Thanks, the first thing I did was check they are mechanically set up the same and they are as spot on as can be, so I don't think that is the cause.

They are run separately on different channels and mixed on the tx, hence I asked whether it could possibly be a software fault, where the slave servo is running slightly behind the master servo in the tx software processing etc. ?

I know some radios like the 9ZAP has the facility to allow the servo speeds to be altered to fix problems like this, but I don't think my 9CAP (FF9) does. Maybe I could try a bit of exponetial on the faster servo to allow the slower one to catch up.