Why is the PI front-end and clamping designed like this?

[CrizzyD]

Hi all,
I've inherited a PI circuit that uses a high-side switching topology for pulsing using a P-channel MOSFET. What I'm not so sure about is the immediate front-end circuit before the clamping diodes:



I get that C23 and R33 are for a low pass filter, but what is the purpose of R34 and R32? Won't the entire behaviour of this part of the circuit be dependent upon whether the clamping dioes are conducting?

[Carl-NC]

Assuming J18 connects to an RX coil, looks to me like R34 is a damping resistor and R32+C23 form a LPF @ 1.6MHz. The LPF might help tamp down the peak flyback voltage or it may be for EMI reduction. When the clamp diodes conduct they alter this a little but not a whole lot.

[CrizzyD]

J18 connects to a receive coil of about 2000uH inductance. The coil topology of this system is separate transmit and receive coils. Carl-NC thanks for the advice, what about R33?

[Carl-NC]

J18 connects to a receive coil of about 2000uH inductance. The coil topology of this system is separate transmit and receive coils.

That was my assumption. R32+R33 is the clamp resistor.

[detectormods]

I think that a better way to retain sensitivity and utilise the benefits of low noise components is to use the dual series depletion mode mosfet circuit or use the active N series mosfet switched circuit that Minelab uses, Less loss , less noise more signal. IMHO.

of course there are better ways but that is in an egg I’m hatching.

[Auto-Mation-Assist]

Capacitor C23 will cause a large DC offset on the output of the first stage. The best thing to do is to minimize capacitance here and not add to it. The capacitor will be charged by the recoil voltage and take a substantial amount of time to discharge. This is a really bad place to place a capacitor.

There will always be some voltage there that will tend to charge the cap to some degree especially in a PI detector. I think that R34 may be part of the TX coil load if it is not a DD only input stage.

[Auto-Mation-Assist]

I think that a better way to retain sensitivity and utilise the benefits of low noise components is to use the dual series depletion mode mosfet circuit or use the active N series mosfet switched circuit that Minelab uses, Less loss , less noise more signal. IMHO.

of course there are better ways but that is in an egg I’m hatching.

This is true, but at times there are drawbacks to that method also. My primary concern with those methods is the generation of inter-mod and the possibility that they have a greater potential of being more susceptible of interference due to the active switching that takes place. When using strictly resistors there is no switching but you do pick up more resistors noise and loose only a very small amount of signal if the input of the amp following has high impedance. The resistor noise will likely dominate when following amp is properly selected and used, thus high quality resistors need to be used in this location.

Any kind of switching at this input point will inject some kind of residue into the first gain stage, and each method that is designed to protect the input amp from high recoil voltage has its own issues. My own circuit that uses only resistors has something like 0.02 db loss at 200Khz.

[detectormods]

This is true, but at times there are drawbacks to that method also. My primary concern with those methods is the generation of inter-mod and the possibility that they have a greater potential of being more susceptible of interference due to the active switching that takes place. When using strictly resistors there is no switching but you do pick up more resistors noise and loose only a very small amount of signal if the input of the amp following has high impedance. The resistor noise will likely dominate when following amp is properly selected and used, thus high quality resistors need to be used in this location.

Any kind of switching at this input point will inject some kind of residue into the first gain stage, and each method that is designed to protect the input amp from high recoil voltage has its own issues. My own circuit that uses only resistors has something like 0.02 db loss at 200Khz.

I came up with a better way to get the noise out from the system, but overall the enemy is resistance, capacitance, intermod in the mosfets, rectification via diodes and diode junctions, self induced intermod, mux leakage, plus many other issues, i cured most internal and external issues in the P.I that i reworked.

https://youtu.be/3aZjqUxoIYs

[Orbit]

What medicine did you use?

[Detectorist#1]

Still no answer what medicine use detectormods for curing of these issues. Top secret maybe?

[Detectorist#1]

Minelab SD2000 uses switched depletion transistors in the front of the first RX stage but this MD is not for very small nuggets. AMA is right - "Any kind of switching at this input point will inject some kind of residue into the first gain stage"

[moodz]

Still no answer what medicine use detectormods for curing of these issues. Top secret maybe?

Hot glue ??