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

Re: Measuring Coupling Coefficients


From: 	Malcolm Watts[SMTP:MALCOLM-at-directorate.wnp.ac.nz]
Sent: 	Wednesday, November 26, 1997 1:40 AM
To: 	Tesla List
Subject: 	Re: Measuring Coupling Coefficients

Hi Terry,

> From:   terryf-at-verinet-dot-com[SMTP:terryf-at-verinet-dot-com]
> Sent:   Tuesday, November 25, 1997 4:23 PM
> To:     Tesla List
> Subject:    Measuring Coupling Coefficients
> 
> Hi All,
>         I have been playing with measuring coupling coefficients lately.  I
> was just wondering if anyone has a relatively easy way to accurately measure
> this elusive factor?
>         I have tried measuring the coils in series so that the inductances
> add and then subtract and then applying the formula M=(La-Lb)/4.  However,
> due to the large difference in the primary and secondary inductances (two
> orders of magnitude) this formula can introduce substantial error.
>         I have been able to load the output with a 10K ohm resistor and use
> a 50 ohm voltage source to determine K.  However, the frequency needs to be
> carefully selected to insure an accurate reading, the math involved in the
> calculation is very complex (the compter does all the work:-)), and it takes
> substantial test equipment to perform.  So far this is the best method I
> have found and it seems to work very well but I just want to be sure I
> haven't missed some simpler method.

My way of doing it (and leaving the secondary behaving as a 
resonator): (this method assumes pri tuned to sec independently and 
the resonator grounded as usual)

     Short the gap and connect a signal generator across the primary 
cap with a resistor of 10k or so in series with the "hot" lead from 
the generator. Monitor the secondary with a scope and an aerial wire 
dangling several feet away. As you tune the generator, you will hit 
two peaks in the system response with a null between them. The null 
corresponds to the centre frequency of the system. Note that 
frequency, then note the frequencies of the two peaks. To an 
approximation, k = (Fhi-Flo)/Fnull

If you only get a single peak at the system frequency, your coupling 
isn't nearly tight enough to produce sparks.
    If you want more bang for buck (pretty hard on the primary 
components), you can try tapping in more primary inductance. You will 
notice when you do the frequency sweep that the lower peak is 
enhanced at the expense of the upper. This is known as offset tuning.
With a static gap system, this causes more erratic operation but the 
sparks are longer. It is probably better suited to a sync RSG.

Malcolm