Pulse Width of a Magnetic Field

[Nash78]

Hello, new kid on the block here.
I have read the rules and hereby present my question.

My aim is to calculate the period of a transient DC magnetic field from a half sine wave magnetiser output into an air core wirewound coil.

Coil Resistance : 14.66 mili ohms (derived using voltage drop by passing 1A of DC current into the coil)

Inductance : ~295 micro henries (measured via inductance meter)

Connecting cable size and lengths : 50mmsq and a total of 6meters (I assume that to be of negligible resistance and inductance for my equation)

Magnetiser resistance and inductance : No idea on how to get that, because the manufacturer does not supply.

Magnetiser capacitance : 25 mili farads

So, basically it is a R-L-C circuit in simplicity. With all the information i have on hand, my calculations come up to around 26 msecs.
But i am not confident of my results, as the formula i used, is to calculate the resonant frequency. This is something i am pretty unsure of, even after checking some technical papers online.

1) Is the resonant frequency = pulse width of the magnetic field? Confused... as the manufacturer of the magnetiser explains the output will be only a half-sine wave. I understand the word "resonant", as a oscillating wave which goes up/down until the amplitude disappears as energy is lost through each oscillation.

2) Can anyone guide me on how to measure the pulse width directly from the coil during the magnetic pulse? I have studied basic electrical engineering and understands my fluke multi-meter measures via high resistance inputs. But the reality of the magnetiser current displaying in excess of 10 kilo-amperes per pulse scares me in performing "live" measurements.

Kind souls, please guide me using simple explanations.:shy:

 

[vk6kro]

Using 25000 uF and 295 uH I get a resonant frequency of 58.61 Hz. This frequency has a period of 17 mSec.

However, the reactance of the components at resonance is only 0.11 ohms so you may not notice any resonance at all.

If it is a sinewave, you would get opposite magnetization on each half cycle of the sinewave assuming you used repeating cycles. But it would be roughly sinusoidal in shape, so the field would also be varying in amplitude. Each half cycle would have a period of 17 / 2 or 8.5 mSec.

25000 uF capacitors would invariably be electrolytic, so they may need some DC voltage on them to maintain the integrity of the capacitors.

To verify the magnetic field, you would probably use a Hall Probe. Several good Hall Probe chips are available with built-in amplifiers.

10000 AMPS! Are you sure about this? What are you doing?

.

 

[Nash78]

Using 25000 uF and 295 uH I get a resonant frequency of 58.61 Hz. This frequency has a period of 17 mSec.

However, the reactance of the components at resonance is only 0.11 ohms so you may not notice any resonance at all.

If it is a sinewave, you would get opposite magnetization on each half cycle of the sinewave assuming you used repeating cycles. But it would be roughly sinusoidal in shape, so the field would also be varying in amplitude. Each half cycle would have a period of 17 / 2 or 8.5 mSec.

25000 uF capacitors would invariably be electrolytic, so they may need some DC voltage on them to maintain the integrity of the capacitors.

To verify the magnetic field, you would probably use a Hall Probe. Several good Hall Probe chips are available with built-in amplifiers.

10000 AMPS! Are you sure about this? What are you doing?

.

Hmmm, yours and mine have different values of the resonant frequency.

We are getting a reading of actually 11,200 amperes from the display of the magnetiser.
This is only a transient current, not a steady state current. So the coil does not blow in my face when i pump the current in.

From my own understanding, this is how the thing works. The magnetiser "pumps" all the charged energy from the capacitors into the coil. The pulse width is limited by all the R + L + C of the entire setup. The manufacturer claims it is only ONE half sine wave, either positive of negative, depending on how we connect the output cables.

So it means all the energy has to be expended into the coil in that half sine cycle.

We reached this value of 11,200 amperes by testing the saturation of the magnet. The magnet only seems to hit saturation at this current and above. Currently, this is the only way of doing without any measurement instruments to determine the gauss value of the field.

 

[Bob S]

See thumbnail. I used 1000 volts. Max current in coil ~ 8,500 amps.

 

[Bob S]

Hi Nash-
This is a follow-up to my previous post.
First, in my thumbnail, the black curve is voltage, use vertical scale on left. The red curve is current, use vertical scale on right.
The calculated resonant frequency is (1/2 pi)*(1/sqrt(LC)) = 58.6 Hz. The calculated half-cycle time is 8.5 millisec, about what is seen in my posted thumbnail.
Make sure you have included all losses, such as capacitor internal series resistance. You might have to look at the capacitor itself, get mfgr and type, and look on internet.
Is the capacitor unipolar or bipolar? Does it have a reverse diode on it to block reverse voltages such as seen in my thumbnail. Lastly, note that the second reverse polarity pulsed current peak (at about 16 msec) will reverse magnetize your sample. Are you concerned about reverse polarity magnetization, and whether it might demagnetize your sample?
Bob S

 

[Nash78]

Hi Nash-
This is a follow-up to my previous post.
First, in my thumbnail, the black curve is voltage, use vertical scale on left. The red curve is current, use vertical scale on right.
The calculated resonant frequency is (1/2 pi)*(1/sqrt(LC)) = 58.6 Hz. The calculated half-cycle time is 8.5 millisec, about what is seen in my posted thumbnail.
Make sure you have included all losses, such as capacitor internal series resistance. You might have to look at the capacitor itself, get mfgr and type, and look on internet.
Is the capacitor unipolar or bipolar? Does it have a reverse diode on it to block reverse voltages such as seen in my thumbnail. Lastly, note that the second reverse polarity pulsed current peak (at about 16 msec) will reverse magnetize your sample. Are you concerned about reverse polarity magnetization, and whether it might demagnetize your sample?
Bob S

Hi Bob, i appreciate the follow-up. I have done quite some homework before on a magnetizer, especially what the manufacturer has told me. We have on hand a Sine-Half Wave Magnetizer, which typically produces only a half sine wave magnetic pulse to any connected coil.

The magnetiser internal inductance and resistance is unknown. This information is not disclosed to us nor will it ever be ( i think...). I have absolutely no idea on how to perform my own measurements to get them.

From the schematics, there is a diode in series with a scr, connected to one of the outputs. So the answer to your question, it should be a YES, there is a diode to block the reverse polarity. The capacitors are bipolar and there is NO label on the caps, so no way to know who manufactured them too.

I just hope that i am not going in the wrong direction

 

[Bob S]

Hi Nash-
I put a generic diode in, and it seemed to work. See thumbnail. Black is coil voltage, left scale. Red is current, right scale. Blue is capacitor voltage, left scale.
NOTE that there is voltage left on the capacitor after the pulse.
Bob S

 

[Nash78]

Dear Bob,

Your example makes perfect sense.

SO all the calculations have been right! Maybe i should have better confidence in what i am doing...

All i need to do is to get the resonant frequency and divide it by 2.
Amplitude just depends on the voltage upon the pulse into the coil, that is for the later part which i will try to calculate how much heat my coil dissipates at each pulse.

I greatly appreciate your help! And in return, i will post my later findings for sharing with the rest of the forum members.

PS: Is your software freeware or those expensive ones?

Rgds
Nash

 

[Bob S]

Hi Nask-
1) To get the pulse width of half cycle, calculate resonant frequency, get reciprocal to get period, then divide period by two.

2) To estimate the peak coil current, equate stored energy in capacitor and in coil:

(1/2)CV² = (1/2)LI²
Using C=.025 F, V=1000, and L = 295 uH. I get 9205 amps peak.

3) Software is FREE! I am using LTspice IV from www.linear.com.

Bob S

 

[Nash78]

Hi Bob,

Im beginning to use LT Spice, is there any transformer and bridge rectifier component somewhere? I can't seem to find them.

Thanks to you, I'm completing my work soon.
Really appreciate the help.

Rgds,
Nash

 

[Bob S]

High Nash-
For a bridge rectifier, I would use 4 generic diodes, or you can choose a specific diode from a list. I have not tried it, but for transformers, you use a spice directive and two inductors (with the polarity circle) back to back. The relative inductance is the turns ratio squared. See HELP in the top toolbar, then type "transformer" in the search box.
Bob S

[Edit] I added the attached thumbnail jpg showing a 10:1 (120 volt to 12 volt) transformer using a spice directive (under Edit in top LTspice toolbar).
Bob S