PI current pulse

[mickstv]

Here's another little circuit for anyone wanting to experiment with a flat current TX pulse as opposed to the normal sawtooth current pulse.

Changing the value of the 0.82 ohm resistor changes the max amount of current through the TC coil.

[Jose]

This is interesting, while I look for this scheme or something similar to experiment.
Thanks mickstv.

[mickstv]

Forgot to mention the transistor is a 2N3904 and the coil I'm using is low resistance 0.5 ohm.

[Tepco]

Also resistor of at least few hundred Ohm is needed between (low impedance) pulse source and MOSFET gate. Classic constant current source, interesting for experimentation, I tried something similar with small addition. 1K resistor between 2N3904 base and MOSFET source and another transistor (or analog switch) to “ground” 2N3904 base can switch circuit operation from constant current mode to normal linear ramp mode. Naturally pulse width need to be adjusted for given inductance and supply voltage to get desired flat part width in CC mode or desired peak current in standard mode, or to equalize them.

[mickstv]

Also resistor of at least few hundred Ohm is needed between (low impedance) pulse source and MOSFET gate

Hi Tepco, you don't need a resistor that high, I'm using a 10 ohm resistor between the gate and the TC4424 mosfet driver IC.

[Tepco]

This is OK, 2N3904 will withstand bit higher peak current during short pulse period, dissipation is low enough. But can be advantageous to use higher (few hundred R) value for MOSFET turn-on and CC period, and diode in parallel (cathode toward 4424) to speed up turn-off.

[mickstv]

This is OK, 2N3904 will withstand bit higher peak current during short pulse period, dissipation is low enough. But can be advantageous to use higher (few hundred R) value for MOSFET turn-on and CC period, and diode in parallel (cathode toward 4424) to speed up turn-off.

Just tried that mod but it actually caused the turn-off to be slightly slower.


I'm now running a 200us pulse with the same circuit (my first post) and 0.75 ohm resistor. With my ailing CRO set to 5us per div I can see the current switch off occurring in less than 0.5us, I would say thats quick enough.

[greylourie]

Hi mick,

With this circuit would the fly-back duration continue to be the same, even if something under the coil were to affect or alter the inductance of the coil ? Does it adapt ?

[mickstv]

Hi mick,

With this circuit would the fly-back duration continue to be the same, even if something under the coil were to affect or alter the inductance of the coil ? Does it adapt ?

No the purpose of this circuit was to control the maximum current to the coil.

[mickstv]

Here's a couple of CRO shots taken at 150us and 300us.

[Mechanic]

HI Mick,

Thanks for posting this. It looks like a very simple and effective way to make a CC pulse! Well done.

Cheers Mick

[mickstv]

I set up two samples one at 10us and the other at about 200us, and noticed that with smaller AU coins 5c 10c 20c 50c target response didn't change much between the 150us and 300us pulses. But when I tested $1 and $2 coins and also bigger targets like pliers lead weights, the response increase was higher for the 300us pulse.

Not sure whether this might lead to better differentiation between short and long TC targets.


I also need to check this against the typical response given by normal TX methods.

[mickstv]

Here's a picture of the gate drive for the TX mosfet, I'm using a lower value resistor .5 ohm at the moment.