joe said:
PI uses 100 cycles per second and VLF close to 1000 per second.
Very inaccurate information although not particularly harmful.
Current marketed VLF detectors operate at 2500 cycles per second to 71000 cycles per second and use a Sine wave to transmit signal. The coils typically have a much higher number of finer gauge windings than a PI coil and also must be balanced on an individual basis. On a DD this is done by moving the two coils in relation to each other and in a concentric coil it is done by adding a third small coil to balance the two coils against each other.
PI Detectors run maybe as low as 100 pulses per second and as high as ~3200 pulses per second (Whites TDI). I am not sure what the GPX runs at but I have seen figures as high as 5000 although I personally have not scoped my GPX. PI detectors transmit a square wave pulse and then wait a set period or multiple set periods to take a reading/s of the residual current in the coil (a coil will 'hold' onto a current when it is switched off). Coils on PI machines are much less sensitive to variances in inductance and balance in the case of DD coils. Whites 'Dual Field' coil, although it looks like a concentric is actually a dual mono.
I think it is possible to incorporate a Induction Balanced VLF detector and a PI detector into one machine and one coil but there is going to be a massive tradeoff. Lets put it this way, would you be happy if your new top of the line dual mode detector costing many thousands of dollars was about effective as a chinese knockoff VLF or as a 100us delay PI machine? Not to quote hard numbers but I think that would be a starting point for the crossover of efficiency.
That thread Joe leaves a funny aftertaste... a lot of it is very interestingly anti-minelabesque I'd be taking it with a large grain of salt.