Time to design a all-round 5V metal detector?

[6666]

I'm sharing my solutions with you, people. Whining doesn't help anyone.

Goodbye.

[bklein]

Basically go 3.3V with everything but the TX drive circuit. I don't think a 5V coil drive circuit will compare to a higher voltage one. Take any existing design and substitute your TX circuit with a 5V supply. Even if you get close, 5V Vcc is going to cost you extra or not perform as well. But go for it and prove me wrong.

[Dave J.]

As I see it, it's almost all +12V.

Dave Johnson says if that's what you see, you need to take another look.

[jladre]

Seems it works..the ace 250 can work on 4x 1.2 rechargeable nimh bat. So 4x1.2=4.8volts. anyway i have no idea on ace circuitry but power supply definetely work on around 5v.

[Teleno]

Take any existing design and substitute your TX circuit with a 5V supply. Even if you get close, 5V Vcc is going to cost you extra or not perform as well. But go for it and prove me wrong.

In principle the only difference with a higher voltage Tx circuit is a longer duration of the ramp (I = V/L * t). This can mislead into thinking that the 5V Tx circuit consumes more power, but this is wrong. The amount of electrical work to reach a given I at the coil and maitain it goes up with the square of V, therefore a 5V Tx driver is much more effcicient.

[6666]

Dave Johnson says if that's what you see, you need to take another look.

THANK YOU DaveJ

[Teleno]

Dave Johnson says if that's what you see, you need to take another look.

Your Tx circuit looks like a sine wave generator to me. What's the current waveform in the coil?

[Dave J.]

voltage is rectangular and current is triangular. the patent is probably kicking around on Geotech someplace, it shows the waveform.

are you looking at the wrong schematic someplace? what I'm talking about is the Fisher Impulse.

[Teleno]

voltage is rectangular and current is triangular.

That's what I thought.

My design is a totally different approach: constant current drive and sampling during on-time by an IB coil. Best for maximizing both short and long decays (1us < Tau < 1ms).

The following is a traditional (not CC) driver adapted to 5v by use of a depletion type MOSFET, which considerably reduces the component count. A uC output pin can drive this directly.

[rickb]

The posted pdf of the Fisher Impulse is a mix of 12v, 6v, and 2.5v, i.e. you're both wrong.

[Dave J.]

Audio runs on unregulated 12. Everything else runs on 6. 2.5 is just the analog ground.
A fairly similar circuit in induction balance configuration demodulated on and flyback time and added VLF type operation to pulse induction. That however was never manufactured.

[moodz]

In principle the only difference with a higher voltage Tx circuit is a longer duration of the ramp (I = V/L * t). This can mislead into thinking that the 5V Tx circuit consumes more power, but this is wrong. The amount of electrical work to reach a given I at the coil and maitain it goes up with the square of V, therefore a 5V Tx driver is much more effcicient.

Its quite possible to effectively remove the time factor from the I = V/L * t equation above .... by recycling most of the coil energy its possible to get very large coil currents with a couple of volts .... 5 volts is almost overkill.

See attached ltspice :-) ... remove the .txt to use ltspice.



Coil current reaches approx 4 amps with 5 volt supply , 300uh , 1 ohm , 330 volt peak flyback and sampling under 10 us.

[moodz]

Correction ...Coil current reaches approx 4 amps with A 3.3 volt supply ... 5 volts is too much

[Teleno]

Hi moodz,

I'm not going to publish my PI circuit, but indeed I'm using an energy recovery scheme too. Something like the Vallon PI does but with a flat top for sampling on-time (Vallon causes a ramp top).

[moodz]

....so you are sampling during TxON ;-) .... ...the current pulse rise time is fast enough ... but your pulse fall time looks a tad slow and overshot ...

BTW have you seen this patent ... http://www.geotech1.com/pages/geo/patents/US4157579.pdf