MENR

menr1232

00:17

I can chat here now, if someone still wants to

DKNguyen

Hey

The effect you're seeing is more due to your unusual measurement setup than any fundamental property of inductance though I haven't entirely thought it through

menr1232

Hi, that is how I thought too at first. Maybe it's how I measured with the shunt. But the simulations show the same mechanic.

devnull

Note how high Vd gets

DKNguyen

For example, try running your simulation with the current sense resistor at the drain. You need subtraction or a differential probe to measure that but in simulation you can just get it to directly give you the current or the voltage across the current sense resistor

You're saying your sim gives that even without a resistor at the source terminal?

menr1232

@devnull Yes it reaches up to 700 volts in my setup. This I expected. Not the negative current.

@DKNguyen exactly

I can remove the shunt and it has the same behaviour

DKNguyen

00:29

Hmmmm. I'll have to think on that one.

menr1232

You can read the answer from gbg, especially with that analogy seems to make sense to me.

DKNguyen

No, that's basically what I Said. There's nothing in there about reversing currents.

UNless he's referring to some other effect

ANd when he talks about pressure that's voltage, not current.

menr1232

"So, now the coil is inside of the magnetic field that induces an opposite (with negative sign) current in the coil."

DKNguyen

THat's the only time he talks about a negative current but nothing else he says explains that

Because negative current would mean the water flowing backwards

not the pressure (equivalent to the voltage) reversing

So I don't get where he's coming from.

menr1232

Yeah this is confusing me again...

DKNguyen

00:36

I've never seen anything like that in the inductors I've worked on

menr1232

Yes, I'm wondering how this translates to something like a buck converter inductor

DKNguyen

In your simulation or circuit if you leave the transistor off, does the current actually reach zero and stay at zero?

menr1232

Because the whole premise is that the current wants to flow in the same direction and starts freewheeling.

DKNguyen

I guess your frequency is reall high and inductance is massive so it would take a while to reach zero.

menr1232

I can try it by decreasing the frequency in simulation

DKNguyen

00:40

What do you see if you use an inductance of something like 100u?

that works too

devnull

It I stop the pulses it keeps oscillating the current at 224 kHz

menr1232

I guess that's the series resonance of the inductance + filter capacitor

devnull

And if I connect a 100p capacitor between drain and source the osc. frequency changes to 137 kHz

I disconnected the 200u

menr1232

Ok that should then be the parasitic capacitance of the mosfet

DKNguyen

the MOSET capacitances too

yeah

menr1232

00:49

@DKNguyen I increased the inductance and indeed the current becomes 95% one sided

DKNguyen

one sided as in one direction?

menr1232

yes in the simulation it's negative but in reality you would count it as positive

DKNguyen

increase it from what to what? it's already at a massive 100mH in the simulation isn't it?

or 7mH, either way.

menr1232

yes from 7 to 100

DKNguyen

By the way, in case you haven't figured it out in LT SPice, current directions are predefined through components

even symmetrical ones like resitors

so you can flip the component

to get the proper current polarity through it, as indicated by the arrow when you mousover a component

and for voltage across a component you hold shift or alt, i cant remember which and click or drag between two terminals for the direction

devnull

00:53

Same here. But, when you stop the pulses it goes back oscillating initially with high peak-to-peak voltage

*with higher...

DKNguyen

Does that happen with a drain current sense resistor rather than a source follower?

devnull

No shunt anymore

DKNguyen

what if you toss on a drain resistor just as a load or current limiter or damper or whatever

I'm kind of surprised LTSPice will simulate it without a resistance

devnull

There is a 40mOhm resistance at the main voltage source

DKNguyen

oh, well severely underdamped in any case. heh

menr1232

00:56

There is also a series resistance of the inductance

parasitic*

devnull

Indeed. It was not marked as visible

menr1232

It seems one of the answers were removed

DKNguyen

I had asked GBG to clarify his answer

Not sure if he removed it in respnse to that

I mean he's sitting in this channel

but hasnt said anything

I'm going to have to step away pretty soon though but just as well. This computer is a USB stick so it can run LTSpice or anything

*cant

menr1232

All good, I appreciate your effort though :)

devnull

Ok. Thanks for the chat. It was an interesting problem

menr1232

01:04

I hope this topic doesn't die and we can get some more answers tommorow. Thank you guys.

Goodbye!

DKNguyen

01:34

@menr1232 I think it is your coupled inductor

I tried it with a regular 1uH inductor and what I see is a negative current spike distinct to reverse recovery behaviour similar to that on a diode but on a MOSFET .

which matches my initial suspicion but I had never seen a negative spike so large and long like the one in your sim before

well I reran it with your inductor but 10uH and it matches the normal induuctor now

its reverse recovery of something in the MOSFET. possibly the body diode unless the MOSFET itself has some similar behaviour

Transcript for