Posted by Dave Furnish on May 09, 192001 at 02:45:02:
In Reply to: Re: To much voltage posted by Mike Gray on May 09, 192001 at 01:07:15:
: : I made and inverter circuit with a 1.7 A 240V to 6V transformer reversed and driven by a 555 ocillator coupled to a MJ 10012 transistor fed by a 6V battery.
: : I expected to get a 240V reading on my multimeter, but what I got was a high voltage arc at the test probes. Needless to say that was the end of the meter. Across the terminals of the transformer I get a strong arc of 3- mm in length which would suggest thousands of volts.
: : So much for simple logic. Any suggestions on how to get rid of the unwanted cefm?
: : Thanks for listening.
: Hi Ned,
: The 555 timer produces a rectangular waveform. Whenever you drive an inductor with a non-sinusoidal waveform, you will end up with much higher voltage than expected! Your circuit is switching the voltage on the primary of the transformer on and off, just as if you were doing it with a mechanical switch. The sudden build-up and subsequent collapse of the magnetic field will produce very high voltage spikes in the secondary winding. You can swamp much of those spikes with a resistor/capacitor network across the seconday winding. Values will have to be selected that are large enough to reduce the hv spikes, but not so large as to put too much load on the power output circuitry. I don't know how to calculate the values. Mayby someone else can help here.
: Regards,
: Mike Gray
Here's one way of figuring C for RC across sec. - With secondary open circuit apply normal voltage to primary and measure primary current. From this, energy stored in the primary inductance can be figured by Z=V/I, Z is approx. XL and XL=2pifL. Energy stored in the primary is around (1/2) LI^2 joules. To match energy stored in winding, and hence limit Vsec to Nsec(Vpri)/Npri, make C such that capacitive energy is equal to inductive energy - (1/2)CV^2=(1/2)LI^2. This may be overkill for you, but might be a starting point. First cut at R can be rather arbitrary, keep low to force most voltage across C. If switching frequency is up there, attention must be paid to power rating, as Ic can be significant.
Snubber capacitance will be reflected to the primary circuit by the turns ratio squared - if snubber C = .01 uF, reactance seen on the primary side would be approx. .01 uF * (240/6)^2 or approx. 16 uF. Current through the primary switch can get quite high if another circuit isn't used to limit dV/dT.
Hope board usage increases, 'cause stuff like this can cause us all to think a little more, take a little more time when considering current precepts about design engineering... "we've always done it this way..." anyone ever got trapped into that mode of thinking? Yeah, history is great (I've got 6 binders full of info from the '40s - '60s which is invaluable), but sometimes independent thought is useful, if not required, to make any progress. A lot of what goes on has been done before... many times. Reinventing the wheel is OK I guess if someone makes money on it...
Any feedback on this idea? All magnetic materials reach a Curie temperature, beyond which they don't support a magnetic field. And V is proportional to d(theta), theta being magnetic field. Does this work to generate electricity?
Sorry for the rambling -
Dave