Can anybody tell me how to set the throttle position sensors as its getting anoying at traffic lights with the engine reving on its own?

thanks

 

http://fc3s-pro.com/TECH/HOWTO/TPS/tps.html

OR

Search button at the top.

 

I had the same thing going on when I bought my car try the screw at the top of the VDI first. If that does not woik do a search there are like 10-50k threads on it I did a search on that a while ago.

 

He has a TII.


Is it a S4, or a S5? Cause the method that hybrid posted won't work for a S5, cause the plug is different.

 

Its a 1990 matey

The pictures on the FC3S site on how to do not match my TPS mine has two sensors to the throttle

thanks

 

Ok, heres how you do it.

Either buy a tester from mazda, or go to radio shack and buy 2 12v low wattage LED's. Twist the read lead from each LED together and crimp a male spade connector to them. Then attach one male spade to each of the black leads individually.

Somewhere in your car there is a bright green 3 prong connector. The service manual says it is between the airbox and the passenger side strut tower, however, mine was near my solenoid rack, right near my tps. This connector has 3 prongs. 2 going parallel to each other, and a 3rd going perpendicular to the others.

Warm up your car for 10-15 minutes, until it is fully warmed up and operating temperature. Shut the car off, and plug the tester lamp you made into the plug. You will want to put the black leads into the 2 parellel sockets, and the red leads into the other one. Turn the key to the "on" position (do not start the car). If both lights or neither light light up, then you need to adjust the tps. If only one is lit up, then it is in adjustment.

To adjust it, pull the intercooler off, and locate the screw underneath the TB. Turn it counter clockwise until it is lose, but don't completely remove it. Then slowly turn it clockwise until one light lights up. It should not be more than 1 full turn until the second light turns on. Once both lights are lit up, turn it back the other way about a quarter turn until only 1 light is lit up.

Put the intercooler back on, remove the test lamp, and drive the car again. Come back and repeat the test with the lights, because it is likely that the car cooled down in the time you were figuring this out. Put the lamp back in, and see what happens with the lights. If it needs adjustment, pull the intercooler and try again.

 

thanks matey I will go and get the equip. needed

 

A lot of people use the two lights to set their TPS, but don’t realise what they’re actually setting it to. If you look at the ECU wiring diagram (page 50-25 for Turbo or 50-31 for NA in the S4 FSM) the plug to connect the lights (check connector EM-18) into is simply tapped into the three wires for the switching solenoid, the relief solenoid and switched 12V. If the solenoid is on, the light connected to its wire is also on. So all the lights actually tell you is whether or not those solenoids are open or closed. When you adjust the TPS you can hear the solenoids click as the lights come on.
The TPS setting is obviously important to the correct operation of these solenoid valves when they’re being used as Mazda intended, but most people are setting their TPS for drivability resons, not emmissions, and if you’ve removed the Air Control Valve, the solenoids are completely redundant, and setting your TPS this way is pointless.
Measuring the voltage or resistance of the TPS is a better way, and you need to find the exact figure that works best for your car through trial and error. The FSM says 1.0V or 1.0k? across the TPS’s brown/white (Vref) and green/red (signal) wires at idle. I found the smoothest on/off throttle transition isn’t necessarily at these figures, but that’s where you should start tuning from.

 

Here's my experience. On my S4 TII which has ACV and everything else, I set the TPS with the lights (made a tool to adj. screw w/IC on) a long time ago before I knew of the DMM method, and it's idled pretty steady ever since.

I've been hesitant to mess with a good thing even though I'm tempted at times to try the other method. Could it be better? Maybe. Once in a blue moon the idle will vacillate/oscillate a couple times then be cool.

I just installed a S-AFC and it reads about half a Volt when warm and close to 1V cold. Hmm? That should be just as good as measuring at the check connecter huh?

That's interesting info. Jason. I didn't know that. BTW, I love that sig.

Just thought I'd throw that out there.

 

Next time you use the two light method, check the voltage at the green/red wire on the tps. You'll find that the difference b/t both lights being on, and then turning the screw to get just one light on, is 1volt plus or minus .05 volts. Been there, done that.

See the attached jpg out of the Series five manual that shows the approx output of the tps.

The online manual is at http://www.iluvmyrx7.com, and the setting of the tps is in the FUEL section.

I just use a meter shoved up the back of the green/red wire and adjust the screw until I get one volt (with a HOT engine, idling). Series four.

 

any idea how to do this on a S5? The only writeups I could find were the light method, and then a bunch for the S4, but none for the S5. And I can't find a wiring diagram for the S5 to figure it out myself from.

Do you do this while the plug is connected, or do you take it off and check it (obviously you can't for voltage)?

 

This http://www.johnr.com/tps.html plus the jpg that is attached should clarify things for a series five. Monitor pins A and E.

 

This http://www.johnr.com/tps.html plus the jpg that is attached should clarify things for a series five. Monitor pins A and E.

 

Ok, that helps.

One question. He says at the end check back to the LED method to make sure it is adjusted properly. Is that something I would still want to do since I have my ACV removed and it will effect it like NZ said?

 

No. The acv has nadda to do with it. Its just that the Relief(blue) and switching (grey) solenoids have their wires spliced into the same wires as the Green check connector.

You could have chunked the Relief and Switching solenoids and cut their plugs off, but the Green check connector will still get the ground signal from the ECU.

The bottom line is that when the TPS output is 1volt to the ECU, the ECU will send a ground signal to the Relief solenoid AND one of the sockets on the Green connector. That ground on the Green check connector will turn one of the lights on.

Ah, shoot, I can tell this isn't working out. Bottom line is that yes, the Green check connector is still active whether or not the solenoids have been dumped, shot, mangled and or pooped on, etc.

 

So therefore the LED method should be a valid method of setting the tps, correct?

How I read it is, you are trying to get the Narrow range of the TPS to read ~1v. When it is ~1v, the ECU sends a signal to the test connector, which lights up the light. Therefore: if 1 LED is lit > ECU is reading ~1v from tps > TPS is adjusted.

So if that is correct, why do people call the LED test invalid?

 

Not quite. The check connector is simply wired in parallel with the relief and switching solenoids. The ECU sends a signal to the solenoids to open, and if you have an LED checker plugged in, the appropriate LED will also come on.
All the lights are for is to indicate the whether the solenoids are open or not, that's it. The TPS setting affects when they open and close, and this is important for the correct and complete functioning of the secondary air injection system.
If you're having driveability problems, adjusting the TPS may help, but the correct settings for the emission controls may not give the best on/off throttle transition (which is what most people are tryijng to fix). Unless you need to pass strict emissions tests, the operation of the solenoids is irrelevant, particularly if they're not even on the car any more! After 10+ years things won't necessarily work exactly as they did when brand new. Try all three methods (lights, resistance, voltage) to see what works best for your car. It'll probably take a bit of trial and error.