If you swap out the factory exhaust manifold for a header that eliminates the pre-cat, will it cause the ECU to throw ant code ?
I'm thinking of getting the AWR mild steel cat and eliminates the pre-cat .

 

It can. You need to remount both O2 sensors in the exhaust system after installing the headers. You should mount another cat in the exhaust system somewhere behind the headers, and place one O2 sensors before the cat, and one after. A catalytic convertor by itself is not necessarily a power restriction. The stock exhaust manifolds short runners, and primary cats location, are the problems with our stock exhaust. The AWR header is not available in a streetable version yet. Tony is currently working with Mazda to produce one.

 

From what I know about OBDII systems it's only the second sensor that'll know something is up, and it may not notice at all. Even if it does notice it might be an intermittent light. Either way the second sensor has no input to the fuel mixture control. It's just there because the EPA want's a second opinion of how everything else is working.

It is way too easy to remove it and never see the light come on, assuming you have 4 wire O2’s – I haven’t looked yet. The actual sensor part is two wires that generate its own voltage in the presents of oxygen when at temperature. The ECU monitors that voltage for leanness of burn. The other two wires are for a heater. So all you need to do is use a pair of resistors (Radio Shack $.05 each) as a voltage divider on the heater supply voltage to send the ECU the voltage it expects.

When I get around to it I plan on measuring the supply voltage and the O2 sensor voltage so I have a baseline record of what the ECU likes. This also determines the resistor size for the divider. Summit sells these ready made cheaters on a connector for $50 each, like it takes a rocket scientist.

 

I did it. Removed the secondary cat on my 99 Pro with the Bosal Federal exhaust system and pushed out the catalyst material from the primary cats housing. Check light came on after a few days then I got the OBD II cheater from Caspers Electronics in IL for imports got the link off of Summits site. Everything is working fine now. FYI the exhaust system did not increase in loudness/volume from the cat removals. Did get better pickup, I guess cause engine breaths a lot easier now. Gas mileage seems a little better too. I'll hopefully be getting a header some day, I'm 1.6er.
Later

 

The ECU monitors the voltage from each O2 sensor, and the difference between the two O2 sensors in order to verify the cat is working properly. The ECU will turn on the check engine light, and set a code for incorrect operation of the catalytic converter, if the cat is removed from between the sensors.
The 3rd gen Protege does use two heated O2 sensors (four wire) the voltage on the two wires to the heater element should be the same as the charging system/battery voltage (~12volts). The OS sensors on the Protege are 1 volt systems. They output from 0 to 1 volt based on the amount of oxygen in the exhaust. The second O2 sensor should output a lower voltage than the first if the cat is working properly.
I have yet to hear of someone actually using resistors to reduce voltage and successfully fool the ECU. It's been a while, but unless all my knowledge of electronics fails me, resistors restrict current flow. They do not change the potential energy (voltage) applied to a circuit. A transformer, with different length windings is normally used to increase or decrease voltage. Same way the ignition coil (dual stage transformer) works to build voltage to a level capable of creating a spark across the gap between a spark plugs' electrodes. Even if they did, whether this method would successfully prevent a check engine light is questionable, since the output from an O2 sensor varies depending on temperature, and engine condition/operation (RPM, load, state of tune, etc).
If you have the parts, try this experiment at home. Use a power supply to output 12volts DC to a breadboard. Insert a resistor, 10K ohms or more, across the terminals of the breadboard. You have a complete circuit with 12 volts of potential energy, and a load (resistor) of 10K ohms. That should keep the current flowing through the circuit low enough to prevent overheating the resistor and burning it up. Using a voltmeter, measure the voltage at one side of the resistor, then the other. They should both read 12 volts, unless you used a lower ohm resistor that burned up and caused and open circuit. Replace the resistor with a 100K ohm resistor and measure the voltage at the same points. Same voltage right? What changed? The current flowing through the circuit was cut to a tenth of the original current. Resistors are rated by colored bands on there body for their resistance, and also wattage capability that tells you how much heat they can dissipate. Just in case anyone was interested.

Volts divided by resistance equals current flow Volts / Ohms = Amps
Volts times current flow equals wattage Volts * Amps = Watts

 

For those of you interested in doing this. Here is the link for the device that Biknman used to prevent the CEL. Not as cheap as I would have hoped, but it does the trick.

 

you can drop voltage with resistors, its called a 'voltage divider circuit' which is a series circuit. current is common in a series circuit. i think the law goes something like this: the sum of the voltage drop in a series circuit is equal the source voltage. what you're reading is the voltage across the resistor. current can't change in a series circuit, it must be constant. you're thinking of a parallel circuit.

 

I've worked in OBDII, Eric is correct.

 

eric's information was false