Need help about calculating damping resistor

[javahitman]

You can, but you may still need an oscope. We have a parallel RLC circuit, and we want to achieve critical damping. This occurs when:



L is the coil inductance (measure it), so all you need is C, which is the total parasitic capacitance of the coil, cable, and circuitry (mostly the MOSFET). The cable and circuitry can be estimated, but you will need to measure the self-capacitance of the coil. This is usually done by measuring the self-resonance with an oscope.

As an example, a 300uH coil has 100pF of self cap, so

= 866 ohms

That's in the ballpark of most damping R's.

- Carl

Wonder if anyone tried using potentiometer-variable resistor on the coil or the control unit
Say if the formula gives you 866 ohms and if you put 1000ohm Trimpot than you would have some room to adjust it either way?

[Carl-NC]

Yes, people have done that. I'm personally wary of putting a trimpot on a flyback, but the worst that could happen is it opens up to a high impedance.

[javahitman]

Digital trim pot

If it's microchip based PI maybe could use a digital trimpot which has a higher resolution
And can possibly add push buttons for adjusting and display values on an LCD

[Tepco]

Something didn't look right. Mikebg's and Carl's formulae did not look like they would jibe. But when I made comparative calculations using what I thought were realistic variables, the two formulae agreed to within about 1%. So I set out to try and prove equivalency.

[I kept getting negative results until I realized that does not equal (arrooo).]

After working the proof below, I see that Carl's and mikebg's formulae are approximately equivalent. Mikebg formula includes an error term and is slightly more accurate. But that does not make Carl's formula wrong, only simplified.

I refreshed some basic algebra skill, and got a lesson in LaTex all at the same time.
(and it's time for a break)

This way you never get correct dumping resistor value, not even close. Circuit is all but critically dumped RLC. Consider cable capacitance ( lumped in this case), usually larger than coil self-capacitance (distributed) then shield capacitance (also distributed). Then power switch (MOSFET or something), large highly nonlinear capacitance vs. voltage during flyback, like varicap. Then, at some point, flyback will hit transistor breakdown voltage, introducing another (nonlinear) resistance in the circuit and prevent energy release (coil is overdumped in this phase). With two nonlinear parameters simple RLC circuit calculation is not even nearly correct and will always produce wrong results. Use scope and variable resistor to adjust it right. Then, if you change anything in the circuit, pulse width (affecting energy stored in the coil and avalanche time), transistor type (affecting capacitance and breakdown voltage) etc, do it all over again.
Even then, some metal inside coil (coil itself) will have some finite eddy current decay time, depending on actual construction…Not exactly exact science, like standard model…Measure, observe actual waveforms, don’t calculate anything! Or you will be off at least 30%, probably much more. Formula in #11 will not hold in reality, just close.

Time variable dumping resistor (MOSFET) is patented at least few times, i never get any benefit of using it, always can be done using different, less complicated method. Lower capacitance transistor, high voltage one to prevent breakdown using short pulses, to reduce stored energy, later compensated by increasing frequency etc…Interesting machines to play with.