3316601/ You need to click on magnifier icon glass to make it readable. The only transformer load is old Tektronix P6015 HV probe (Cin=3D3pF). There is no sparking/breakdowns (as far as I can see/hear). Primary voltage is measured (Tektronix P6139A) across =93auxiliary=94 one turn winding placed next to the primary winding. Transformer primary/secondary voltage is red/magenta trace, left Y- axis. Secondary voltage is divided by fifty (turns ratio). Questions I have:
I assume that (very fast) negative voltage transition corresponds to core saturation. It makes sense at 0.19us mark. Does negative overshoot correspond to ringing with mu=3D1? What happens at 0.36us mark?
Huge current spike at 0.19us mark - ??? Where does this current flow?? Not in 3pF of scope probe... How come current polarity changes??
How do you know that the voltage out of the current probe is=20 actually related to current, and not a capacitively coupled=20 signal related to voltage swing on the conductor through the=20 pickup?
Is the current probe on the hot or ground side of the winding?
You don't indicate what is happening at the fet switch. Can we assume it is turned on, and left on? It doesnt seem to be very well controlled, given the slow rise of voltage on the monitoring winding. Why not measure the primary directly?
Where is the current transformer. Is this fet/capacitor current? For one-shot waveforms, you should be able to establish a pretty reliable
0A base-line on your display.
Although you've saturated the core in one direction, it is only maintained in saturation in the presence of unidirectional forces. As your circuit rings (series cap and stray parallel capacity - leakage and nonlin inductance terms), the desaturating inductance can develop voltage on waveform reversal. The inductance varies dramatically, as is evidenced in the varying fundamental frequency of the scoped voltage and current waveforms.
I'd be interested in viewing the switch gate and drain waveforms, to see if the switch wasn't actually contributing to the nonlinearity, thanks to a high-z gate drive.
The indicated current appears to be a reasonably accurate representation of the rate of change of the voltage. It appears the primary dv/dt is around 200V/25nsec, and the secondary is (if I've understood what you've written correctly) something around 400-500V/ nanosecond. It wouldn't take much to couple that capacitively into things in the general area. I'd look for a measurement problem... but then I also don't know from your base posting in this thread just how you are driving the primary. What's the primary inductance? What's the primary current as a function of time?
Hint about core saturation: you should be able to see that with no secondary on the core. If you put a step voltage into an inductor with zero initial current, you should see a linear ramp of current starting at zero, if the inductance stays constant and there's no resistance. Resistance causes the current to follow an "exponential" just like voltage across a capacitor charges through a resistor, with a time constant of L/R. Saturation, of course, causes the inductance to drop and the di/dt to therefore rise. So if you monitor current while applying a voltage pulse of adjustable width, you can start with the pulse narrow and adjust it till the current gets as big as you care to test to. Be sure, of course, to let the core re-set between pulses; beware of flyback voltages when you turn the current off.
If you don't know much about the ferrite of the core you have, you can at least determine its permeability from the inductance of a coil wound on it and its dimensions, and from that you can at least make an educated guess about its saturation characteristics, from published data on typical ferrites. A measurement should be able to tell you if your guess was about right.
The current transformer is on the return side of the (sic!) secondary. The only load is oscilloscope probe (and whatever winding parasitics are (more about it below).
What is desaturating inductance? The core will desaturate when H get into "hysteretic" region of B-H curve, won't it?
The switch is IRFR12N25D driven by transformer-coupled UCC37322. The switch stays ON longer than above mentioned 0.22uF capacitor driving transformer-in-question primary winding gets discharged.
... OK, I didn't say to get 40A. The implication to me is that it MAY be finding another way into your measurement. You have (apparently) a lot of secondary voltage, happening very quickly. Again, to a pretty good approximation, it appears that the indicated current is proportional to the derivative of the voltage, and in that I suspect lies an important clue.
If you disconnect the HV probe and the primary voltage probe and have only the current monitor, do you get the same indicated waveform?
Yes I do. Kind of... Current spike becomes smaller by factor of two. It indicates that it's parasitics...20 pF cause 40 amps current... It means 2 kilovolts per nanosecond dV/dt. I measure 0.8kV/ns with
500MHz probe and 500MHz oscilloscope... Capacitance meter sees ~6pF across the HV probe. The current drops by ~50% when HV probe is removed. It hints ~12pF total parasitic capacitance - I have to cogitate on it. It still does not explain
why I get such a huge dV/dt value when core saturates
You mean to tell us that your current transformer, in fact, isn't?
A current transformer proper is inseperable from its burden resistor. Without a burden resistor it is a little series inductor and utterly useless for current measurement.
Tim
--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
From the first wave form it looks like you have exceeded the volt seconds of the primary. Then the core saturated, the current went nuts and more unwanted things happen.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.