Some more refinements, I have a GPX-4500 sitting on the work bench at full gain running a mono coil and the detector is just about dead silent except for a very small amount of 50 hz EMI from mains electric cables. small gold performance is exceptional and large deep target response is way beyond anything that i have seen before. I was thinking that removing the high frequency signal component would screw up the ground balance but this is not the case, ground balance is exceptional. this detector upgrade will be the one to beat. As a side note to anyone experimenting with detector input circuit design, any type of diode in the receiver path used as a a clamp or speed up setup on an opamp can cause serious intermodulation from mixing unwanted signals with the dv/dt of the input switching, once it occurs it is near impossible to remove, some soft knee fast recovery diodes can be substituted but its not the be all end all cure, dc servo steering of response time and also keeping dc offset and drift to tight specifications using chopper stabilised servo control has merits, just need to be careful in filtering out any feed through clock noise such that can be observed in the LT choppers in some detectors that i have inspected, a simple 10 ohm resistor and 3300pf capacitor removes most of the noise at some small expense to THD and common mode rejection. When dealing with nanovolt signal levels there is now scope for greater improvement in the signal to noise levels, the limiting factor is still in the input series Fets that can gate modulate the the input signal and resistive and capacitive losses and stray coupling on the input circuit. All these improvements will get more gold, that is 100% for sure.
