Quick question about this h-bridge circuit

**broken link removed**

1. What are the 100 Ohm resistors for? I understand that with a 100/10k ratio, the voltage drop is minimal, so what is the purpose?

2. I understand why the P-Type FET's need 12V at the gate, but why do the lower N-Type's at the bottom require a 12V gate voltage?

All Power FETs (which are not sold as Logic-Level FETs) , be they N or P, require a gate drive of ≥10V. N-Fets turn on when the Gate is 10V more positive than the source; P-fets turn on when the gate is 10V more negative than the source.

That is a very poor H-Bridge circuit. The author that posted it is very naive. To turn on either the N-Fet or the P-Fet, the current required to charge the gate capacitance must come through the 10K resistor. The 10K and the gate capacitance forms an RC filter where it takes several ms for the device to turn on, during which time the FETs are Dissipating up to a tens to hundreds of Watts.

To rapidly turn on/off the Fets, which is what you want to do to minimize their power dissipation, the gate driver circuits must be capable of sourcing/sinking up to a good fraction of an Amp. The whimpy 10K resistor can only source a max of 12/10K = 1.2mA. The gate is discharged through the 100Ω resistor, which can sink 11/100, about 0.1A, which is better, but still falls short.
This is why an IC H-Bridge gate driver interface chip is a good idea.

What's the switching frequency?

If it's only a couple of hundred Hz I don't see the problem, the time constant will only be 15µs for a MOSFET with a capacitance of 1.5nF.

How much current does the motor need?

The problem with the circuit is that the MOSFETs turn ON fairly quickly (through 100 ohms) but turn OFF very slowly (through 10k ohms). During the overlap time, a shoot-through current flows which is limited only by the resistance of the battery, the MOSFETs, and the wiring. This current is far beyond the maximum rating of the devices.

Problem is made worse because breadboard conditions (high interconnect resistance) tend to conceal this problem. When you build a proper PCB with good connections, they blow up without explanation.

The controller should be able to mitigate any shoot tough.

It the controller goes wrong and makes both outputs high simultaneously there will be brief interval when both MOSFETs are on but I don't see how a short surge, lasting for about 50µs will cause any damage.