New to the forum, Long time believer in OU

[mscoffman]

Forest,


You should do a bit more analysis IMHO. RF frequency forces any RF current to flow in the surface of a wire conductor.
I'd wonder about the distribution of current there because if more flows in the surface the conductor, it could overheat
and create surface oxidation in the heater element, causing the surface to oxidize and perhaps flake off. Unfortunately
this is known as Skin Effect and wikipedia drops out of analytic mode before it gets to nichrome, because it has human
shock effects of the same reason.


I think would want to find a cheap slow power diode that can down convert AC at your frequency. Perhaps one would
suffice, which doesn't cause the final amp circuitry to increase in temperature.



:S:MarkSCoffman

[MarkE]

Hmm..I've got 1096W real power at 46,9khz from your formula. Why such difference of 96W ?

The formula is correct:  I get 997.3W woith 26.5 Ohms and 997.1W with 26 Ohms.  You may have rounded pi, or the result of your arc tangent calculation.

X_L = 2*pi*46,900*90E-6 = 26.521 Ohms
X_L/R = 2*pi*46,900*90E-6/26.5 = 1.0008.
theta = atan(2*pi*46,900*90E-6/26.5) = 45.02 degrees.
P_APPARENT = 230²/(2*pi*46,900*90E-6² + 26.5²)^0.5 = 1.41098kW
cos(theta) = .706822
P_REAL = cos(theta) * P_APPARENT = 997.3W

[MarkE]

Forest,


You should do a bit more analysis IMHO. RF frequency forces any RF current to flow in the surface of a wire conductor.
I'd wonder about the distribution of current there because if more flows in the surface the conductor, it could overheat
and create surface oxidation in the heater element, causing the surface to oxidize and perhaps flake off. Unfortunately
this is known as Skin Effect and wikipedia drops out of analytic mode before it gets to nichrome, because it has human
shock effects of the same reason.


I think would want to find a cheap slow power diode that can down convert AC at your frequency. Perhaps one would
suffice, which doesn't cause the final amp circuitry to increase in temperature.



:S:MarkSCoffman

1) RF penetration depends on the conductivity / resistivity of the conductor.  In a high resistance material like nichrome, the fields penetrate much further than they do in a high conductivity material like copper.

2) Heat is released through the surface, not the core of the conductor.  The interior to exterior temperature of the conductor is worse at low frequency when more power is dissipated in the core of the wire, and then that heat has to travel to the surface where the heat sink (air) is.

3) The slower the diode the less efficient the diode will be, turning the diode and the switching transistors into big heaters.

[forest]

MarkE


One more question : rectifying such high frequency current with super fast diode bridge will give me rectified AC. Will it be enough to avoid inductive reactance due to coiled heater element ?

[MarkE]

MarkE


One more question : rectifying such high frequency current with super fast diode bridge will give me rectified AC. Will it be enough to avoid inductive reactance due to coiled heater element ?

Inductance is forever.  Inductance resists any change in the flow of current.  If you chop DC or AC, the resulting signal has the chopped frequency content as well as the original DC or AC.  If you rectify and filter then you can remove the chopping frequency content.  So if you do not filter, as you can see from the graph I published, you will get less maximum power from the same heaters by chopping at any frequency above about 1kHz.

I know you are set on chopping.  That adds a lot of cost and it is not at all clear what value you expect to get for that extra cost and complexity.  If you are heating stuff, the thermal inertia of the oven/stove and material you heat makes a huge low pass filter.  Skipping AC half-cycles is the tried and true: reliable, safe, and low cost way that people control resistance heater driven stoves and ovens.  The circumstances where you would want something with faster response are limited to where you need either ultra-fine resolution, or have a small thermal mass in both the heaters and the things you are heating. 

You should be wary about building big power electronics without first acquiring a fair amount of skill and experience.  An AC half cycle skipping scheme requires only large SSR's with very modest spike protection and appropriate fusing for the power section.  A chopper is many times more complex.  If you want to learn about  chopping I recommend that you make your first project something that operates below 42V and less than 50W power.  230Vac can kill you straight away.  If you make a circuit design or construction mistake there is considerable risk of: electrocution, explosion, and fire.