Circuit Analysis of the 1.8kW Induction Hotplate

COOL THING #4 – Lazy man’s freewheel allows low-voltage IGBT turnon.

Now we can get to the really cool stuff.  When SWREF falls past TOPREF and the comparator trips, it’s output voltage goes low.  This pulls one side of the cap low, which initally pulls the other side of the cap down with it.  But the 62k pullup resistor begins charging up that side of the cap, which presents a rising ramp voltage to the noninverting input of the second comparator.

Now think about where we are now.  The SW node fell and the first comparator tripped and pulled it’s output down to 0.  The series cap begins charging and the (-) input of the second comparator is a ramp voltage, rising up from 0 in an R-C waveform.

So if we apply a DC voltage to the other input of the comparator, say 1v or so, the ramp will start below it and eventually rise above it.  The second comparator output will initially go high (when the DC voltage is above the ramp) but later it will go low (when the ramp rises above the DC voltage).  Therefore the output of the second comparator is a pulse whose duration is adjustable by changing the DC voltage.  Brilliant!

Fig 4 – Ramp Generator Feeds the Pulse Generator

Fig 5 – Simulation Traces of the Ramp Generation and Pulse Generation Stages

There you go, you’ve got an adjustable pulse that is automatically triggered once SW falls low enough to safely turn on the switch.  Just feed this pulse to a gate driver to ACTUALLY turn on the switch, and you’re in business!

COOL THING #5 – Built in PWM generator!

That one is worth discussing a little more.  It’s really quite cool that the circuit detects is own resonance, and even cooler that it can generate an adjustable pulse for gate driving so simply.  Much respect to the dude that first made a quasi-resonant PWM modulator out of a dual comparator.   Let us know if you’ve seen it before this, we’d love to see the first ever publication of this little chunk of genius!

On to the last circuit element – the gate driver.  This one also has a pretty sweet advantage in that since the SW node is already low and is going to stay low, there’s no rush in turning on gate of the power switch.  Therefore, we don’t need to use a fancy 5A driver IC and we can just drive the gates with stacked NPN/PNP emitter followers.

In this configuration, the NPN is on the TOP and the PNP is on the bottom to take advantage of a strong emitter to pull both up and down.  The base can be driven relatively weakly (10-20mA) and the gates will be driven with a half-amp to an amp, which will get them up and down pretty quick.   Not fast enough for any heavy duty hard switching, but in the resonant (or quasi-resonant as the case is here) mode it will be good enough.  And only takes two transistors!

This entry was posted in Circuits and tagged . Bookmark the permalink.

108 Responses to "Circuit Analysis of the 1.8kW Induction Hotplate"

Leave a reply