That's a peculiar design, isn't it? Can someone explain to me what that CD4007 is doing there? Seems like only a single PMOS is in use. Why not use a discreet device then? Is separate access to the body diode the issue? There must be discreet devices that have that and are smaller than a 4007, no? I am not clear on what the function is here. Is is just a switch between the R and R/C inputs of the 4047? If so, wouldn't and old JFET do?
Also, what's with the 4049? If maximum current is of the essence, why no parallelize three per channel and kick the redundant single inverters before the pairs? Outputs 10 and 11 of the 4047 are already buffered, are they not?
Most perplexing design choices these are...
Cheers and sorry for highjacking,
Andy