of modules. Looks like the modules contain op-amps made up of discrete components (rather than 'chips'). They might be encapsulated with a potting compound to prevent users from getting inside.... but I don't know that for fact. Can those modules be easily opened or are they sealed?
According to the schematic there appears to be a switch somewhere that selects direct-coupled (DC) output or AC-coupled through a capacitor. It could be the toggle switch on the PCB shown in the photo but I'm not sure. There also appears to be DC-offset trim-pots for each (Left and Right) line-level amplification module. You can try adjusting the Left channel trim-pot if you can find it while monitoring the DC off-set. Make sure to put the trim-pot back to its original position if it has no effect.
You can also try finding the switch for AC / DC coupling. AC coupling would eliminate the 0.7V DC offset at the output (assuming the cap isn't shorted) but something is still not correct if only the Left channel has that much offset. Based on the 0.7V off-set I would guess that a silicon component inside (maybe one of the bipolar transistors) is defective as 0.7V is the forward voltage drop of most PN junction silicon devices.
On a related note, you can characterize (take measurements) for the good channel and compare against the bad channel to see where the problem pops up. But the challenge might be getting at the components within the suspect module.
You can download the PDF schematic below. I found it online and put it on my site. Good luck!
Mark Levinson inside photo:
