Advice on calculating line impedance

[Mosaic]

Hi folks ,note this graph with a pulsed voltage (yellow) vs current (blue). The line inductance is clearly not saturated as the current is still rising by the time the voltage pulse stops.

Is there any way to derive line impedance (deltaV/deltaI) from this data? Or do I need to wait until the current saturates?

 

[ronv]

Hi folks ,note this graph with a pulsed voltage (yellow) vs current (blue). The line inductance is clearly not saturated as the current is still rising by the time the voltage pulse stops.

Is there any way to derive line impedance (deltaV/deltaI) from this data? Or do I need to wait until the current saturates?

View attachment 93674

It looks more like diode ringing. Is there a zener in there somewhere?

 

[Tony Stewart]

It looks like line inductance is 0.36uH assuming 10V step applied from Cap bank discharge. THis would appear to say the Cap bank is precharged to 13.2+10V.
Extended the duty cycle would appear not to saturate the inductance and the current may reach 800A steady state.
If the cap bank dropped to 7V above the battery rest voltage for 800A that would imply a battery ESR of 7/800=8.75mΩ

If peak current rises over time, that would indicate reduced ESR. Since the power Switch ESR is included, MOSFETs should be expected to rise in ESR with temperature.
MOSFET ESR ought to be much lower than Batt. ESR. I expect you can achieved ~1mΩ, which would be excellent.

Looks like a good sulphate sledgehammer.

What type of choke?

 

[Tony Stewart]

 

[Mosaic]

No choke, direct cap discharge via a quad set of FETs netting around 0.6 mΩ. The pulse starts out at about a 22V step. The blue current transformer trace runs @ 40A/V.

 

[Tony Stewart]

Looks perfect then, as I expected.

Litz Wire or many more parallel wires can reduce the inductance. Air inductors dont saturate.

 

[Mosaic]

Ok, thx.
I am using 4 x 12Awg x 1' paralleled hyper flex cabling per polarity now. The battery pole heats up quite a bit, more than the core.

Would you consider the voltage vs current wave to be in phase here? Given the ramping?

 

[Tony Stewart]

Ok, thx.
I am using 4 x 12Awg x 1' paralleled hyper flex cabling per polarity now. The battery pole heats up quite a bit, more than the core.

Would you consider the voltage vs current wave to be in phase here? Given the ramping?

By virtue of the inductive integration, the fourier harmonics of current will be shifted 90deg in phase. Therefore less reactive energy and more real harmonic energy can be transferred with a faster rise time. The concept is that some of the pulse harmonics are absorbed by the sulphate crystals and are ablated to clean the plates.

 

[Mosaic]

Would you think correcting the phase shift (adding a capacitor to the battery) for an improved power factor would be beneficial?

 

[Tony Stewart]

No that will bypass the current spike away from where you want it. Litz wire is the best way to reduce rise time for aggressive desulphation. It is the rise time that determines how much spectral density occurs to be absorbed by the lead sulphate crystals which in turn causes a piezo effect to vibrate and separate from the lead plates..

If the rise time is too slow, it does not excite the resonance of the crystals.

As I have said in past conversations, you are using a blow torch, which may work, but we also had an effective solution many years ago with a BIC lighter by comparison with Tr= 50ns pulses over DC vs your Tr=25us with 640A peak. We had the small unit wired directly between the battery posts.

P.S. the battery post and connector must be brushed very clean to achieve <1 mΩ contact resistance , otherwise it will be lossy and get warm in your case.