I have made some progress today. The negative voltage is not getting through the 1 M resistor feeding the grid of the EF83. When I short this out, the negative DC reduces the audio input voltage so the output voltage is reduced, but the gain remains roughly the same at about 20x. weird! The EF83 tube spec sheet mentions the source resistance should be less than or equal to 220k, but it is often more than this on many circuits as there is usually a resistor from the preceeding EF86, I am not sure how much this affects things.
best
DaveP