1998 Mazda Protege Rough idle when brake light switch activ

This is a strange one. At an idle, if you turn the headlights on, does the rpms do the same thing? If the engine idles and accelerates normally without the brake being pressed, I have to believe the TPS is set right. You may want to check battery voltage with the motor idling and then with the brake depressed and see if there is a difference.

Other electrical loads do not affect idle. I can turn on the headlights, rear defroster and AC and the idle compensates for the load like it should. I will try checking battery voltage at idle with and without the brakes depressed.

OK, let me know what you find

I checked out the battery voltage under several conditions and this is what I found:

Engine off (cold outside below feezing): 12.6v
Engine on (cold) with brake depressed: 14.8v
Engine on (cold) with brake released: 14.81v
Engine on (warm) with brake depressed: 14.8v
Engine on (warm) with brake released: 14.83v

It does not seem to have much effect on voltage. Of course once the idle drops low for several seconds, battery voltage measures around 12.6v (the alternator isn't turning very fast at that point).

I also turned on the headlights and it did not have much effect on voltage or idle. The voltage does seem a bit high, shouldn't it be around 13.8v?

Any other ideas? I'm stumped. Is there a better way to test the IAC other than a resistance test?

Battery voltage should run from 13 to 14.5 volts, so you are right where you should be. Try pulling the vacuum line off the brake power booster and plugging it. See if it makes a difference. Keep in mind, you won't have any power assist to your brakes.

I disconnected the brake booster vacuum line and plugged it. The idle behaves exactly the same as before. Anything else?

Also, in cold weather lately the idle has sometimes hunted at idle, dropping and returning to normal every 1-2 seconds. This occurs at a stoplight without the brakes depressed. Very strange. If I depress the brakes at idle RPMs drop very low and stays there until I release the brake pedal. If I leave the brakes depressed and allow it to idle rough for too long (about 12-13 seconds) I get a check engine light and an "idle control system failure" code. The code clears itself after a few hours of driving.

Judging from your description, I would suspect your idle air control valve is acting up. This a little motor controlled valve that compensates for different load demands on the engine at idle. ie, ac compressor, electrical and changes in vacuum. It is located on the intake manifold and may just need to be cleaned. An ordinary can of carb cleaner will usually do the trick if you remove the valve.

Hi, thanks for the info. I did recently remove the IAC and clean it. (there was a little carbon in it.) I also tested it for proper resistance and it passed. I have considered replacing it to eliminate that possibility, but it is a very expensive part. Is there a more comprehensive test I could perform to verify if it is good or bad? I would hate to replace a good part for $300.

Do you have, or have access to, a fax #? If yes, I can fax you the test procedures for your IAC and EGR and some diagrams.

That would be wonderful. Please fax to 801-852-6965 (attn: Alex)

Thanks for the help.

Fax is on it's way

Hi Jim,

Thanks for the fax. I've looked it all over, and the parts described in the fax do not quite match what I have under the hood. For one, the EGR is electronic and has no diaphram it in and no vacuum line attached, just a screw type valve that opens and closes between air ports. There also appears to be no EGR postion sensor, but there is a "boost sensor" that appears to have similar terminals to the drawing. (It is marked "Hitachi Boost sensor, Z5A6). When I apply vacuum to this boost sensor, there is no change in resistance across A,B,C terminals in any combination. This sensor only has one vacuum line connection that attaches to another device that appears to be a vacuum solenoid of some type. This car is strange since it never quite matches the manuals and documentation I have. Could the boost sensor be the culprit?

Thanks,
Alex

Just a follow up to this strange problem, I finally found the problem and everything is running normally now. When I bought the car used, someone had installed and then removed an after market security system. Where they tapped into the tail light circuit, the connector rubbed loose and was contacting the frame of the car. For several months I had problems with fuses blowing. After fixing the problem and still having no lights, I noticed that the wire going from the fuse block to the marker lights was melted. I bypassed it and didn't think of it again until today. That melted wire is still inside a large bundle of other wires. I suspected that where it has melted it was contacting several other wires in the bundle. I tore it apart this morning and found that this was indeed true. What I had was an old burned wire that had burned into several surrounding wires and was causing shorts between several wires. This explains the erratic idle behavior I believe. I test drove it after repairing the wiring and all seems fine! Weird problem, weird solution.