New Version of Bipolar PI MD with GB and DD Coil

[Detectorist#1]

Hi all PI MD enthusiasts,
Now, I propose my last seriously improved version of this project. The main improving are addition of 100% reliable Fe recognition when "Fe" is reported and adding of pseudo "static" mode very useful for serious detecting. These improving are very useful for reducing of the digging efforts and for more easy pin-pointing of the targets. I hope, this project will be very popular among DIY people in the near future.
The schemes of the "Main Board" and "Power Board" of this project are attached.

[Detectorist#1]

This version has the ability to further reducing of the first delay below 5us. In the coming days I will test where the reasonable limit is.

[Nightryder]

I hope it will get the attention it deserves. Thanks for your time and effort!

[Detectorist#1]

Hi Nightyder,
All my efforts in last 3 years are pointed to help for DIY people in this exciting hobby!

[Detectorist#1]

It's time to dethrone the "Delta Pulse" project from the top spot in DIY PI projects!

[algan]

It's time to dethrone the "Delta Pulse" project from the top spot in DIY PI projects!

Delta pulse works in static mode, suggest a static mode detector here, does your project work in the same mode, is it suitable for coils with a large frame 1meter x 1meter?

[Detectorist#1]

Hi algan,
Delta pulse project works well only with big targets. When you try to reduce the first delay and the sample's duration, this project is not capable to work in real ( or near to real) static mode. Also, the pretentions for ground balance possibilities are very doubts. The case with PI MD with 1m x 1m mono search coil is totally different from 15" x 12" or 12" x10" DD coil. No universal solution for all soil's conditions and for all target's sizes. This is the reason for existing of so much different schematics of metal detectors.

[eduardo1979]

Hola Detectorista #1. Felicitaciones por tu trabajo y el gran esfuerzo que estás haciendo para aportar cosas interesantes a este foro.
Me gustaría experimentar con tu proyecto, así que tengo algunas preguntas: ¿
Funcionaría tu proyecto con una bobina Minelab Mono original? No quiero discriminación de metales ferrosos porque sé que tendría que construir una bobina casera con los mismos valores para TX y RX, y también sé que el campo magnético terrestre afectaría negativamente la estabilidad del detector en modo de discriminación de metales ferrosos. Quiero construir un detector para encontrar pequeñas pepitas de oro, con buena cancelación de efectos de campo electromagnético (EFE) y buen balance de tierra. Sé que los detectores bipolares cancelan bien el campo magnético terrestre. Por ahora, estoy rediseñando tu proyecto para adaptarlo a mis necesidades. Me gusta mucho la topología que Eric Foster usó en sus proyectos. Es la abrazadera activa, formada por el CD4066 y el amplificador operacional para el reinicio automático de los dos canales SAT. Para la etapa de audio, me gusta usar el CD4046 y amplificadores de valor absoluto para el circuito de umbral, que constan de un amplificador inversor y otro no inversor. Mi única duda ahora es dónde conectar la entrada no inversora, el pin 2 del preamplificador, el que está conectado a la bobina RX, ya que voy a usar una bobina mono.

[Detectorist#1]

Hi algan and eduardo,
No problem for working with MONO coil with any size. Pins 2 and 4 of the coil connector have to be connected together. The resistor R18 have to be removed and the values of R24 and R25 have to be corrected according the parameters of the coil. In this case, the "Fe" mode will not work and the minimal delay will be higher. Eventually, it will be necessary to correct the duration of the TX pulse to ensure that the peek voltages on the coil are under 450V.

[Detectorist#1]

Hi eduardo,
For your goal to build PI MD for small gold nuggets, you need to follow the following steps:
- if the search coil is MONO type, the size have to be relatively small - (5" to 8")
- the inductivity of the coil have to be also relatively small - (75-100-150) uH. This will allows more fast sampling delay (some more complex technics allows low delays with coils with higher inductivity)
- follow all requirements for making of a fast coil ( Teflon isolated wire, two layers of PVC spiral wrap and scotch 24 screen)
- use flexible coaxial cable with 40-50pF capacitance per meter for connecting of the coil to the control box.
- use max TX current, but the peek voltages on the coil have to be under 500V (limited by the max Uds of P-ch MOSFET)
- use metal box for control unit with good, sure electrical connections of all parts of the box to GND plane of the control board. The cases of the external connectors also have to be grounded.
- bipolar TX pulses and possibility to switch-off of GB channel are obligatory. Bipolar pulses reduces the requirements for low frequency noise of the first OpAmp in RX channel and eliminate the need of EFE pulses.
- if all this not satisfy your requirements, you need to use balanced DD of OO coil (hard task for DIY people) for reducing of the value of the sampling delay in the target channel

[pito]

+/- 260 V , its that means you have 260V if you are using one MOSFET ?

[Detectorist#1]

Hi pito,
Bipolar sequentially pulses exists only with two MOSFETs - N-ch and P-ch

[pito]

I mean you have +260V from one MOSFET and -260V from second MOSFET, its that correct ?

[Detectorist#1]

Yes, two peek voltages are generated from two MOSFETs. And adding of C17 (330pF) allows the starting of the TX coil peek voltages with equal delay after the end of the TX pulses. The capacitances of N-ch MOSFET are 3 times lower and this capacitor add some delay in switch-off moment for N-ch MOSFET for equalizing of this time for two MOSFETs.

[pito]

is that so?