ARMD (ARMRADIO based Metal Detector) VLF IB PROJECT

[Carl-NC]

Your sines do not look so good, try TC4420.

It's probably the result of using a series resonated H-bridge. Even to get it this good you have to carefully match the drive frequency to the LC resonance.

[pito]

This is what I have on output of Rx amplifier, 250 mVpp, gain 100, distortion = hardly to see the spikes



carefully match the drive frequency = for that purpose variable frequency generator is used.

[Carl-NC]

Might be the analog scope is helping it look better. I sometime switch to my analog scope when I don't want to see all those glitches.

[moodz]

Carl is right it is a series resonant CCPI bridge.

L1 and L2 form a pulse transformer.
However it is important to note the pulse width is set by the TX drive frequency ... NOT the LC combination the harmonic distortion is the main effect.
anyway you will see if you model it.

To get PI mode you close the switch across C2. Very simple.

Pito the glitches are far away in the spectrum from the principle sine wave. In the VLF mode we are using Frequency domain DSP and the glitches just dont have any affect ( gotta love the brick wall filter ).
You can see from the attached spectrum that the glitch energy is above 1 Mhz approx. The harmonics are sine distortions ( not due to glitches ) ... but we have DSP and those harmonics are not a bad thing. ( esp because we can adjust to have a multi frequency spectrum ).

As regards the voltage spikes on the TX waveform .. this is OK ... there are no current spikes and its all contained within the sheilding ... and they actually are necessary to get the efficiency.

Note the addition of the schottky diodes ... this works in real life ( body diodes of most mosfets are not schottky ) ... but ltspice does not model the body diode properly.

[pito]

This is what happened to my H-bridge.
1/2H, 12V/100ma, U@50cm = 0.5V
H, 12V/550mA, U@50cm = 1V
5 times more current 2 time more voltage.

I thought maybe my H-bridge wasn't good. I checked with the BTL amplifier = same story.

[moodz]

This is what happened to my H-bridge.
1/2H, 12V/100ma, U@50cm = 0.5V
H, 12V/550mA, U@50cm = 1V
5 times more current 2 time more voltage.

I thought maybe my H-bridge wasn't good. I checked with the BTL amplifier = same story.

You will have to decrypt this for us ?

[moodz]

For an exercise ... added a 7 - 10 Khz 5 stage cheby filter to the TX output coil L3.
Bandpass so we can still adjust the frequency.
This just demonstrates that it is compartively simple to clean the TX spectrum.

The filter would have to be bypassed also for PI mode.

[pito]

In those circuits Tx series resonance coil can be connected as 1/2 H bridge or as full H bridge. Simple field strength meter was used to compare how much energy is emitted from the coil at distance 50 cm.

[pito]

compare to single IC

[pito]

[moodz]

I think we are on different pages Pito ... the circuit I proposed is not a mosfet driver .... all the circuit elements ( ie mosfets ) are power switches - not drivers.

This implementation is an improvement on this patent https://patents.google.com/patent/US20230375738A1/en
I have added an output pulse transformer.
Additionaly I dont use a damping resistor and the body diodes (+ additional shottky diodes ) of the mosfets are utilised ( missing from the patent ).
The flyback is recycled to the next pulse .. via the C not the L. The L generates the flyback and the C conducts it.

Also the detail of the mosfet drivers is not shown ( pulse generators ) .. its a model from LTSPICE.

moodz

[Carl-NC]

pito , this thread will give you more background on this circuit:

https://www.geotech1.com/forums/foru...ve-to-h-bridge

[ivconic]

A possibly stupid question! In low power design; can we use such drivers for direct coil drive?

[pito]

@Carl-NC

Thanks, I see this design is not new, but there is no video of how it is performing, interesting is the very low voltage.

[ivconic]

@Carl-NC

Thanks, I see this design is not new, but there is no video of how it is performing, interesting is the very low voltage.

Bingo!
Unless Carl manages with his spare time and finish it.
Otherwise we will never see it materialzed.