[Prev][Next][Index][Thread]

Re: Ignition coil measurements




>Original Poster: Scott Stephens <Scott2-at-mediaone-dot-net>
>
>I measured an ignition coil and found the following:
>
>Core in:
>primary:
>L=6mH R=1.6 ohms Q=20 (tested with .1uF -at- 3.8 KHz, sharp response)
>step up ratio=86 (measured with 600 Hz signal)
>
>2ndary:
>L=67H R=8K ohms Q=3 (tested with .1uF cap -at- 65 Hz 3dB BW; sharp lower break
>point, upper response rolls off more gradualy)
>
>K= 0.81  which I derived from (correctly? seems .3 too high for solenoid)
>Step-Up Ratio Measured (86) / SquareRoot [L-2ndry (67H)/L-pri (.006H)]

    It's in a tin can, and that changes the magnetic ciruit alot, because
it's a perfect closed loop, and they use alloy wire which is partly
ferro-magnetic.  You could easily check that with a magnet.


>
>Self tuned peak frequency response at 3KHz
>Mu-eff core = 5 (including extreme air-gap of solenoid)
>
>Core out:
>primary:
>L=3mH R=1.6 ohms Q=17
>
>2ndary L= 13H R=8K ohm Q=8.8 (c=.001uF -at- 1.6 KHz, 3dB BW)
>Peaks frequency response at 5.8 KHz, no capacitors (except meas. circ.
stray)
>
>step-up ratio= 40 (measured with 600 Hz signal)
>K= 0.6
>Step-Up Ratio Measured (40) / SquareRoot [L-2ndry (13H)/L-pri (.003H)]
>
>I have another ignition coil, that I more quickly characterized, without
>tearing it out of its can:
>
>Pri: L=1mH, R=1.4 ohms; 2ndry: L=102 H, R=8.8K ohms; step-up ratio-81,
K=0.25
>
>My measurements were made with an audio generator, a scope & DVM. I would
>hope they are within 25%. The values of "K" or coupling coefficient between
>primary and secondary are much higher than Terry reported in a post a few
>weeks back. I derived my value of K from dividing the measured square-root
>of the impedance transformation (or the step-up/down ratio) by the value
>calculated from the ratio of the primary and secondary inductance. Maybe my
>step-up ratio measurements were affected by interwinding stray capacitance.
>My measurement seem too high for K, and Terry's, around .005 IIRC, IMHO too
>low.


>
>Maybe the most significant is Q. The secondary loss is awfull. At a coils
>best frequency, 6KHz, X-L(13H) is 500Kohm; now 500Kohm/8000ohm
>(DC-resistance) = 61. So Q (unless my capacitor was crap, very unlikely)
>should be 61, not 3!!!

    When you change the capacitor's value, the Q of the circuit changes.
Q=XL/Rw   were  XL = L(F)2pi  therefore, the greater the Frequency, the
greater XL will be.


>
>Is the skin effect is kicking the hell out of the coils Q? Too bad. I tried
>using a 3/8" powdered iron stick in place of the steel core, but both cores
>resulted in a 3dB core loss. They do wonders for the inductance & coupling
>though. If you can cool a steel or ferrite core better than those insulated
>windings, it would be worth keeping, provided it doesn't saturate & suffer
>hysterisis and other losses at much higher power levels.


    Put it in a ferro magnetic can, or container again, and the thicker that
is the lower the loss will be, but the frequency response will drop as well.
I built a transformer that used a couple of cookie tins, and I put the
primary on the outside can, and the secondary on the inside, but at this
moment I cannot test it accurately.  Okay, they build the standard power
transformers with the same goal in mind, but they've worked so hard at
eliminated eddy currents from the core, by laminating the core.  The
principal of the can inside a can doesn't elinate the eddy currents, unless
your windings go onto a stack of laminated washers, and there's a second
larger stack of laminated washers that either the primary or secondary
windings are incased in.  The third stack would contain the both the primary
and secondary windings, and the keepers for the cores would be made of cut
stips that when layed side by side like planks would make a circle, then
each layer would be on a different radian.  To avoid all of this, you can
use powdered iron, but who's going to make the forms?  Unless you can buy
powdered iron by itself, and mix it with fiberglass resin to a point that
you can kneed it like bread dough, before you add the hardener I doubt it.
You would need to have plaster, or concrete{No Sand} molds of your fittings,
before you could attempt that.  Wear heavy gloves fiberglass resin isn't
good, but I would use a hard coat, or finish coat, because those are the
thinnest resins, and most likely to remain workable.  You can also make your
own iron oxide, using ultra fine steel wool, a platic bucket with a lid, an
airstone, an aquarium air pump, a unpainted but clean red clay pot for
plants without the drain holes, 1 gallon of water, and 1/3 cup Chlorine
bleach.  The red clay pot is used to filter out the oxide from the liquid
once you left it in a plastic bucket with a 1/4 inch hole in the lid, and
the air stone in it with the aquarium air pump on for one week to digest.
The oxidization results will appear as though they were instant, but
literally they are not, and iron sediments and fibers will still be on the
bottom of the bucket in the liquid.


>
>So how many watts can it dissipate? I'll guess around 100 Watts in air for
>30 second runs, and 1000 Watts in oil. Now if I fig'r right, the envelope
>period = 2*Q/2*PI*F, so at 6 KHz, 2*(9)/2*PI*6KHz = 500 micro-seconds. So
>you can let a car coil 'ring up' for half a millisecond at 6 KHz, before
you
>reach a point of diminishing returns, you can integrate 3 cycles of 6 KHz
>energy. You want tight coupling to transfer energy quickly, before the coil
>burns it.
>
>This is significant for me, as I would like to pump it with a dual-scr 120
>VAC inverter, and let the coil ring up 30 KV or so across a 2nF glass plate
>capacitor, at 3KHz, about the size of a 4" cube. 100mJ for the TC, 100mJ of
>heat dissipated by the ignition coil, up to 1000 bangs/second for 100 watts
>average.
>
>My second design would use a 1200V MOT & it's 1uF 2KV cap to break-over 3
or
>4 series SCR's into a 900uH ignition coil primary. Higher energy, but uses
>bulky MOT. Can SCR's be abused in such a way (operating like a spark gap)?
>Higher output voltage (+100KV!) too.
>
>